This screen looks at where U.S. solar stands today, with a focus on utility-scale and commercial solar, the demand and supply picture, costs vs. competing power sources, recent policy changes, the bull case and risks, and three stocks I think offer attractive exposure to the theme.

Disclaimer: Nothing posted by P14 Capital should be considered financial advice. The author of this post may hold positions in the tickers discussed. Please consult a financial advisor and/or conduct your own due diligence before making investment decisions.

Modern photovoltaic solar traces back to Bell Labs’ practical silicon cell in 1954. High costs led to early applications being concentrated in satellites and remote power systems, but this changed drastically during the 2010s. Manufacturing scale, particularly in China, pushed module costs sharply lower as cell efficiency improved. The installed cost of U.S. utility-scale solar fell ~75% from 2010 to 2023 and the panel has since become a mass-produced component, leaving a larger share of project cost in trackers, inverters, electrical equipment, labor, development, interconnection, and financing.

The supply chain spans polysilicon → ingot/wafer → cell → module → inverter/tracker and balance-of-system equipment → EPC/developer → utility or IPP owner → power buyer. Polysilicon, wafers, cells, and crystalline-silicon modules are highly sensitive to manufacturing utilization and trade policy. Trackers and inverters offer greater product differentiation, and development depends on securing land, permits, transmission access, interconnection, and an offtake agreement.

The IEA estimates that China holds ~85% of the broader solar supply chain and ~95% of wafer capacity. This manufacturing buildout has created excess supply, with global cell and module factories operating at only ~50% utilization. China is by far the largest demand market, installing ~370 GW of PV in 2025, while the EU, India, and the U.S. form the next major tier of markets.

U.S. manufacturing has grown fastest at the final module-assembly stage. Domestic module production reached ~70% of 2025 U.S. installations, while operating cell capacity was only 3 GW entering 2026. U.S. module factories are still dependent on imported cells and wafers. India, Indonesia, Laos, Malaysia, Thailand, and Vietnam supplied 78% of U.S. cell imports in 2025.

The U.S. finished 2025 with 279 GWdc of solar capacity after installing 43.2 GWdc during the year. Q1 2026 added another 7.8 GWdc, including 5.9 GWdc of utility-scale and 523 MWdc of commercial solar. Solar supplied 60% of all new U.S. generating capacity added during the quarter; solar plus batteries supplied 91%. In May, total solar generation reached 47.1 TWh, or about 12.9% of U.S. electricity generation, exceeding both coal and wind for the month.

Wood Mackenzie expects total U.S. solar capacity to reach 769 GWdc by 2036, up from 279 GWdc at year-end 2025, representing a ~9.7% CAGR in cumulative installed capacity. Utility-scale solar will drive most of this growth.

Supply and Demand

BloombergNEF estimates that 95 GW of modules are already in inventory, enough to meet demand through 2027. So, 30-40 GW of annual installations does not require another subsidized manufacturing boom. The key bottleneck lies in the grid interconnection queues.

Wood Mackenzie expects 211 GWdc of utility-scale solar additions from 2026 through 2031, and contracting has already strengthened alongside higher power demand. Developers signed 6.3 GWdc in Q1 2026, +15% Y/Y, led by Texas. U.S. electricity consumption is entering its strongest growth period in decades as AI infrastructure, semiconductor plants, manufacturing reshoring and electrification add load, while retiring generation still needs to be replaced. NERC’s latest assessment now projects North American summer peak demand to rise 224 GW over the next decade, vs. only 55 GW in its 2022 assessment, with data centers responsible for much of the increase.

BloombergNEF expects 71 GW of additional U.S. data-center capacity through 2030, with 57 GW, or 80%, concentrated in Texas, PJM, MISO and the Southeast. These regions also account for much of the solar development expected to move forward after the current federal tax-credit pipeline begins to wane. For data-center developers, power availability, timing and price matter much more than the generation source itself, which favors technologies capable of adding large amounts of capacity within a relatively short construction window.

Natural gas is a credible alternative, but turbine availability has become a major constraint. GE Vernova has a backlog of 116 GW of gas-power equipment and slot reservations and expects annual turbine output of only ~20 GW in 2026, 24 GW in 2028 and 30 GW by 2030. Reuters reported in June that new turbine deliveries can take around 5 years, while large solar-plus-storage projects can be built in ~18-20 months once development and interconnection work is complete. The crucial point here is that the core supply of modules is already secured, unlike with natural gas turbines. Solar projects can still spend years waiting for permits or a grid connection, but once those issues are cleared, construction is considerably faster than for a new gas or nuclear plant.

Cost also reinforces solar’s advantage over other energy sources. Utility-scale solar costs ~$40-$98/MWh ex-subsidies, vs. $51-$129/MWh for new combined-cycle gas and $175-$255/MWh for new nuclear. Unsubsidized utility-scale solar + four-hour storage costs ~$61-$156/MWh, falling to ~$43-$135/MWh with the federal ITC depending on bonus-adder eligibility. This is competitive with new gas across much of the range and far cheaper than nuclear, while also having a quicker build time than both. Natural gas turbines are also exposed to greater underlying commodity price risk, whereas solar has no fuel bill.

Battery technology can improve solar further over time. LFP already accounts for ~90% of stationary battery deployments, and more projects are moving toward four-hour and longer durations. Sodium-ion is beginning to scale and could become attractive for stationary storage because it does not require lithium and energy density matters less than in EVs. Longer-duration technologies are also moving into demonstration, including flow batteries and iron-air systems capable of multi-day storage; DOE is supporting two planned 10 MW/100-hour iron-air projects. Four-hour batteries cannot cover several days of weak solar output, but cheaper and longer-duration storage would allow solar generation to serve a larger share of demand outside daylight hours. Gas and nuclear already provide stable generation, so improvements in batteries provide much more incremental value to solar.

It is likely the upcoming decades will involve a combination of energy sources to respond to the rise in electricity demand. I think solar and storage will take a disproportionate share of incremental U.S. power demand for four reasons: 1) module manufacturing capacity already exists at enormous scale; 2) solar can be built relatively quickly once permitting and interconnection are cleared; 3) utility-scale solar is one of the cheapest sources of new generation, and storage extends its output into evening demand; and 4) solar is modular and can be deployed from a few MW to multi-GW projects.

Policy

Federal policy has shaped solar for more than two decades, beginning with the 30% ITC in 2005 and later expanding through trade restrictions, domestic-manufacturing incentives and the IRA in 2022. Section 45X supported U.S. manufacturing, while Section 201 tariffs protected domestic producers from low-priced imports before expiring in February 2026. Recent policy looks negative at first glance, but the actual impact on the industry is more mixed.

The OBBBA accelerates the phaseout of tax credits by requiring utility-scale and commercial projects to have begun construction by 07/04/2026 or be placed in service by 12/31/2027 to retain Sections 45Y and 48E. As expected, most developers spent 2025 and 1H26 qualifying projects ahead of the deadline, leaving a large backlog still eligible for the prior credits. Research identifies 110 GW of eligible solar through 2030, with 76% expected to commission in 2026-27, which will keep near-term utility-scale installations supported even as the broader 2026-30 solar forecast was cut 27%, from 251 GW to 183 GW, following OBBBA.

On the other hand, domestic solar manufacturers are likely to experience a pricing tailwind as the new Section 232 regime sets a $0.38/W minimum import price for modules and adds tariffs across the solar supply chain. Research estimates some imported modules could cost ~$0.48/W, but the ~95 GW of modules already in U.S. inventory should delay much of the impact until ~2028. So, 2028 is the period when much of the subsidized backlog begins to roll off, but this does not mean the solar industry will cease to exist or that growth will turn negative. There are few alternatives that can match solar + storage on both cost and scalability. Utilities and data centers are likely to gravitate toward solar as they look for cheaper generation that can be deployed at scale and quickly. Rising electricity demand should offset part of the impact from expiring solar credits.

The U.S. needs to win in AI, and for that to happen, the power constraint needs to be solved. Solar and storage are likely to be a major part of that solution.

Catalysts and Risks

The fundamental catalyst for solar is electricity demand. The near-term pipeline is already large enough to support years of utility-scale construction; data centers and manufacturing projects now need to turn today’s load forecasts into actual consumption. Faster interconnection approvals, additional transmission and higher power prices would allow more projects to move from development into construction.

In trading and investing, one must separate industry fundamentals from equity flows. The short-term equity catalyst is the midterm elections. I think U.S. listed solar equities are highly likely to experience bullish flows as the midterm election volatility kicks in. The industry has clearly been out of favor, reflected in the underperformance of the Invesco Solar ETF.

I think policy has become as unfriendly as it gets. If electricity demand grows broadly as expected, I think much of the policy risk is already reflected in solar equities. A Democratic gain in either chamber would support a re-rating and positive flows as investors price a lower probability of further policy tightening and better odds of a more supportive policy environment after 2028.

The largest risk is that electricity demand comes in below expectations. Data-center queues contain projects that will never be built, and the EIA has already revised Texas load expectations lower as proposed projects face greater scrutiny. Cheap natural gas, faster gas-turbine deliveries, higher financing costs, transmission delays, permitting restrictions and continued interconnection congestion could also slow installations.

Equity risk is different for each company as global module oversupply can compress manufacturer margins while lower panel prices benefit developers. Tariffs can support domestic manufacturers while raising costs for developers. Higher steel costs can pressure tracker margins, while lower PPA prices can squeeze project returns. Solar deployment can grow materially without every company in the value chain benefiting equally.

If you are willing to buy and hold for a quarter (or more), these solar stocks should be at the top of your watchlist.

The Watchlist

FSLR - First Solar, Inc.

Price: $204.46 | Mkt Cap: $21.9B | EV/FY27E EBITDA: 6.1x | FY27E EBITDA Margin: 56.3% | EV/FY27E Sales: 3.5x | FY27E Sales Growth: 16.9% | FY27 P/E: 8.9x | Consensus PT: $280

FSLR is my top pick in the U.S. solar space and the de facto domestic solar champion. The company makes money primarily by selling Series 6 Plus and Series 7 CdTe thin-film modules under long-term supply agreements to utility-scale developers, IPPs and utilities, with 96% of 2025 sales tied to the U.S. market. CdTe avoids the crystalline-silicon wafer and cell chain used by most competitors and performs better in high-temperature conditions. FSLR is also rolling out CuRe, or Copper Replacement, which replaces copper in the module’s semiconductor structure with alternative dopants to reduce long-term degradation and increase lifetime energy output. Q2 ended with 45.1 GW and ~$13.6B of contracted backlog through 2030, with fully integrated U.S. production substantially committed through 2028.

I think the market is overstating FSLR’s exposure to OBBBA. The earlier phaseout of 45Y/48E can slow new utility-scale projects later in the decade, but FSLR’s 45X manufacturing credit stays at 100% through 2029 before stepping down to 75%/50%/25% in 2030-32. FSLR expects $2.10-$2.19B of 45X credits in 2026, and Q2 gross margin reached 57% with a 61% adjusted EBITDA margin as U.S. production increased. PFE rules and Section 232 also favor FSLR’s domestic CdTe supply chain. Recent U.S. bookings came in at ~$0.36/W including adjusters, versus a $0.302/W average backlog ASP, and the new trade rules are expected to push competing crystalline-silicon modules into the low-to-mid $0.40/W range. FSLR therefore has room to sign new contracts at higher prices without giving up its competitive position.

Shares have given back the post-Q2 and Section 232 gains even as FY27 has become more attractive. FY26 sales guidance of $4.9-$5.2B reflects lower international Series 6 production and technology-upgrade downtime, masking higher U.S. volumes. Louisiana is ramping, South Carolina is coming online, and U.S. nameplate capacity is expected to increase from 14.9 GW in 2026 to ~17.1 GW in 2027. Bloomberg research sees FY27 sales approaching $6.5B, implying >20% growth and ~10% upside to the consensus estimates as backlog converts and new capacity ships. At 8.9x FY27 EPS and 6.1x FY27E EBITDA, I think the current multiple gives too little credit to the 2027 volume and pricing ramp. Q3 bookings are the next catalyst after Section 232 raised the price of competing imports, with any positive midterm reaction providing additional upside.

NXT - Nextpower Inc.

Price: $85.44 | Mkt Cap: $12.9B | EV/FY27E EBITDA: 12.8x | FY27E EBITDA Margin: 21.2% | EV/FY27E Sales: 2.3x | FY27E Sales Growth: 21.2% | FY27 P/E: 18.9x | Consensus PT: $142

NXT is another solid pick in utility-scale solar. The company is the world’s largest solar tracker supplier, with ~30% global market share. Its systems rotate panels throughout the day to increase energy production and are sold mainly to EPCs, developers, IPPs, and utilities. Trackers are more differentiated than modules because project developers prioritize reliability, wind performance, software, bankability, and lifetime energy output. NXT has also expanded into foundations, electrical balance-of-system equipment, power conversion/inverters, battery storage through Prevalon, TrueCapture software, and fixed-tilt systems through Zimmermann. Non-tracker products reached 14% of FQ1 revenue, eBOS revenue doubled Q/Q, PowerMerge backlog has now passed 2 GW, and Prevalon added more than $300M of storage backlog. Management wants non-tracker products to reach 32% of sales by 2030 and sees its product set covering ~27% of U.S. utility-scale project CAPEX vs. ~9% from trackers alone.

OBBBA can slow U.S. tracker demand later in the decade as the safe-harbored utility pipeline gets built, but current orders are more than healthy. FQ1 bookings exceeded $1B, book-to-bill was above 1x, and tracker backlog increased to more than $5.5B, excluding Prevalon. U.S. tracker shipments are expected to rise from 25 GW in FY27 to 29 GW in FY30 before a larger drop in FY31, giving NXT several years of volume growth before the policy reset comes into play, assuming current policies continue. NXT also continues to benefit from 45X, with FY27 credits estimated around $367M, and international expansion reduces its dependence on U.S. solar policy. The Saudi JV has 12 GW of annual capacity and recently secured a 2.25 GW commitment for the Bisha project. At current levels, I think shares are pricing in a steeper near-term slowdown than what the backlog and booking activity are currently suggesting.

FQ1 revenue grew 8% to $935M, AEBITDA reached $233M and AEBITDA margin was 24.9%, vs. 22.0% consensus, although tariff recoveries helped the beat. FY27 guidance now calls for $4.1-$4.4B of revenue and $870-$930M of adjusted EBITDA, or a 21.2% margin at the midpoint. Current FQ2 estimates are ~$1.0-$1.1B of revenue and $210-$225M of adjusted EBITDA, and I think there is room for another EBITDA beat. Tariff recoveries are expected to continue into FQ2, PowerMerge can now ship following UL approval, eBOS is ramping and U.S. mix has been strong, partly offset by acquisition spending and 45X monetization costs. Power-conversion products are also expected to carry higher gross margins than trackers, while software can add revenue with little manufacturing cost. At 18.9x FY27 EPS and 12.8x FY27E EBITDA, I like the r/r for a company growing sales >20%, carrying more than $5.5B of backlog and selling more equipment into each utility-scale project it already wins.

SHLS - Shoals Technologies Group, Inc.

Price: $7.14 | Mkt Cap: $1.2B | EV/FY27E EBITDA: 9.1x | FY27E EBITDA Margin: 21.8% | EV/FY27E Sales: 2.0x | FY27E Sales Growth: 10.8% | FY27 P/E: 12.8x | Consensus PT: $11.6

SHLS makes electrical balance-of-system equipment used across utility-scale solar, battery storage and data-center power infrastructure. Its wiring harnesses, combiners, recombiners, disconnects and monitoring equipment move electricity from panels or batteries to the inverter and grid, with most sales going to EPCs building large projects. System solutions generated 78.7% of 2025 revenue, and more than 80 GW of Shoals equipment is under contract, under construction or already operating. EBOS is only ~2.5% of total utility-scale solar project cost, yet failures can shut down an entire project, favoring suppliers with a long operating history and proven products. Storage is also becoming meaningful, with ~$20M of BESS revenue in Q2 and PowerHub now being used across grid storage, solar + storage and data-center UPS systems.

The stock has been punished for the wire-insulation issue, weak margins and poor cash flow.Defective Prysmian wire used in older harnesses led to an estimated $73M loss, with $71.8M already incurred by Q2 and only $1.2M left in the recorded wire-specific warranty liability. Q2 gross margin fell to 30.3% as the new Tennessee plant ramped, rework costs hit the quarter and tariffs pressured input costs, while 1H operating cash flow was negative $34.6M. Demand looks much better than the current valuation suggests, with backlog and awarded orders reaching a record $801.4M, +19% Y/Y, and ~$700M scheduled for delivery over the next 12 months. OBBBA can pressure utility-scale solar later in the decade, but Shoals already has strong near-term order coverage and is adding exposure to storage and data centers, so its growth is becoming less tied to solar installations alone.

Shoals has started shipping PowerHub systems for a 1.1 GW hyperscale data-center project with ON.energy, with ~360 PowerHub systems expected across the project; the first phase is complete and additional phases are underway. More ON.energy projects were booked after Q2, and the TerraFlow agreement could support storage deployments of up to 5 GW annually, with revenue expected to begin in 2027. Q3 guidance calls for $150-$170M of revenue and $32-$37M of adjusted EBITDA, with margins expected to improve as the new factory becomes more efficient and higher-margin products contribute more. At 12.8x FY27 EPS and 9.1x FY27E EBITDA, I think the current price still reflects the wire issue and recent margin pressure more than the record order book, improving factory output, early data-center/storage growth, and future margin expansion.

With the company reporting 2Q26 earnings this Tuesday (08/11), I wanted to take a shot at bottom fishing and see whether there is enough upside to justify the risk. I have come to the conclusion that the risk/reward is attractive for a non-core position in a market-share-leading, single-category company. If the thesis works, I think the stock could double relatively quickly. If it does not, I see downside of 20%–30%.

OWLT is by no means a “fantastic business,” but it piques my interest because I have a variant view on how the business could perform over the next two years. This pitch will be relatively short because I do not have high conviction yet. I plan to use Tuesday’s print to decide whether to hold, add and make it a core position, or cut it entirely. If I decide to add, I will supplement this mainly qualitative quick pitch with a deeper thesis and a model.

Disclaimer: Nothing posted by P14 Capital should be considered financial advice. The author of this post holds a long position in the stock discussed. Please consult a financial advisor and/or conduct your own due diligence before making investment decisions.

Owlet, Inc.

NYSE: OWLT | 08/07/2026

Business Overview

Owlet was founded in 2012 and incorporated in 2014, initially developing the Smart Sock before commercial sales began in 2015. The company added Owlet Cam in 2018, Dream Sock in January 2022, FDA-cleared BabySat and Dream Sock with Health Notifications in 2024, and Owlet360 and Dream Sight in 2025. Today, Dream Sock is a $299.99 OTC infant wearable that monitors pulse rate, oxygen, activity and sleep; Dream Sight Gen 3 is a $99.99 video monitor; Dream Duo Gen 3 bundles the two for $379.99; and BabySat is a $699 prescription pulse oximeter for physician-supervised home monitoring. Owlet360 adds personalized health and sleep analytics for $9.99/month, while Owlet One bundles Dream Sock, Dream Sight and Owlet360 for $99 upfront plus $19/month with a 6-month minimum. More than 2.5M families have used Owlet across 30+ countries, and well over 500K parents purchase a new device each year. Owlet went public through a July 2021 SPAC combination with Sandbridge Acquisition Corp. at $10/share, or $140/share adjusted for the 2023 reverse split. The company is now building a broader pediatric health platform around monitoring, software, data and AI.

Revenue increased from $54.0M in FY23 to $78.1M in FY24, +44.5% Y/Y, $105.7M in FY25, +35.4%, and $107.1M LTM through 1Q26, +26.8% Y/Y. On an LTM basis, the U.S. generated $87.0M, 81.3% of revenue and +29.2% Y/Y, the U.K. $6.5M, 6.0% and +16.8%, and other international markets $13.6M, 12.7% and +17.9%. In 1Q26, hardware generated $19.8M, 88.1% of revenue and -4.5% Y/Y, compared with $2.7M from Owlet360, 11.9% of revenue and +569.3%, with subscription gross margin of 67.4% versus 52.8% for hardware. Gross margin increased from 41.8% in FY23 to 50.4% in FY24 and 50.6% in FY25, with LTM gross margin at 50.9%, down 137 bps Y/Y. Adjusted EBITDA improved from $(16.3)M in FY23 to $(1.8)M in FY24, $2.0M in FY25 and $0.4M LTM, equal to a 0.4% margin. Results are seasonal, with Q4 typically the strongest sell-in period from holiday promotions and Amazon 1P volume, followed by a Q1 step-down and stronger Q2 demand around Mother’s Day.

Owlet is the U.S. baby-monitor market-share leader, with 41% of category dollar spend in 4Q25. Its competitive advantages come from scale, FDA clearance, its pediatric dataset, brand and the integration of wearable monitoring, video and software. The regulatory path took years of product iteration across multiple generations of sock hardware, and the October 2021 FDA warning forced Owlet to stop U.S. Smart Sock sales before securing BabySat clearance in June 2023 and first-of-its-kind Dream Sock De Novo authorization in November 2023. Dream Sock’s application was supported by 9 formal clinical evaluations and data from more than 300K infants, and Owlet’s broader dataset now includes more than 17T heartbeats. Masimo, a well-capitalized competitor, received OTC FDA clearance for its Stork monitor in May 2024 but exited the product within a year, showing that clearance alone does not guarantee commercial success.

The broader smart baby-monitor market is projected to grow from $1.7B in 2024 to $3.2B by 2032 at an ~8% CAGR, and Owlet targets markets representing 33M+ annual births. Management is focused on increasing LTV through a 4-year family relationship built around subscriptions across the first two years of each child and the typical two-child family cycle.

Co-founder Kurt Workman returned as CEO in April 2026 after Jonathan Harris resigned without a stated reason. Amanda Crawford, CFO, has been with the firm since 2022. As of 1Q26, Owlet had $35.5M of cash, $6.3M of WTI principal and $13.4M drawn on its old ABL. In June, the company replaced both facilities with a $25M Wells Fargo revolver, expandable to $35M, priced at SOFR +2.00% to 2.25% versus SOFR +7.50% to 8.50% previously and maturing in June 2029; liquidity was $33.8M at closing. Workman’s current package includes a $500K base salary, 70% target cash bonus and 850K RSUs. The annual bonus program has been tied to EBITDA, and performance RSUs granted to other executives are tied to cumulative net revenue through 2028, aligning compensation with profitability and revenue growth.

The Quick Thesis

The primary reason for the YTD flush has been the sharp deceleration in top-line growth, made worse by management’s ugly FY26 guidance reset.

Owlet beat 4Q25 consensus estimates but initially guided to a weak 1Q26, which is normal given the seasonality of the business. More damaging was the initial FY26 outlook for $126M-$130M of revenue, or 21.1% Y/Y growth at the midpoint, a sharp slowdown from FY25’s 35.4% growth. Management also guided to $3M-$5M of adjusted EBITDA, implying ~120bps of margin expansion at the midpoint. The 1Q26 update then cut FY26 revenue guidance to $118M-$122M, or 13.5% growth at the midpoint. At the same time, management doubled the adjusted EBITDA midpoint from $4M to $8M, taking implied FY26 adjusted EBITDA margin expansion to ~475bps.

For a formerly fast-growing company still GAAP unprofitable, resetting expected revenue growth from 21% to 13.5% is an obvious negative on first glance. The market reacted accordingly, or at least most participants think it did. OWLT’s EV/Sales multiple has been crushed from 4.6x TTM at the close of 12/31/2025 to just 1.4x today. I think Mr. Market has focused almost entirely on the top-line reset and given little credit to what is happening underneath it.

This multiple reset is where I think the long opportunity lies heading into the print.

Let us address the short-term catalysts that show up in 2Q26, the first near-term test. Q1 is consistently Owlet’s weakest quarter following the holiday period, with lower promotional activity and a seasonal sell-in step-down. Management expects revenue to recover in Q2, followed by a slight sequential decline in Q3 and an annual high in Q4. Historically, Q2 earnings have also tended to produce positive returns on the stock.

The FY26 revenue cut is not driven by a deterioration in end demand. A major retail partner reduced inventory requirements from 8-10 weeks of supply to 4-6 weeks. This has created a broader ~$7M inventory headwind, which has already passed. Management described the Q1 rightsizing as a one-time event, so this portion of the reset has already occurred. The revised outlook also embeds a more conservative sell-through assumption for the rest of the year, along with the decision to defer lower-return markets including India, Hong Kong and Singapore and pull spending tied to lower-margin, high-burden channels. No quantitative 2Q26 guidance was provided.

Consensus expects 2Q26 revenue of $31.9M, up ~22% Y/Y versus 26% growth in 2Q25. I think both revenue and margins can beat.

Reported 1Q26 hardware revenue was hurt by the retailer inventory reset, yet U.S. Dream Sock and Duo sell-through still grew 10.5% Y/Y. Duo units increased 45% and Dream Sock increased 3%. Owlet was also the only brand growing in a weak category. Excluding Owlet, baby-monitor category dollars fell 19% Y/Y in Q1, compared with an 11% increase for Owlet.

More importantly, in the 1Q26 call, management already gave us a read on 2Q demand one month into the quarter. After optimizing marketing and retail placement entering Q2, QTD sell-through had accelerated to more than 30% Y/Y for both Dream Sock and Dream Duo. Management specifically said this strength was not incorporated into the revised FY26 outlook because it wanted more sell-through data first. Registry additions also entered Q2 strong, with Dream Sock +31% Y/Y and Duo +44% in Q1.

Management then leaned heavily into promotions throughout the rest of the quarter instead of waiting for Prime Day. The Mother’s Day campaign ran May 3-17 across Amazon, Walmart, Target, Best Buy, Babylist and Owlet.com, with $60 off Dream Sock, $80 off Dream Duo 3 and $20 off Dream Sight. Owlet also partnered with Dani Austin’s Divi brand, offering a free $48 Divi serum with Dream Sock or Duo purchases on its own site, followed by a $349.99 Duo 3 + Divi bundle. Memorial Day followed from May 22-25, with a free $99 Dream Sight included with a Dream Sock purchase. Owlet then ran 15% off Dream Sock, Duo and Dream Sight from June 1-13, another 15% sitewide Father’s Day promotion from June 19-21, and Prime Day from June 23-26. Virtually all of May and June had some form of consumer incentive.

Traffic has also been improving with Owletcare.com receiving 349K visits in May, up 36.2% M/M from 256K in April, followed by another increase to 416K visits in June. Additionally, commentary during the 06/02 TD Cowen conference suggested that April momentum has continued on to May.

Prime Day is another important 2Q26 catalyst. Interestingly, this was only the second time Prime Day landed in June, after 2021. Adobe measured $26.4B of U.S. online spending over the four-day event, up 9.3% Y/Y. Adobe also said electronics sales were +120% versus average June daily sales, while baby products saw major spikes, including strollers +195%, car seats +140%, and formula and diapers/wipes +75%. Discounts also drove purchases of higher-ticket discretionary products, including electronics, toys and children’s products. This is directly relevant to a $190-$270 promotional baby-monitor purchase.

Numerator found average Amazon order value declined to $47.66 from $53.34 and household Prime Day spending declined Y/Y. I think this makes Owlet’s promotional positioning more important. Nearly half of Prime Day shoppers said they had been waiting for an item to go on sale before purchasing it. A $299 Dream Sock is exactly the type of discretionary baby product parents can delay until a major promotional event.

Owlet also received unusually strong placement during the event. Dream Duo appeared in Amazon’s own official Prime Day deal coverage at up to 33% off, while Forbes highlighted Dream Sock among the best nursery deals at 37% off.

The quarter also had incremental product support outside the U.S. In May, Owlet launched the full Dream ecosystem in Australia and New Zealand with Dream Sight and Dream Duo 3, building on the existing Dream Sock footprint. Owlet OnCall also went live inside the app for select participants during the quarter.

2Q26 consensus calls for a 50.65% gross margin and 7.77% adjusted EBITDA margin. At $31.9M of consensus revenue, this implies ~$16.2M of gross profit and ~$2.5M of adjusted EBITDA.

With this promotional activity, consensus already assumes meaningful gross-margin compression. The 50.65% estimate is 385bps below the 54.5% reported in 1Q26 and below the 51.3% reported in 2Q25.

Owlet360 generated $2.7M of revenue in 1Q26 at a 67.4% gross margin, well above the 54.5% consolidated gross margin. Paying subscribers surpassed 115K, MRR reached $1M, and U.S. Dream Sock subscription penetration reached 34%. Subscription revenue exceeded 10% of total revenue and increased sequentially again, giving Owlet a much larger high-margin recurring revenue base entering 2Q than it had a year ago.

Owlet generated a 54.5% gross margin in 1Q despite a 480bps tariff headwind. Management subsequently lowered its tariff assumption for the rest of 2026 to 15%, versus the prior 19% assumption for Thailand and 20% for Vietnam. Consensus should already capture some of this improvement, but tariff costs below the 15% baseline would create additional upside.

Management also attributed part of the 1Q gross-margin beat to a more favorable mix of Dream Sock versus cameras. Dream Sock was one of Owlet’s most heavily promoted products during Prime Day, so strong Sock volumes could help offset some of the lower promotional ASP. Duo has more camera exposure, but stronger Duo sales also expand the installed base available for Owlet360 conversion and increase LTV beyond the original hardware purchase.

Volume itself can help. Owlet generated a 51.3% gross margin in 2Q25 despite promotions and tariffs, helped by higher revenue, favorable product mix, lower direct product and fulfillment costs, and better fixed-cost absorption. If sell-through stayed anywhere close to the >30% Y/Y growth management disclosed in early May, higher hardware volume should provide another source of gross-margin support.

Management eliminated planned headcount additions, deferred lower-ROI projects outside the core 0 to 24-month market, paused new-country launches carrying regulatory, engineering, quality and marketing costs, and shifted spending toward higher-return projects. These changes could reduce the 2026 cost structure by ~$8M.

2Q is the first full quarter where some of these savings should begin flowing through. If revenue beats the $31.9M consensus estimate, the incremental earnings impact could be meaningful. At a 50.65% gross margin, every additional $1M of revenue produces ~$507K of gross profit before incremental operating costs. With planned hiring reduced and much of G&A and R&D already in place, a revenue beat should carry much stronger flow-through to adjusted EBITDA than the consolidated margin suggests.

Long-term thesis

I think we are likely to see a beat and raise on both revenue and adjusted EBITDA. If I like what I see in the 2Q print, I will expand on the following long-term drivers in a separate post.

• The international opportunity

• Insurance reimbursement

• More on the subscription mix shift and why it will work

• The telehealth opportunity

• Increasing penetration in U.S. and international markets

• The high probability of a buyout

• Dilution and other key risks

• And more…

07/31/26-08/03/26

After a historic early-year rally to an intraday high of $5,595/oz in January, gold has struggled to catch a bid amid geopolitical, inflationary, and demand headwinds. What was once treated as a pure safe-haven asset has recently traded with heightened sensitivity to real yields and the U.S. dollar.

At a spot price of ~$4,000, I think gold presents an attractive long opportunity heading into the remainder of Q3 and Q4. I do not have a hard price target, but I think a rally back to flat YTD, or ~$4,300, is certainly plausible. To supplement my bullish thesis, I selected 3 gold stocks for my watchlist to capture positive beta and idiosyncratic catalysts that are difficult to ignore.

Before getting into why P14 is bullish on gold, let us dissect the basics.

Commodity Overview

Gold is one of humanity’s oldest monetary assets, serving as a store of value and medium of exchange for millennia. Following the end of the Bretton Woods system in 1971, it transitioned into a freely traded macro asset. Over the trailing five years, gold has returned ~123%, compared with ~71% for the S&P 500. This outperformance has continued over the past year, with gold returning ~20% versus 18% for the S&P 500.

Today (07/31), spot gold trades at $4,043/oz, down 6.4% YTD relative to the S&P 500’s 9.6% gain. Its underperformance has been driven by the unwind of a speculative, retail-driven rally fueled by leverage during the first 28 days of the year.

Market participants can access gold through physical bullion, physically backed ETFs such as GLD, IAU, and GLDM, CME/COMEX futures, and gold-mining equities or ETFs such as GDX. Combined OTC and exchange trading volumes average more than $240B per day, providing deep liquidity for institutional allocations.

What Drives Price

Gold prices are primarily driven by real interest rates, the U.S. dollar, inflation expectations, the trajectory of sovereign debt, and net purchases by central banks.

If the world were a simple place, then:

Interest Rates: Gold is a non-yielding asset, so lower real interest rates decrease the opportunity cost of holding it.

U.S. Dollar: Gold is priced in USD globally, so a weaker dollar makes it cheaper for foreign buyers.

Inflation: High inflation erodes the value of fiat currencies, leading participants to buy gold to preserve purchasing power.

Sovereign Debt: A high and rising national debt burden can force governments to monetize debt, elevating credit risk and concerns over currency debasement. Gold benefits as a reserve asset with no counterparty risk.

Central Banks: Geopolitical risks and a confluence of the factors above lead foreign central banks to hold gold as a reserve asset, currency hedge, and source of monetary backing.

However, the world is not a simple place.

Research indicates that real yields have historically been among the strongest predictors of gold prices. Inflation affects interest rates and other variables, including the dollar, creating second-order effects on gold. Historically, real yields and gold have exhibited a near-perfect inverse relationship. More recently however, during parts of 2024 and 2025, gold continued rising despite higher real yields. The relationship appears to have normalized following the unwind of this year’s early speculative rally.

Supply

Unlike industrial metals such as copper, gold supply is less constrained. Global supply averages ~5,000 metric tonnes annually and is relatively inelastic. Mining output typically accounts for 70% to 75% of supply, led by production in China and other major mining jurisdictions. Mine supply cannot respond quickly to price changes because new projects often require 10 to 15 years to reach production. The remainder comes from recycled scrap gold. Remarkably, despite gold reaching an all-time high in January, scrap volumes remained flat.

Demand

Annual demand typically matches supply across four key sectors. Consumer jewelry accounts for ~50% of annual demand, although its share has declined over recent decades. Private investment and central banks account for ~45%, representing an increasing share over time, while the remaining 5% comes from industrial applications. China is the largest buyer of gold.

The Path Forward

Where are we today? Real yields are near GFC-era highs as the market begins to question the Fed’s credibility in controlling the long end of the curve. The dollar has weakened from its YTD highs amid intervention in the JPY, while oil remains in a seesaw as the conflict moves through a perpetual cycle of escalation and de-escalation. If that were not enough, the midterm elections are approaching.

Before getting into my thesis for the remainder of the year, let us first discuss the long-term tailwind for gold. What is the one factor that could prevent gold from crashing? Look no further than national debt. With relentless fiscal expansion and persistent inflation concerns, the debt burden is unlikely to decline anytime soon. This creates a floor for gold as allocators increasingly hedge dollar exposure through the metal.

Now back to my thesis. The Fed’s decision to delay a hike has led participants to question its credibility, sending long-end and real yields toward GFC-era levels. That seems negative for gold, right? Not necessarily. The core rationale for hiking is to combat supply-side inflation caused by the spike in oil. Once, and if, the conflict ends, oil is likely to move lower. Services inflation continues to ease, while economic growth is beginning to decelerate.

My view is that peak growth and peak inflation for the year have probably passed. The market prices a 72% probability of a September hike and is split on another hike in December. I think the Fed hikes once this year. This could cause long-end and real yields to cool as the Fed restores credibility, benefiting gold. In my view, there is no need to raise rates aggressively.

On the other hand, if oil cools sustainably and job growth follows the decline in labor-force participation, I think the Fed would hold rates steady and the long end would adjust on its own.

In terms of real yields, gold looks attractive near its $4,000 support level. Under a September hike, which the market views as highly likely, the long end could begin to cool. Under a no-hike scenario, oil has likely entered a more sustained decline.

After a muted 1Q26, central banks are resuming purchases at a faster rate. The main driver of it all remains China. The PBOC has extended its purchasing streak to 20 consecutive months, with June marking its largest monthly purchase since October 2023. China appears committed to increasing its gold reserves to support the yuan and hedge geopolitical risks.

To determine whether a market trend is shifting, look to institutional managed-money positioning through the CFTC Commitment of Traders data. Managed money entered the year with elevated long exposure and underwent substantial deleveraging during Q1. Long positioning was rebuilt through June and July.

Another tailwind for gold is its seasonal performance during midterm-election years. Since the beginning of the available midterm-election data, gold has outperformed the S&P 500 from August through December in 8 out of 12 cycles.

To summarize my thematic thesis, gold has spent 43 trading days below its 200-day moving average. The last time this occurred was between August and October 2023, when the market was pricing in a higher-for-longer rate regime. Following its reclaim of the 200-day moving average, gold recorded its third-longest streak above the level in 50 years, lasting 660 trading days.

I am by no means suggesting that gold will embark on another historic run during the remainder of the year. However, the setup looks compelling. A Fed hike or a sustained decline in oil prices could weigh on real yields, the dollar has weakened significantly amid JPY intervention, positioning is cleaner and now long-biased, China has reaccelerated purchases, and the ever-growing national debt burden, combined with Treasury demand shifting toward gold, has created a long-term floor.

The risk/reward looks attractive, with my first target at flat YTD, or ~$4,300 spot, and a stop-loss at $3,925.

My preferred way to play this setup is through the GLD ETF (calls and shares). For additional beta exposure, below is my list of 2 miners and 1 royalty co trading at attractive valuations.

Watchlist

RGLD - Royal Gold, Inc.

Price: $201.97 | Mkt Cap: $17.1B | EV/FY26E EBITDA: 11.3x | FY26E EBITDA Margin: 82% | P/NAV: 1.2x | Consensus PT: $301.17

Whenever I am bullish on a commodity, I often gravitate toward royalty companies before miners. The reason is simple: you participate in commodity upside with limited capex and operating-cost risk. As gold prices rise, revenue expands against a relatively fixed cost base, allowing margins and FCF to scale without the same exposure to mine-level cost inflation, project overruns, or dilutive M&A. The key is finding a royalty company with experienced management and a strong acquisition record. RGLD has invested ~$10.7B in royalty and streaming assets, with impairments excluding Pascua-Lama equal to only ~1% of invested capital.

RGLD is the third-largest precious metals royalty and streaming company in the world, but trades at a significant 25% to 30% valuation discount to larger peers WPM and FNV. This gap looks too wide for a business with 78% of 2025 revenue from gold, an 82% adjusted EBITDA margin, interests in 367 properties, and 68% of revenue generated in North America. The Sandstorm/Horizon and Kansanshi transactions diversified the portfolio and should drive more than 30% GEO growth in 2026, followed by ~20% organic growth through 2030. RGLD guides to 430koz to 480koz of GEO production by 2030, with further upside from Fourmile, MARA, Great Bear, Warintza, Platreef, and other development assets (50koz+). Its 1.58 beta to gold and 0.56 beta to the S&P 500 support its role as higher-beta gold exposure with lower broad-market sensitivity.

At $201.97, RGLD trades at 11.3x FY26E EBITDA and 1.2x NAV, below its long-term EBITDA range of 14x to 15x. The balance sheet is also rapidly improving. RGLD repaid $300M of debt in 1Q26, could repay the remaining ~$600M by 4Q26, and has authorized a $500M share repurchase program. The Hod Maden restructuring reduced RGLD’s equity stake from 30% to 15% while increasing its NSR royalty from 2% to 4.5%, reducing future capex exposure and bringing the asset closer to RGLD’s core royalty model. If my gold thesis plays out and spot returns to ~$4,300, RGLD offers upside from higher realized prices, accelerating volumes, balance-sheet deleveraging, and multiple normalization.

BTG - B2Gold Corp.

Price: $3.85 | Mkt Cap: $4.9B | EV/FY26E EBITDA: 2.3x | FY26E EBITDA Margin: 58% | P/FY26 Cash Flow: 4.4x | Consensus PT: $6.31

BTG is probably one of the cheapest gold miners in the market. B2Gold operates four mines across Mali, Canada, the Philippines, and Namibia, with Fekola expected to contribute nearly half of 2026 production. Goose is the main growth asset, Masbate provides steady output, and Otjikoto is transitioning from its depleted open pit to the higher-grade Antelope underground deposit. The 2026 outlook has already been reset lower. Goose guidance came in at 170koz to 230koz versus prior expectations of ~250koz, while the April crushing-circuit fire reduced Q2 guidance from 29koz to 18koz to 20koz. Consolidated guidance of 820koz to 970koz is below the 980koz produced in 2025 due to deferred stripping at Fekola, lower Otjikoto output, and the slower Goose ramp.

The stock has absorbed a barrage of negative news: the guidance reset, repeated Goose crushing issues, the April fire, elevated costs, lower Otjikoto production, and another delay to the Fekola Regional exploitation permit. Mali remains the largest risk. Fekola Main is operating normally under the 2012 Mining Code through 2040, while Fekola Regional falls under the more onerous 2023 code and still requires a separate permit. Missing the June 30 deadline placed 60koz to 80koz of 2026 production at risk. Mali’s tax claims, ownership demands, employee detentions, and temporary control of competing assets justify a discount, although B2Gold has settled its outstanding tax and customs assessments and the current dispute affects incremental Regional ounces. At 2.3x EBITDA, the market appears to price the delay closer to a permanent impairment than a timing issue.

BTG’s $2,400 to $2,580/oz AISC reflects $265M of deferred stripping and underground development, the Goose ramp, and ~$525/oz of royalties and production taxes based on $5,000 gold. At my ~$4,300 target, AISC would decline by ~$84/oz to $2,316 to $2,496, leaving margins of ~$1,800 to $2,000/oz. At the midpoint of guidance, every $100 increase in gold adds ~$79M of annualized pre-tax margin after royalties and taxes. Cash generation should also improve over the next several quarters. BTG produced $362M of FCF in Q1 while still delivering ~66koz per quarter under its prepaid sales obligation, which ended in June and could restore ~$110M of monthly cash receipts during H2 at current gold prices. Goose production is weighted toward Q3 and Q4 as throughput reaches ~3,200 tonnes per day, before moving toward 4,000 tonnes and ~300koz of annual production in 2027. At the current valuation, upside is substantial if Goose stabilizes, prepaid deliveries roll off, and gold returns toward ~$4,300, even without an immediate Fekola Regional permit.

AGI - Alamos Gold Inc.

Price: $28.80 | Mkt Cap: $11.6B | EV/FY26E EBITDA: 6.6x | FY26E EBITDA Margin: 69% | P/FY26 Cash Flow: 8.2x | Consensus PT: $46.25

AGI offers low-cost gold exposure with nearly 90% of NAV supported by long-life Canadian assets. Alamos operates the Island Gold District and Young-Davidson in Ontario and the Mulatos District in Mexico, with Lynn Lake under construction in Manitoba. Island Gold is the core asset and represented more than half of H1 production, while Young-Davidson provides a long reserve life and Mulatos contributes lower-cost production from La Yaqui Grande. The company has 15.9Moz of reserves, a 17-year average mine life, and a funded pipeline that could increase production from 545koz in 2025 to ~1Moz annually after 2030.

The stock has fallen ~26% YTD following another guidance reset. Two seismic events and a storm-related power outage damaged access to higher-grade stopes at Young-Davidson, reducing expected H2 mining rates to ~5,000 tpd versus 7,087 tpd in 2025. Young-Davidson’s 2026 production outlook was cut from ~165koz to ~108koz, driving consolidated guidance down from 570koz to 650koz to 510koz to 560koz. Island Gold guidance was narrowed to 290koz to 310koz, while slower recoveries at La Yaqui Grande lowered the Mulatos outlook. Consolidated AISC guidance increased from $1,500 to $1,600/oz to $1,775 to $1,875/oz, primarily due to lower Young-Davidson production, contractor and labor inflation, and ~$10M of rehabilitation and additional ground-support work. The selloff looks excessive relative to the affected asset and the temporary nature of the disruption. Higher mining rates are expected beyond 2026 as the sequence is optimized and additional ground support is installed.

At the midpoint of revised guidance, my ~$4,300 gold target implies an AISC margin of ~$2,475/oz, or ~$1.32B across 535koz of production before taxes and growth capex. Every $100/oz increase in gold adds ~$54M of annual revenue at that production level before higher royalties and taxes. AGI has also eliminated all remaining 2026 hedges and 85% of the legacy Argonaut forward book, leaving only 50koz at $1,821/oz in H1 2027. Cash generation is already strong despite elevated construction spending, with $144M of FCF in Q2 and $245M in H1. The company ended Q2 with $637M of cash, $200M of debt, and $1.2B of liquidity, with another $310M due from the sale of its Turkish assets through 2027. From 2027, the Island Gold shaft ramp should lift underground throughput to 2,400 tpd, followed by the broader expansion to ~500koz of annual district production from 2028. Combined with PDA and Lynn Lake, Alamos targets ~1Moz of longer-term production and an 18% reduction in AISC to ~$1,250/oz by 2028. At 6.6x FY26E EBITDA, the market is pricing a difficult 2026 while giving limited credit to the coming production, margin, and FCF inflection.

The RealReal, Inc.

NASDAQ: REAL | 07/24/2026

Thesis Summary

Up ~100% over the past year but down ~29% YTD, REAL looks like a turnaround losing momentum. Mr. Market has penalized the stock for guiding to flat Q/Q revenue and lower AEBITDA in 2Q26, while assuming difficult 2H26 comparisons will force growth to decelerate. At 24x NTM EBITDA, the market also views the stock as expensive relative to peers and prices the profitability inflection as largely complete.

I believe the market’s view is shortsighted. REAL is solidifying its leadership in authenticated luxury resale, supported by a growing market, strong repeat engagement, and a differentiated consignment model that deepens supply and buyer liquidity. The shift toward watches, jewelry, and other hard-luxury categories is expanding AOV, while B2B drop-ship, international sourcing, Smart Sales, My Closet, and Athena increase the volume and velocity of supply reaching the platform.

I forecast FY26 revenue growth of 13.3% and GMV growth of 15.0%. From FY26 to FY28, I project a 12.1% revenue CAGR and 13.5% GMV CAGR, above consensus estimates of 10.8% and 12.8%, respectively. Growth is driven by higher-value supply, buyer and order expansion, faster speed-to-site, and broader adoption of luxury resale. These drivers support a longer top-line runway than the market currently expects.

The larger upside comes from the margin profile. Athena reduces labor intensity across intake, authentication, pricing, and listing, while higher AOV increases GMV and profit dollars without a proportional increase in processing costs. I forecast AEBITDA margin expanding from 7.1% on an LTM basis to 13.5% in FY28, still below management’s 15%-20% medium-term target. Positive FCF, debt reduction, an extended maturity profile, and management incentives tied to unlevered FCF strengthen the path toward sustained profitability and deleveraging.

REAL trades at ~2.3x EV/TTM Sales and ~1.9x EV/FY27E Sales, below larger marketplace peers despite stronger growth and a credible margin inflection. Applying a 3.0x multiple to FY27E revenue of ~$879M produces a $19.22 base-case price target, representing 73% upside. My bear case implies a ~$7 share price and 37% downside, creating an attractive risk/reward for investors willing to underwrite the business beyond the next few quarters. Luxury resale is becoming a durable consumer channel, and REAL is positioned to capture a disproportionate share of its growth.

Business Overview

Founded in 2011 by Julie Wainwright, The RealReal (REAL) is the largest online marketplace for authenticated luxury resale. Its end-to-end model removes friction from selling high-value goods by managing sourcing, authentication, pricing, merchandising, customer service, returns, and fulfillment across fashion, jewelry, watches, art, and home decor. Since its 2019 IPO, REAL has shifted away from lower-value, high-volume inventory toward a consignment-led model focused on higher-value luxury goods and deeper technology integration. As of 1Q26, the platform serves more than 40M members and ~1.1M active buyers, primarily in the United States, supported by a selective retail footprint in New York, Los Angeles, Chicago, San Francisco, and other major markets. REAL’s broader mission is to extend the lifecycle of luxury goods and support a more circular fashion economy.

REAL generated $692.8M of revenue in FY25, up 15.4% Y/Y, with a consolidated gross margin of 74.6%. AEBITDA increased to $42.1M from $9.3M in FY24, with margin expanding 450 bps to 6.1%. FY25 was the first year in which REAL generated positive AEBITDA in every quarter. GMV increased 16.4% to $2.13B, while active buyers reached 1.06M at year-end.

Revenue is generated across three verticals. Consignment revenue increased 13.2% to $535.9M, representing 77.3% of FY25 revenue, with an 89.4% gross margin. Under this capital-light model, REAL sells goods on behalf of consignors and retains a percentage of the proceeds without taking legal title to the inventory. Direct revenue increased 41.1% to $91.1M, with a 22.4% gross margin, and consists of merchandise purchased and resold by REAL as principal. Shipping services revenue increased 5.4% to $65.9M, with a 26.0% gross margin, and primarily includes delivery, handling, and consigned-product return fees.

REAL operates an omnichannel marketplace, with its website and mobile app generating most transaction volume and its flagship and neighborhood stores supporting customer acquisition, brand trust, and supply growth. Active members spend more than 40 hours annually on the platform. Physical locations serve as convenient drop-off and valuation points, allow consignors to interact directly with luxury specialists, and generate ~25% of new consignors.

The company manages two inventory models based on title and control. Under consignment, REAL takes physical possession of an item and manages the sale, while legal ownership stays with the consignor until the item is sold and the return period expires. Under the direct model, REAL purchases and takes title to merchandise, primarily through its Get Paid Now program and out-of-policy customer returns.

REAL’s core differentiator is its fully managed, “A to Z” service. Items are collected through in-home appointments, store drop-offs, or direct shipment and transferred to authentication centers in Arizona or New Jersey. Specialists, including gemologists and horologists, authenticate and grade each item using human expertise, proprietary data, and the Athena AI system. REAL then creates the listing, performs the technical copywriting, takes detailed photographs, and sets the price using data collected from more than 50M items sold over 15 years. This hands-off model is designed for consignors who prioritize convenience and do not want to manage photography, pricing, listings, buyer inquiries, returns, or fulfillment themselves.

Processing historically required ~14 days from receipt to listing. Once an item is live, REAL distributes the listing across its digital and physical channels and manages customer service, picking, packing, delivery, and returns. ~50% of items sell within 30 days and 80% sell within 90 days.

Platform economics are driven by Gross Merchandise Value (GMV), calculated as orders multiplied by average order value, and the take rate earned on completed consignment sales. GMV has more than doubled since FY20 to $2.13B in FY25, while the reported consignment take rate has ranged between 35% and 38%, reaching 37.7% in FY25.

REAL’s customer base is concentrated among affluent Millennial and Gen Z consumers, who represent more than 50% of new consignors and form the largest cohort of “RealRealers,” members who both purchase and consign through the platform. ~40% of new consignors report household income above $150K, while more than 40% of new consignors originate from REAL’s existing buyer base. This overlap strengthens marketplace engagement because buyers can monetize prior purchases and use the proceeds to fund additional transactions.

Women’s fashion represented $1.43B, or 67% of FY25 GMV. Jewelry and watches contributed $520M, or 24%, followed by men’s fashion at $170M, or 8%, and home, art, and other categories at $14M. The mix provides exposure to both soft luxury categories such as apparel and handbags and hard luxury categories such as watches and jewelry.

REAL participates in an addressable market anchored by ~$200B of luxury goods held in U.S. closets. This supply is replenished by ~$80B of annual U.S. primary luxury sales, with consumers typically holding items for 3-5 years before recirculating them. ~50% of REAL’s new sellers are also new to luxury resale, indicating that significant supply has yet to enter the secondary market.

The global secondhand apparel market is projected to increase from $257B in 2025 to $393B by 2030. The U.S. resale market is expected to grow from $29.7B to $48.3B over the same period, representing a CAGR of ~10%. Gen Z and Millennials are expected to generate 71% of global secondhand-market growth through 2030.

TAM growth is driven by structural changes in how consumers purchase, own, and monetize fashion. Younger consumers increasingly view secondhand luxury as a regular purchasing channel and are aging into their peak earning and consigning years. Resale value is also influencing primary-market purchases, as consumers evaluate luxury goods based on their potential future recovery value. Demand is further supported by affordability, sustainability, and access to vintage, archival, and discontinued products unavailable through traditional retail. Improved authentication, digital pricing tools, physical drop-off locations, and full-service marketplace models are reducing trust and convenience barriers and bringing additional buyers, consignors, and previously inaccessible closet supply into the market.

Luxury resale is competitive, with platforms differentiated by price point, category focus, authentication standards, and level of service. eBay provides broad horizontal reach through a largely peer-to-peer model, while Poshmark emphasizes social discovery and mass-market fashion. ThredUp operates a managed and highly automated model concentrated in lower-value apparel. Curated platforms such as 1stDibs focus on narrower categories including furniture, fine art, and collectibles. REAL occupies the premium end of the market across both soft and hard luxury categories.

REAL’s competitive position is supported by scale, brand trust, liquidity, and its proprietary data set. The platform’s 15 years of transaction data support pricing and demand forecasting. More than 15,000 items enter the platform daily, creating broad and constantly refreshed selection. Greater supply attracts more buyers, while deeper buyer liquidity improves sell-through and attracts additional consignors.

These network effects are reflected in marketplace retention. More than 80% of GMV comes from repeat buyers, more than 80% of quarterly supply is generated by repeat consignors, and more than 50% of GMV comes from RealRealers who both buy and sell through the platform.

Buyer acquisition cost also declined from $114 in 2019 to $103 in 2025 despite the platform’s larger scale, indicating that marketplace awareness and repeat engagement are reducing the cost of incremental growth.

REAL’s 37.7% reported FY25 take rate is best evaluated against the full range of services included in the model. The company performs authentication, photography, copywriting, pricing, merchandising, insurance, customer service, fulfillment, and returns in-house. Peer-to-peer sellers typically perform these functions themselves or incur separate fees. REAL also absorbs the operational burden created by a 24% to 26% return and cancellation rate. Returned goods stay within its centralized network and can be relisted without requiring additional work or logistics from the consignor. Measured against gross consignment GMV (P14 FY25 estimate: ~$2B), FY25 consignment revenue represented ~26.5%, providing a more comparable view of REAL’s effective revenue yield.

REAL has materially strengthened its balance sheet, generating positive FCF in both FY24 and FY25 while reducing total debt by more than $80M over the same period. As of 1Q26, the company held $123.9M of cash and equivalents, down from $151.2M at FY25 year-end. The company historically relied on convertible debt, including major issuances in 2020 and 2021, to fund growth and preserve liquidity during the pandemic. Balance sheet risk has since declined through debt exchanges completed in 2024 and 2025, which extended maturities and reduced principal outstanding. In FY25, REAL recorded a $40.8M gain on extinguishment after exchanging higher-principal 2028 Notes for lower-principal 2031 Notes. Net leverage has improved to ~5.0x on an LTM basis.

REAL is led by CEO Rati Sahi Levesque, a 15-year company veteran who was appointed to the role in October 2024, and CFO Ajay Gopal, who joined in March 2024. Levesque played a central role in shifting the business toward higher-value supply, stronger unit economics, and sustained profitability. Executive compensation follows a pay-for-performance structure. On a normalized basis, including the annualized value of Levesque’s multi-year equity awards, ~89% of target compensation is at risk. Prior PSU awards included stock-price hurdles of $5.00, $7.50, and $10.00 per share, while current performance awards are tied primarily to cumulative unlevered FCF over multi-year performance periods.

The Top-line Thesis

At ~2.3x EV/TTM Sales, REAL trades at a persistent discount to larger marketplace peers. The valuation implies that Mr. Market views recent growth as part of a finite turnaround and expects the top line to decelerate as comparisons normalize.

P14 expects REAL to sustain revenue growth above the ~10% CAGR projected for the U.S. resale market. Continued execution would support upward estimate revisions and narrow the valuation discount as the market gives greater credit to the duration of REAL’s growth runway.

Product Mix Shift and AOV Expansion

REAL’s shift toward hard luxury allows GMV and revenue to grow faster than order volume. Watches and jewelry carry materially higher selling prices than apparel, increasing the value processed per transaction. The global pre-owned luxury watch market is estimated at $35.7B in 2026 and projected to grow at a 13.3% CAGR through 2033. REAL can capture this opportunity by converting existing fashion and fine-jewelry customers into watch buyers with limited incremental CAC. First-time watch buyers increased 46% in 2025, while 1Q26 AOV rose 15% Y/Y to $646.

REAL’s sourcing network also provides access to original-owner watch inventory that traditional dealers may struggle to reach. In-home appointments and physical stores allow consignors to monetize entire luxury collections in one transaction, including apparel, handbags, jewelry, and watches. REAL’s managed, end-to-end service reduces condition risk, expanding the potential buyer base for transactions ranging from $5,000 to more than $25,000.

Hard luxury also increases GMV throughput without requiring proportional growth in order volume or physical capacity. Illustratively, a $15,000 watch requires a similar storage and fulfillment footprint as a $300 apparel item while generating ~50x more GMV. Watches and jewelry also retain value better than seasonal apparel and can be purchased, worn, and later re-consigned, allowing REAL to monetize the same asset across multiple transactions. The mix shift supports sustained AOV expansion and raises the revenue capacity of REAL’s existing marketplace and logistics network.

Scaled B2B and International Drop-Ship Networks

REAL’s Business Sellers and drop-ship programs expand its supply base beyond individual closets into brands, retailers, boutiques, and professional resellers. These partners can list samples, broken size runs, past-season merchandise, dealer inventory, and other stock that is difficult to liquidate through primary channels. REAL protects the seller’s identity and reports that ~94% of Business Seller items sell within 90 days, providing a discreet channel for converting commercial inventory into cash without severe markdowns. This institutional supply opportunity sits within a global secondhand luxury market that reached ~€50B in 2025 and is incremental to REAL’s consumer closet sourcing.

Drop-ship is a scalable way to grow this business. Select vendors retain custody until an item sells and ship directly to the buyer, allowing each relationship to add commercial-scale supply without routing every item through REAL’s authentication centers. REAL manages risk through partner diligence, provenance requirements, quality checks, and audits. The model expands assortment and GMV capacity without requiring proportional growth in warehouse space or domestic intake volume.

REAL is extending this network internationally through partners in Italy, France, and Japan. European partnerships provide access to boutique and professional reseller inventory, while Japan offers a deep pool of vintage and pre-owned luxury goods. Japan alone holds more than $20B of vintage and archival luxury inventory.

Because REAL does not handle drop-ship inventory, take rates are lower at ~25% vs. ~37% for traditional consignment. B2B drop-ship GMV increased 120% Y/Y to $16M in 1Q26, demonstrating the scalability of the channel. While take rates are lower, near-zero variable intake and storage costs result in a significantly higher contribution margin.

AI’s Impact on the Top Line

Most of REAL’s near-term AI benefit will appear in margins, but the revenue opportunity is also meaningful. Every product is a unique, single-unit SKU, making top-line growth dependent on how quickly new supply reaches the marketplace and how efficiently buyers find relevant inventory.

REAL began rolling out Athena in 2025. Athena is its proprietary AI-enabled intake system, which uses item images and historical data to populate product attributes and descriptions, suggest pricing, assign counterfeit-risk scores, and route lower-risk items through automated processing while directing complex goods to specialists. Athena processed 27% of units in 3Q25 and 35% by FY25 year-end, with management targeting ~50% by the end of 2026. REAL expects Athena to reduce average receipt-to-listing time from 14 days to 7, increasing the volume of supply that can be listed and converted into GMV through the existing authentication-center footprint.

AI also improves price realization and buyer conversion. REAL’s AI-driven pricing engine set launch prices for 85% of items by the end of 2024, while recently added image embeddings allow its models to account for visual characteristics when selecting comparable transactions. This reduces underpricing that sacrifices GMV and overpricing that slows sell-through. On the demand side, REAL is rolling out personalized recommendations and visual, conversational, and agentic search through 2026. These tools are particularly valuable for one-of-one inventory, where buyers may search by image, era, style, color, or other attributes that traditional keyword search handles poorly.

AI is also increasing the productivity of REAL’s supply organization and creating more opportunities to recirculate existing purchases. Smart Sales, launched in 3Q24, uses customer, transaction, and market data to rank consignor leads and identify which clients are most likely to sell high-value inventory. Value generated per sales representative increased 15% in 2024, and REAL added real-time valuation estimates for its sales team in 4Q25. Reconsign launched in 2Q25 as the first component of My Closet, creating a digital catalog of prior purchases and enabling one-click relisting. The broader suite being developed through 2026 adds real-time valuations, earnings estimates, photo-based intake, price tracking, and proactive selling alerts. Smart Sales increases supply acquired per representative, while My Closet turns prior buyers and purchases into recurring sources of GMV.

Direct Revenue and Incremental Monetization

Consignment is REAL’s primary revenue engine, while direct purchases provide an additional top-line lever. Through Get Paid Now, REAL evaluates, authenticates, and prices select watches, handbags, and fine jewelry, then purchases the item before resale. Sellers receive direct-deposit payment within 48 hours or immediate site credit with a 5% premium.

REAL also earns incremental revenue through buyer subscriptions and platform fees. First Look costs $12 per month and provides 24-hour early access to more than 10,000 daily listings and exclusive promotions, while the $49.95 Platinum tier adds upgraded shipping benefits. Subscription revenue was still immaterial in 1Q26, though it offers modest upside as the buyer base grows. Shipping services provide a larger monetization opportunity and benefit directly from higher order volume.

Revenue Build

I forecast FY26 revenue of $785M, up 13.3% Y/Y and modestly above the high end of management’s guidance, on GMV of $2.45B, up 15.0% and in line with guidance. Web traffic and app engagement through 2Q26 point to another strong quarter and management’s flat Q/Q outlook leaves room for another beat and raise.

Consignment drives the build, supported by greater supply of higher-value goods, faster speed-to-site through Athena, and continued buyer and order growth as luxury resale adoption increases. I forecast a decline in the effective consignment revenue yield but expect it to remain significantly above peers due to REAL’s luxury focus and unmatched end-to-end service. Direct revenue provides additional support, but its lower-margin profile will keep it as a secondary growth lever. Over the forecast period, I project a 12.1% total revenue CAGR and a 13.5% GMV CAGR, above consensus estimates of 10.8% and 12.8%, respectively, and ahead of management’s HSD-to-LDD medium-term growth framework.

The Margin Thesis

While top-line growth will keep REAL as an attractive candidate in GARP portfolios, the real juice of the investment thesis comes from the margin expansion opportunity inherent in its platform business model. LTM AEBITDA margin is 7.1%, compared with management’s medium-term target of 15%-20%, leaving substantial room for operating leverage as revenue scales across REAL’s existing infrastructure.

Athena as a Margin Multiplier

REAL carries a heavier physical cost base than most platform businesses. The company operates 4 processing facilities across Arizona and New Jersey totaling ~1M square feet, with $21.5M of facility lease expense in FY25. Its retail network also carries fixed occupancy costs, which totaled $38.7M. These facilities are necessary to support intake, authentication, merchandising, and fulfillment, but create meaningful leverage as throughput increases.

Personnel expense represents the largest variable-cost opportunity. Total people costs reached $308.9M in FY25, up 9% Y/Y and equal to ~45% of revenue. REAL incurs an estimated $78 of operations and technology expense per order, including ~$53 of variable labor tied to authentication, photography, and listing.

Athena reduces labor intensity across intake and authentication. AI-enabled tools have cut processing time for certain high-value items from ~30 minutes to 5–6 minutes and eligible inventory is processed ~20% faster and at a lower cost than under manual workflows. This allows transaction volume to scale faster than personnel expense and is the primary source of operating leverage in the model.

AOV and Unit Economics

The shift toward watches, jewelry, and other higher-priced categories improves profit dollars despite lower percentage take rates. REAL has simplified its accepted categories and stopped taking items expected to sell below $100, removing inventory with weak fulfillment economics.

Processing cost does not increase proportionally with item value. A $5,000 watch can require less labor than a $200 apparel item, particularly when it arrives with supporting documentation, while generating materially more GMV and revenue per unit handled. Higher AOV therefore improves labor productivity, warehouse utilization, and contribution profit per order.

Direct revenue is the main mix risk. Direct AOV is ~10x consignment AOV, but its 21.3% gross margin (1Q26) is materially below the 89.4% consignment gross margin because REAL takes inventory ownership. Management continues to treat direct purchases as a selective sourcing tool, with consignment serving as the core volume and profit engine. A few quarters of faster direct growth can pressure reported gross margin, but do not change the longer-term operating leverage available within the consignment model.

High-Margin Revenue Streams

Retail media, advertising, warranties, and return insurance provide additional margin upside with limited incremental operating expense. These businesses are currently immaterial to total revenue, but leverage REAL’s existing traffic and transaction base and can generate attractive contribution margins as adoption increases.

EBITDA Build

I project gross margin to expand by ~23 bps over the forecast period, driven by modest improvement in consignment gross margin, partially offset by lower direct gross margin.

The larger inflection comes from operating leverage across operations and technology and SG&A, supported by Athena-driven productivity gains. Rather than assuming absolute cost reductions, I conservatively model both expense lines growing more slowly than revenue and declining as a percentage of sales. This drives a GAAP profitability inflection with operating margin expanding by 1,018 bps from -3.5% in FY25 to 6.7% in FY28.

After accounting for D&A and adjustments, primarily SBC, I forecast AEBITDA margin to expand by ~745 bps from 6.1% in FY25 to 13.5% in FY28. My estimates are 8.3% in FY26, compared with the 8.0% midpoint of guidance and 8.3% consensus, 11.4% in FY27 vs. 10.5% consensus, and 13.5% in FY28 vs. 13.3% consensus. The FY28 estimate is still below management’s medium-term target of 15%-20%, leaving additional upside in a bull-case scenario.

Valuation and Base Case

REAL trades at ~2.3x EV/TTM Sales, in line with ThredUp but below Etsy at 3.2x and eBay at 4.5x. ThredUp is the closest resale peer, though its ~$25 AOV and focus on lower-priced apparel limit its exposure to the hard-luxury and AOV tailwinds supporting REAL. No listed peer fully captures REAL’s combination of luxury positioning, consignment economics, network effects, and margin expansion potential.

My base case applies a 3.0x multiple to FY27E revenue, matching REAL’s NTM Sales multiple at FY25 year-end and below the current multiples of Etsy and eBay. This compares with REAL’s current multiple of ~1.9x EV/FY27E Sales. A 3.0x multiple is justified if revenue grows 13.3% in FY26 versus 12.6% consensus and ~12% in FY27 versus 10.4% consensus, while continued execution places AEBITDA margin on a credible path toward management’s 15%-20% medium-term target.

Applying 3.0x to FY27E revenue of ~$879M produces an enterprise value of ~$2.64B. After incorporating a 12% increase in diluted shares from FY25, ~$85M of cumulative debt repayment through FY27, and FY27 cash of ~$176M, I derive a base-case price target of $19.22, representing 73% upside. The expected FCF inflection, combined with management incentives tied to unlevered FCF, provides adequate capacity to address the ~$147M debt maturity in 2029.

Risks and Bear case

Macro sensitivity. Inflation, weaker consumer confidence, or a prolonged economic downturn could pressure discretionary spending and reduce luxury resale demand. Lower buyer activity would weigh on orders, GMV, and revenue, although tighter consumer budgets could also increase consignor supply.

Competition. REAL competes with peer-to-peer marketplaces, managed resale platforms, and traditional luxury retailers. Platforms such as eBay charge lower fees, while larger competitors could invest more aggressively in authentication and resale. Few competitors match REAL’s end-to-end service, but maintaining this differentiation requires continued spending on technology, specialists, logistics, and customer service.

Authentication and reputation. Failure to identify counterfeit or misrepresented goods could damage REAL’s reputation and weaken buyer and consignor trust. Negative publicity related to authentication, shipping delays, returns, or service quality could reduce engagement and increase customer acquisition costs.

Operational infrastructure. REAL depends on a limited number of large authentication and fulfillment facilities, creating concentration risk from fires, natural disasters, labor disruptions, or other interruptions. Cyberattacks, technology outages, and communication failures could also disrupt intake, listings, transactions, and fulfillment.

Supply and labor. GMV growth depends on attracting new consignors, retaining existing sellers, and maintaining a consistent flow of high-quality inventory. REAL also relies on specialized personnel, including gemologists and horologists. Automation should reduce labor intensity across lower-risk inventory, but skilled specialists will continue to be required for high-value and complex items.

In a bear case, revenue growth slows to the mid-single digits as supply, buyer demand, or execution weakens. Applying a 1.5x multiple to FY27E Sales results in a ~$7 share price, representing 37% downside from current levels.

Conclusion

I view the risk/reward as attractive for a long position in REAL. The company is the market leader in authenticated luxury resale, with a longer top-line runway supported by category mix, new supply channels, and continued resale adoption. AI-enabled productivity, higher AOV, and operating leverage support continued AEBITDA margin expansion and stronger FCF, while debt reduction and maturity extensions have improved the balance sheet. With aligned management and the stock trading below larger marketplace peers, successful execution provides a credible path to substantial upside.

CLOSED FOR +0.87% since pitch price (08/27/26). Thesis invalidated with the 2Q27 print + guide.

Build-A-Bear Workshop, Inc.

NYSE: BBW | 07/07/2026

Thesis Summary

Mr. Market prices BBW as a post-COVID mall retail beneficiary whose earnings have peaked, with DTC sales and e-commerce traffic expected to enter terminal decline. Shorts have been rewarded immensely this year with shares down more than 50% YTD. The selloff has followed a barrage of negatives: weaker sentiment toward specialty retail, revenue misses, the retirement of longtime CEO Sharon Price John, 1Q27 marking the first revenue decline in eight quarters and first negative traffic quarter in seven, and a cut to FY27 revenue guidance. FY26 tariff costs and AI-driven search changes added further pressure on margins, online demand, and planned store visits. BBW’s TTM P/E has compressed by more than 50% since the start of the year, its EBITDA multiple is near a two-year low, and short interest is near a record 25% of float.

I think 1H27 marks the operational bottom, with several dynamics supporting a recovery in 2H27. The first catalyst for shares is a low bar into the 2Q27 print. Management’s commentary came only 3.5 weeks into the quarter and extrapolated weak 1Q27 traffic and digital trends through the remainder of Q2, leading consensus to expect another revenue decline, with 2Q27 projected to be worse than 1Q27. Since then, BBW has had a full quarter of its new Hello Kitty stores, the Toy Story 5 launch, strong graduation demand, NBA and World Cup-related sales, and new promotional activity. Alternative data points to a Q1 bottom and improving traffic from late May onward. With short interest near 25% of float, even a modest beat or improved 2H27 commentary could trigger a short squeeze.

Beyond the near-term print, social arbitrage between customer demand and the stock price continues to widen. Wearable plush, bag charms, Sanrio, and Pokémon releases are gaining traction with teens and adult collectors, while BBW posted its best Valentine’s Day in North American history, a solid Easter, and continues to sell more than 20K Birthday Treat Bears per week. The 2H27 calendar also improves with the world’s largest Build-A-Bear store opening at ICON Park, the company’s 30th anniversary, a refreshed Harry Potter collection, Halloween, and Black Friday. The licensed slate stays strong into FY28 with Shrek 5, Spider-Man: Beyond the Spider-Verse, Bluey, and Avengers: Secret Wars. BBW has relationships with the studios behind these franchises, giving the company several opportunities to launch licensed products around major releases. The stock price gives little credit to the customer response or improving product and event pipeline.

Mr. Market is also giving little credit to the mix shift in the business. Commercial revenue has grown at a 54% CAGR since 2020 and increased another 43.6% Y/Y in 1Q27 as BBW expands through partner-operated locations and wholesale distribution. Partner locations generate a 106% ROIC versus 21% for traditional retail, while Commercial carries a 53.1% pre-tax margin versus 9.2% for DTC. International Franchising adds another capital-light growth avenue, with a 26.4% pre-tax margin, 49% ROIC, and room to grow from 112 locations toward management’s long-term target of 300. More than 40% of the global footprint is already outside traditional malls, and the Walmart rollout has shown that BBW products can sell well outside the workshop environment.

The $13.2M tariff refund supports FY27 earnings, while higher average transaction values, lower merchandise costs, Commercial scale, IP licensing, and International Franchising should support margins over time. BBW has zero debt, stable capex, and no appetite for acquisitions, leaving a large portion of FCF available for buybacks. The company has already reduced shares outstanding by more than 25% from its 2019 baseline and has accelerated repurchases during the recent selloff.

Under my base case, revenue grows at a 4.1% CAGR through FY29, pre-tax margin expands 83 bps, and diluted EPS compounds at 10.3%. Applying 9x FY28 EBITDA gives me a price target of ~$65/share, or +117% from current levels, versus ~20% downside to ~$24/share in my bear case. A 2Q27 beat and improved outlook could pull the upside forward considerably given the current short interest and depressed multiple.

Business Overview

Build-A-Bear Workshop, Inc. (BBW) is a global experiential toy retailer built around the in-store creation and customization of stuffed animals. Founded in 1997, the brand has expanded beyond its traditional mall base into tourist destinations, partner-operated venues, international franchises, and digital channels. As of May 2, 2026, BBW had 669 locations across 37 countries, including 376 company-operated stores, 181 partner-operated locations, and 112 international franchise locations. The company hosts 50M+ guests each year, with North America accounting for 92% of revenue.

In FY26 (ended January 31, 2026), BBW generated record revenue of $529.8M, up 6.7% Y/Y and 13.2% from FY23, while pre-tax income reached $67.3M. Pre-tax margin was 12.7%, down 84 bps Y/Y, primarily due to ~$11M of tariff-related costs.

The company operates across three segments. Direct-to-Consumer, its largest segment, generated ~$486M of revenue and $121.6M of contribution margin, with 80% of sales coming from company-operated stores and 20% from e-commerce. Commercial generated $38.7M of revenue, up 23.5% Y/Y, and $9.2M of contribution margin through wholesale product sales and IP licensing. International Franchising generated $5.1M of revenue and $1.7M of contribution margin from royalties, development fees, and product sales to master franchise partners.

BBW’s core model monetizes each stage of the customer experience. Guests select a plush toy for $14-$40, participate in the stuffing and Heart Ceremony, then add clothing, footwear, sounds, scents, and accessories for $7-$25. The model drives high-intent traffic, with 80% of store visits planned in advance. Online customers can replicate parts of the experience through the 3D Bear Builder or purchase pre-built licensed products and collectibles.

The customer base has expanded well beyond families with young children. Teens and adults now account for ~40% of revenue, supported by pop-culture and sports licenses, collectibles, gifting, and the age-gated Bear Cave assortment. The product mix combines licenses from partners such as Disney, Nintendo, and Warner Bros. with owned IP, including Glisten and Merry Mission, which has generated more than $150M since launch.

The company executes its sales model through a four-pillar strategy: Organic Growth, Location Expansion, Wholesale/Outbound Licensing, and Gifting/Personalization. The operation relies on ~5,500 employees across 37 countries.

BBW participates across the $108B global toy market, the multi-billion-dollar personalization and gifting market, and the ~$15B global plush market, which is expected to grow at a MSD CAGR through 2033. The company competes with mass-market toy manufacturers such as Mattel and Hasbro, legacy brands such as Ty, and newer plush competitors including Squishmallows and Jellycat.

BBW differentiates itself through the experience attached to the product. Customers create and customize the product themselves, with the Heart Ceremony adding an emotional component that is difficult to replicate in a big-box aisle or through a standard online transaction. The model has helped BBW build a database of 20M+ first-party contacts while supporting gross margins near 54% and ROE near 35%.

The company has also broadened its footprint through smaller store formats, tourist destinations, shop-in-shops, and partner-operated locations. Since January 2024, BBW has added 135 net locations to reach 669 global sites, giving the company more ways to expand beyond its traditional mall base.

BBW has zero debt and $26M of cash. Since 2021, the company has returned more than $180M to shareholders through buybacks and dividends and has reduced shares outstanding by more than 25% from the 2019 pre-pandemic baseline. The quarterly dividend was increased for the third consecutive year in March 2026 to $0.23 per share, equal to a ~3% yield.

CEO Chris Hurt assumed the role on June 11, 2026, after serving as Chief Operations and Experience Officer since 2015. CFO Voin Todorovic joined BBW in 2014 after previously leading finance for Wolverine Worldwide’s Lifestyle Group. Executive compensation is heavily performance-based, with annual incentives tied to revenue and EBITDA margin and most of the CEO’s long-term compensation tied to multi-year operating targets.

FY27: A Tale of Two Halves

In my view, 1H27 represents the operational bottom. BBW’s YTD weakness has been driven by a sharp slowdown in revenue growth and traffic, with the stock reacting exactly as you would expect in a world dominated by pod bros making L/S decisions based on re-acceleration or deceleration in the P&L. For BBW, the latter was particularly evident in 1Q27.

1Q26 was one of the strongest quarters in the company’s history. Revenue grew 11.9% Y/Y, including an $8.7M increase from existing DTC stores alone, as store traffic significantly outperformed national retail trends. The company leaned into an offensive market-share strategy backed by the success of licensed collections such as Hello Kitty and Pokémon, while the launch of Mini Beans added a new high-volume category at a ~$10 entry price that quickly gained traction with both children and adult collectors.

1Q27 was forced to lap that strength with a weaker product lineup and fewer major launches. Retail traffic fell 7% Y/Y, with adverse winter weather costing an estimated $2M of revenue, while e-commerce demand dropped 26% as AI-driven search changes pressured organic traffic. The digital weakness also affected stores, as 80% of visits are planned in advance and the website often serves as a research tool before customers visit a workshop. Combined with weaker consumer sentiment and lower mall traffic, existing-store sales, typically the largest driver of revenue growth, came under significant pressure.

1H26 was a period of unpressured growth, creating the toughest comparisons of FY27 in the first half. The market is treating the 1Q27 slowdown as the start of prolonged weakness, whereas I think FY27 is more likely to be a tale of two halves as comparisons ease and the product and event calendar improves.

Upside to the 2Q27 Print

BBW’s P/E multiple has compressed by more than 50% since the start of the year, with significant pessimism already embedded ahead of the 2Q27 print on 08/27.

Much of that pessimism stems from management’s tone on the 1Q27 call. BBW does not provide quarterly guidance, but in 4Q26 management initially guided for 4-6% FY27 revenue growth, -3% to +3% pre-tax income growth, at least 20% Commercial growth, and 50+ net new locations. After 1Q27 marked the company’s first revenue decline in eight quarters and first negative traffic quarter in seven, management cut FY27 revenue guidance to 0-3.8% growth.

The bottom-line outlook moved in the opposite direction as pre-tax income guidance was raised to +7.1% to +16.1% Y/Y due to stronger margins and the $13.2M IEEPA tariff refund, with $7M recognized in 1Q27 and another ~$6M expected to flow through inventory later this year. The revenue guide was reset lower after management extrapolated weak 1Q27 trends and the first 3.5 weeks of 2Q27 through the remainder of the quarter, while still expecting a recovery in 2H27.

Management also said 2Q27 revenue would decline Y/Y and likely perform worse than the 2.4% decline in 1Q27. Consensus agrees, expecting revenue to fall 2.6% Y/Y as BBW laps 11.1% growth in 2Q26, while a fourth consecutive revenue miss would surprise few investors.

I think the bar is low enough for an upside surprise. The guidance reset came early in the quarter, after only 3.5 weeks, and the next print will be the first under a new CEO who had every incentive to avoid overpromising in the 1Q27 call. Several product, seasonal, and traffic indicators also suggest the quarter could finish better than management’s early commentary implied.

The first full quarter of the new Hello Kitty and Friends themed stores at Mall of America and American Dream should provide incremental revenue, particularly given how important Hello Kitty was to 1Q26 growth.

Toy Story 5, released on June 19, gives BBW another major licensed product catalyst. The collection spans premium products priced at $60-$70 and higher-volume Mini Beans at $12.50, giving the company exposure across multiple price points.

Graduation merchandise also appears to have performed well in June, with reports of early sellouts in several regions. Sports could provide another smaller source of upside. BBW has a broad NBA licensing relationship, while country-themed bears and apparel have reportedly seen strong demand around the World Cup despite the company not holding rights to official tournament collectibles. Neither the NBA Finals nor the World Cup received meaningful attention on the 1Q27 call. The recently launched Summer of Mystery promotions could also help engagement.

The more speculative evidence comes from alternative data. Placer.ai store traffic and Similarweb web traffic both point to a 1Q bottom and improvement beginning in late May. I would not rely on either dataset in isolation, especially for a small-cap company, but the direction is encouraging and runs against the straight-line weakness implied by management’s early-quarter commentary.

Taken together, the bar for 2Q27 looks low relative to the number of potential upside drivers that developed after the 1Q27 call. With the multiple near a two-year low and short interest near a record ~25% of float, even a modest beat or improved 2H27 commentary could trigger a meaningful squeeze.

Social Arbitrage and an Improving Content Slate

Even if 2Q27 comes in weak, the setup for a stronger 2H27 is intact. The social arbitrage lies in the gap between what is gaining traction with customers and the prolonged weakness being priced into the stock.

In May 2026, BBW launched Slushie Plushies, its first wearable plush, as customers increasingly “style and show off” their products. The same trend has extended into $20 bag charms, which management says are “on fire” across stores and are attracting teens and adult collectors. This is part of a broader shift in the toy market, where adults are becoming a larger source of demand. The U.S. toy market grew 13% Y/Y in early 2026, with consumers aged 18+ driving 35% of the growth.

Seasonal demand also held up better than the 1Q27 headline suggests. BBW had its best Valentine’s Day in North American history and a solid Easter, while the Birthday Treat Bear program continues to sell more than 20K units per week. These results are difficult to square with the market treating BBW as a brand entering prolonged decline.

Collector demand provides another source of repeat traffic. The Pokémon Eevee Evolution series includes eight limited-time releases throughout the year, while the Sanrio Ice Cream Shop collection is performing well with the kidult demographic. Limited releases, owned IP, and licensed collections give BBW more ways to drive repeat purchases beyond the traditional family visit.

While 1H27 laps the toughest comparisons in company history, the 2H27 product and event calendar is much stronger. Late summer brings the opening of the world’s largest Build-A-Bear store at ICON Park Orlando, giving the company a new destination location in one of the country’s largest tourist markets with opportunities across retail, events, and tourism traffic.

October marks BBW’s 30th anniversary, giving the company an obvious nostalgia campaign aimed at adults who grew up with the brand and now bring their own children. A refreshed Harry Potter collection is also timed around the new HBO series, while Halloween and Black Friday add two of the brand’s most important seasonal periods.

The licensed slate stays strong into FY28, with Shrek 5, Spider-Man: Beyond the Spider-Verse, Bluey, and Avengers: Secret Wars all scheduled for 2027. BBW has relationships with the studios behind these franchises, giving the company several opportunities to launch licensed products around major releases.

The Capital-Light Strategy and the Mix Shift

BBW is expanding beyond traditional workshops through partner-operated locations and wholesale distribution in venues such as Walmart, Carnival Cruise Line, Great Wolf Lodge, and SeaWorld. In the partner-operated model, the partner bears the cost of store build-outs, real estate, labor, and inventory, which is purchased from BBW on a wholesale basis. Partner locations generate a 106% ROIC versus 21% for traditional retail, allowing BBW to expand faster with little capital at risk.

Commercial revenue has grown at a 54% CAGR since 2020 and increased another 43.6% Y/Y in 1Q27, making it the strongest part of the business even as DTC weakened. The opportunity has also expanded beyond experience-based locations into traditional wholesale, led by the launch of Mini Beans across 1,500 Walmart stores.

The strategy gives BBW another source of growth while reducing its dependence on mall traffic. More than 40% of the global footprint is now outside traditional malls, and partner-operated locations have more than doubled since mid-2023. Wholesale also increases brand awareness and product trial among customers who may later visit a workshop.

The Walmart rollout could open the door to additional large retail partnerships (TGT, etc.) Even without another Walmart-sized deal, management expects Commercial revenue to grow at least 20% for the fifth consecutive year in FY27 and is investing in the infrastructure needed to support a much larger global distribution network.

The mix shift is also highly accretive. Commercial carries a 53.1% pre-tax margin because partners absorb most operating costs, while BBW earns revenue through wholesale product sales and high-margin IP licensing. Commercial and other third-party channels now account for 32% of LTM pre-tax income, up from a loss in 2019.

BBW is also using more flexible store formats to improve returns on new locations. Traditional Discovery stores generate higher margins but require more upfront capital, while smaller Concourse and shop-in-shop formats produce lower margins with significantly higher cash-on-cash returns. The broader range of formats gives BBW more options to enter high-traffic locations without relying on the economics of a full-size mall store.

The market is giving little credit to the margin benefit from this shift. Even through the FY26 tariff pressure, selective price increases and a growing Commercial mix helped protect profitability relative to pre-tariff levels. As Commercial becomes a larger share of revenue and pre-tax income, BBW should be able to grow earnings faster than revenue. 1Q27 margins should not be used as the baseline, however, because they were inflated by the partial tariff refund.

Going Global

International Franchising is the second leg of BBW’s capital-light strategy. Outside the company-operated markets of the U.S., Canada, U.K., and Ireland, management sees room to grow from 112 franchise locations today to as many as 300 over time. The brand has already expanded from 19 countries to 37 in just two years, with recent entries into Germany, Norway, Italy, and the Philippines.

International Franchising carries a 26.4% pre-tax margin versus 9.2% for DTC and generates a 49% ROIC, giving BBW another way to expand with limited capital. Many of these markets are also much earlier in the brand’s development and have less direct competition in experiential toy retail.

International markets have another advantage in the current environment. Product can generally be sourced directly from China without the same U.S. tariff burden, reducing some of the volatility affecting the domestic business.

BBW is also giving franchisees access to the smaller formats developed for its corporate fleet, including Concourse and shop-in-shop locations. These formats lower the upfront investment required to enter new markets and make it easier to secure high-traffic locations where a full-size Discovery store may not make economic sense.

With the current franchise base still well below management’s long-term target, International Franchising provides a long runway for high-margin, capital-light growth alongside Commercial.

Revenue, Margin, and FCF Build

I project FY27 revenue growth of 2.2%, slightly above management’s 1.9% midpoint, driven by a recovery in 2H27. E-commerce should remain a drag this year, but I expect declines to reverse in FY28 as management adjusts to changes in search and improves its digital strategy, followed by flat sales in FY29. In-store sales should recover in 2H27 and accelerate in FY28 as comparisons ease, the content slate improves, and new high-traffic locations such as ICON Park ramp. Overall, I project Net Retail Sales to grow at a 2.8% CAGR through FY29.

Commercial should continue to lead growth. The model is highly scalable, and the Walmart rollout increases the likelihood of further wholesale wins. I project an 18% Commercial revenue CAGR through FY29, driven primarily by distribution and location growth while holding ARPU flat. International Franchising should grow faster than the overall business but slower than Commercial, where I expect management to focus more heavily given the larger opportunity and stronger economics.

In my base case, total revenue grows 2.2% in FY27, 5.7% in FY28, and 4.5% in FY29, or a 4.1% CAGR. I view MSD long-term growth as reasonable given BBW’s repeat-purchase behavior, expanding distribution, and ability to drive demand through new products and seasonal events.

FY27 margins benefit materially from the $13.2M tariff refund. I estimate Net Retail Sales gross margin expands ~190 bps Y/Y, with most of the increase tied to the refund. Gross margin should normalize in FY28 but stay above FY26 levels as higher average transaction values and lower merchandise costs offset the loss of the one-time benefit.

I expect Commercial gross margin to expand ~200 bps in FY27 and continue improving at a slower pace as IP licensing becomes a larger part of the mix and the business scales across more distribution points with limited incremental cost. International Franchising should also benefit from lower tariff exposure and similar scale economics. Across the business, I forecast 62 bps of gross margin expansion through FY29.

SG&A should stay elevated in FY27 due to 1H27 operating deleverage, continued digital investment, and higher compensation costs. I expect SG&A growth to slow in FY28 and FY29 as revenue reaccelerates, resulting in 83 bps of pre-tax margin expansion over the forecast period.

Higher earnings and stable capital intensity should translate into more FCF available for buybacks. BBW is not an acquisitive business, and I hold capex at 4.7% of revenue. The company repurchased 509K shares in FY26 and another 338K through late May, equal to ~2.6% of the share base in less than four months, with the pace accelerating as the stock sold off.

I forecast diluted share count declines of 4% in FY27, 5% in FY28, and 4% in FY29. The buybacks should be especially accretive in FY28, when pre-tax income laps the tariff refund, helping diluted EPS compound at a 10.3% CAGR through FY29.

Valuation and Risks

Under my base case, I value BBW at 9x FY28 EBITDA, implying a price target of ~$65/share, or +117% from current levels, to be achieved in under two years. The stock could reach that level sooner if a 2Q27 beat and raise forces shorts to cover.

BBW currently trades near its lowest EBITDA multiple in more than two years. Under my base case, I expect the multiple to recover toward, but remain below, the prior peak of ~11x. The name of the game is to buy low and sell high.

E-commerce is the most immediate risk. Management has only recently begun adjusting its digital strategy through a mobile-first approach and changes to SEO. Continued weakness in organic traffic could weigh on both online sales and store visits. I would also like to see BBW become more aggressive on platforms such as TikTok, where toy discovery and social shopping continue to gain share.

Weaker consumer sentiment, lower discretionary spending, and softer mall traffic could delay the 2H27 recovery and pressure the multiple.

Tariffs could return as a headwind. A full reinstatement would pressure merchandise costs and reduce the margin upside in my forecast. The thesis also depends on execution around major seasonal periods and product launches, making a weak Halloween or holiday season another downside risk.

BBW also needs to protect the in-store experience. The workshop model is the company’s core competitive advantage, and pushing too far into standard wholesale at the expense of the experiential business could weaken the brand over time.

Under my bear case, I see ~20% downside to ~$24/share.

Conclusion

Build-A-Bear has established itself as the leader in experiential plush retail, with a durable brand, strong unit economics, and an increasingly capital-light growth model. Bears have extrapolated a difficult 1H27 compare into prolonged weakness while giving little credit to the improving 2H27 setup, Commercial growth, margin mix shift, or continued buybacks.

With BBW trading near its lowest multiple in more than two years, I view the risk/reward as attractive and I am long at ~$31/share.

Equity markets are undergoing a transition as the unchecked capex boom from mega-cap hyperscalers runs into endless price gouging from memory makers. Micron’s earnings once again smashed expectations, but the market reaction was telling. Effectively every stock tied to heavy memory usage has been sold off as investors realize memory is taking a larger and larger share of capex and/or cost of goods sold.

Since hyperscalers make up a large part of SPX and NQ, index pressure makes sense. One would think MU’s earnings would benefit the receivers of capex spend, the toll booths and bottlenecks across AI infrastructure. Instead, the opposite has played out this week.

Markets are waking up to the reality that the biggest stocks in the market, which are also the biggest spenders on AI, have a difficult choice to make. They can continue spending and watch their stock prices wane, which is a problem when stock prices are the greatest incentive for management at these companies, or they can revise capex lower and risk a major vol event that sends the AI trade lower. The goldilocks outcome is that 2Q earnings again show immense returns on investment. Of the hyperscalers, I think Microsoft is the most likely to hint at reducing capex relative to expectations, given that it has been the most negatively affected.

Underneath the hood of a pressured index is an endless rotation into what were once ignored sectors. It almost feels like there is a new rotation every day. This has led to one of the most aggressive quarter-end rebalances I have ever seen.

Does this mean the semis trade is over? No. A large chunk of the selling is likely tied to pension funds rebalancing back toward 60-40 portfolios after a historic run in semis that was due for profit-taking. But that does not change the fact that hyperscalers have to make a choice. I actually think semis and hyperscalers may be due for a bounce within the next few weeks, as stocks sold during quarter-end rebalances often recover in the early weeks of the new month.

Next week also happens to be one of the strongest seasonal weeks for stocks, matching the patriotism around the Fourth of July. I still think we are in the final innings of the AI trade as a whole. There will be pockets of outperformance, but it makes sense to diversify away, at least for the summer.

This is one of the main themes in the market today. While semi-only investors continue hunting for bottlenecks in what is likely the late stage of the AI trade, another theme is starting to come out of the shadows.

Another overhang on equities is the path of rates under a new Fed chair. On 06/17/26, the S&P dropped 1.2% after erasing intraday gains and catching many investors off guard. The Treasury market saw a sharp bear flattener as 2Y yields posted the largest single-day jump on an FOMC day since March 2008.

Since then, the yield curve has retraced all of the move and then some after positive developments around oil’s role in inflation.

What shapes my view that rates are more likely to be held or cut than hiked is as follows.

Oil-driven inflation shock is fading

In the May CPI report, all-items inflation rose 0.5% month over month, while gasoline rose 7.0% and the broader energy index rose 3.9%. With gasoline carrying a 3.937% relative importance and energy carrying 7.474%, the direct contribution from gasoline alone was ~0.28 percentage points, while energy overall contributed ~0.29 percentage points.

PCE prices rose 0.4% in May and core PCE rose 0.3%, but the PCE energy goods-and-services price index jumped 4.0% in the month. Because energy spending represented a little more than 4% of total nominal PCE in May, the direct effect of energy was material. Real energy PCE actually fell 0.7% in May even as nominal energy spending climbed, which means the move was mainly higher prices and not booming real demand.

Following the interim peace framework, oil prices have returned to pre-conflict levels. We should see negative month-on-month inflation prints soon. Peak inflation for 2026 has likely already passed.

I think we are entering the disinflationary growth regime we saw at the start of the year.

Market’s inflation expectations vs. Fed hawkishness

The June FOMC median projects the fed funds rate at 3.8% by year-end 2026, 3.6% in 2027, and 3.4% in 2028, versus the current target range of 3.50%–3.75%. This is a meaningfully hawkish official policy path relative to where the market’s inflation expectations are trading. The Fed has turned hawkish for 2026 and expects a hawkish hold in 2027 before resuming cuts in 2028 and the longer run.

On the back of this, the market also prices one hike in 2026 and a hawkish hold in 2027.

As of June 26, the 10-year breakeven inflation rate was 2.20% and the 5y5y forward inflation expectation rate was 2.19%; even the 30-year monthly breakeven was only 2.30% in May. Put differently, the bond market is not pricing a de-anchoring of long-run inflation expectations. It is pricing inflation risk that is far closer to the Fed’s target than to the inflation prints that followed the oil spike.

The drop in 10-year swaps shows institutional bond investors have high conviction in a return to price stability. This collapse in long-term inflation pricing has effectively offset the rise in real short-term rates, anchoring the long end of the nominal Treasury curve. The market is effectively telling the Federal Reserve that further rate hikes are unnecessary because the underlying inflation impulse has already faded.

In my view, the Fed is afraid to call any kind of inflation “transitory” because of the mistake made in 2021/2022. The difference today is that supply already exists for the commodity responsible for the inflation shock, which in this case is oil.

Kevin Warsh may not be as hawkish as the market assumes

The press has characterized newly confirmed Fed Chair Kevin Warsh as an unyielding hawk whose primary objective is to aggressively raise interest rates. This interpretation relies heavily on his past academic writings and his initial press conference on June 17, 2026, where he said the Fed’s commitment to price stability was “unambiguous and unanimous,” adding that “we have missed on inflation for five years and we are going to fix that.”

Warsh’s primary objective appears to be shifting the Fed toward a minimalist, market-driven framework. He is a long-standing critic of forward guidance, the practice of using explicit verbal commitments to steer future rate expectations. He believes this distorts market signals and forces private participants to trade on Fed commentary rather than organic economic data.

At his debut press conference, Warsh moved quickly to shrink the Fed’s communication footprint. The policy statement was cut to 130 words, removing language that markets could interpret as forward guidance. He also declined to submit an individual dot, saying his colleagues filled out their projections “with pencils... that have big erasers” given the rapidly changing global backdrop.

Warsh also launched five task forces covering Fed communications, the balance sheet, data reliance, productivity and jobs, and the inflation framework. The communications review will assess the SEP, dot plot, and public commentary, while the balance sheet review will revisit the ample reserves regime. The other reviews will focus on real-time/private-sector data, AI’s impact on labor markets, and alternative inflation measures such as trimmed mean inflation.

The inflation and balance sheet reviews are the most important for rates. Trimmed mean inflation gives Warsh a basis to avoid overreacting to energy-driven price spikes, especially with Dallas Fed Trimmed Mean PCE holding near 2.4% during the oil conflict. On the balance sheet, a shift away from ample reserves and toward scarcer reserves would let the Fed tighten liquidity by shrinking asset holdings rather than relying only on a higher fed funds rate. That creates a path for rate stability first, followed by eventual normalization.

So, the central goal of new chair Warsh is to reduce the Fed’s role in pricing the forward curve. My initial view is that this is a strong strategy. The downside is that rates, and by extension equities, will become more sensitive to monthly CPI, PCE, and employment prints. If you thought equities were volatile, rates are about to join that party.

In my view, the July meeting is the only real window for a rate hike this year, as I do not think the Fed will hike into the midterms. With inflation trending lower as the oil shock passes, a hike looks highly unlikely.

For Q3, which is what my screen is focused on, I think the market rotates into rate-sensitive equities. I also think Q3 could look like a toned-down version of the start of the year. Go long what worked in January, metals, EMs (outside the semiconductor-heavy markets like Korea and Taiwan), cyclicals, industrials, and so on.

What would render this screen worthless is a meaningful shift in the rate path caused by a return of the oil shock and/or surprise services inflation. A potential entry point could come from a strong jobs report next week, where seasonal hiring strength pushes yields higher and the market incorrectly interprets it as a sign of hikes, similar to the jobs report that caused carnage only a few weeks ago.

Let’s dive into my ideas. Note that this is not financial advice and is meant for research and entertainment purposes only.

Brazil

I have been bullish on Brazil since the start of the year, at least until the USD began strengthening. See my initial Brazil screen and select equities below. My top single-stock idea in Brazil is B3 SA Brasil Bolsa Balcao, the country’s major stock exchange, which I wrote about briefly in the screen below. I also like going long EWZ for broad exposure. With oil having found a bottom, in my view, downside in Petrobras looks more limited, while Vale gives EWZ direct materials exposure. Both are large holdings in the ETF.

Brazil is a major exporter, so it benefits from a weaker dollar, stronger commodity pricing, and improved EM flows. If U.S. hikes are priced out → USD down → EM currencies up → EWZ works.

Brazil is also on a path to easing one of the world’s highest policy rates. The Selic was recently cut to 14.25% despite an inflation print that came in above estimates. This suggests Brazil’s central bank is willing to look through the current oil-driven inflation shock as temporary.

Another dynamic in Brazil is fiscal stimulus in an election year. In my Brazil screen, I briefly discussed why Lula being re-elected is an over-feared scenario given his track record. That overhang is priced in, in my view. What is not priced is the potential upside if the conservative party wins and the market begins considering a more business-friendly policy path.

Apart from B3, Brazilian fintech should generally benefit from easier monetary policy and resilient consumption. Names like PAGS, STNE, XP, and NU are all trading at attractive valuations relative to their averages in what is likely to be an accelerating growth regime.

Metals: Gold, Silver, Copper

I have been burned on gold multiple times this year and have a better track record in copper. Even so, among the metals, I would rank gold and silver as better longs than copper for this specific setup.

I am a long-term copper bull, but given its correlation to the AI trade and the likely volatility ahead for AI-linked equities, gold and silver look like better near-term expressions of the rates-down, dollar-down trade.

Metals should benefit from a weaker dollar. The main reason the USD broke out to new highs this year was the perceived hawkish pivot from the Fed, ongoing inflation pressure, and seasonally strong jobs reports. The hawkish impulse has outweighed geopolitical de-escalation. At the same time, the Eurozone has faced widening growth differentials versus the U.S. and worse terms of trade, while the BOJ’s move to lift its policy rate to 1.00% in June has not been enough to offset the massive U.S.-Japan rate differential, leaving the yen pinned near multi-decade lows.

I do not think DXY breaks above 102 in the near term. The FX market is already responding to the repricing lower in yields and the divergence between Fed officials and market-based inflation expectations.

My metals ideas are focused on both the commodities and the miners. Gold is bouncing near a key $4,000 level with record-high put skew, while silver has been left for dead. I do not expect a January-style squeeze in either, but I do think a recovery back to flat YTD is plausible under a rates-down and dollar-down scenario.

Lower oil prices also help miners through lower AISC. I particularly like miners with exposure across multiple metals. I have previously been long Ero Copper Corp for its copper and gold exposure, along with its operations in Brazil, which is also an oil-producing country. I have also liked KGHM, the leading silver and copper producer in Europe. Within gold miners, I like Kinross Gold Corp for its capital returns and long-term project pipeline, both of which can support multiple expansion.

U.S. Fintech

Amid this oil-driven inflation spike, consumer spending has stayed resilient. A good way to play this resilience is through consumer-facing fintechs, since they benefit from higher nominal spend while transaction volumes hold up.

In this realm, KLAR, Klarna, is my top idea and a position I am holding for the long term. Since my latest update, management has made progress on the metric that matters most: transaction margin dollars. The company is well on its way to becoming a full-fledged neobank, and U.S. consumer penetration should continue increasing with the already successful Klarna card and new features like high-yield savings accounts.

A new catalyst has also emerged through Klarna’s subsidiary PriceRunner and its massive antitrust damages lawsuit against Google. PriceRunner is seeking 77B SEK, or ~$8B, on claims that Google broke EU competition laws.

I do not believe BNPL is a winner-take-all market, and given TAM growth, I also like AFRM, Affirm, as a Q3 idea.

Shift4, FOUR, is another strong pick worth investigating. While I personally have issues with how they present organic growth given their M&A strategy, the stock has been unfairly hit during the fintech selloff. Shift4 is also a direct beneficiary of the 2026 World Cup through payment processing at many stadiums, while also benefiting from foreign-issued card spending through its Global Blue acquisition. The stock trades at a large discount to the sector and its five-year average.

RKT, Rocket Companies, is more of a mortgage-rate and housing-duration trade than a broad fintech trade. With home affordability coming back into focus and hikes being priced out, RKT is a bet on a recovery in mortgage loan origination volume. With the Redfin deal, RKT is now one of the most integrated housing platforms. The Compass partnership adds another demand channel, with Compass inventory expected to appear on Redfin and the potential for more than 500,000 additional listings to flow onto the platform.

A retail favorite that could work is SOFI. If rates move lower while consumption stays resilient, SOFI should benefit directly through loan originations, funding costs, and stronger demand across its financial services platform. The obvious negative is relentless shareholder dilution, but for a quarter-long trade, shares look attractive here with an average sell-side target of ~$20/sh.

Luxury Marketplaces

Luxury marketplace stocks should be one of the better expressions of a rates-down/spending-up trade. Non-store retailer spending surged 12.2% in May, and there are not many pure plays listed in the U.S. Both names below are also undergoing a margin inflection.

REAL, The RealReal, is a direct luxury resale marketplace. 1Q26 marked its fourth consecutive quarter of double-digit top-line growth. The company is adding new high-volume markets in San Francisco and Boston in 2026 and raised GMV guidance on the back of higher AOV. Higher AOV is helping margins, while AI is improving pricing and marketing efficiency. Average sell-side target is ~$17/sh.

DIBS, 1stDibs.com, is the more rate-sensitive and high-beta version. 1stDibs is a marketplace for luxury design, furniture, art, jewelry, and vintage fashion. It should benefit if lower rates revive high-ticket discretionary spending, housing turnover, interior design demand, and wealth-effect purchases. Recent top-line weakness has forced the company to execute on adjusted EBITDA. A recovery in housing should help management achieve its goal of returning to GMV growth by 4Q26.

Value-focused Consumer Discretionary

This is a play on resilient low-income consumption. My top pick here is JAKK, JAKKS Pacific, a toy company set to inflect on the return of FOB ordering and a materially better content slate in 2027. Other picks are mentioned in the screen linked below. Aside from JAKK, DECK looks attractive heading into earnings at the end of the month. The company continues to take share from Nike in footwear.

Among the sea of green in networking, there lies a stock that has not participated in the rally and now trades with unusually attractive risk/reward, a recovering chart, with the stock up ~15% since first alert, and an undemanding valuation.

Relative value is extremely important for single-stock theses going forward. In an environment where momentum stocks occasionally get violently dumped into the bin, the market quickly searches for value within growing sectors. The stock pitched today has this characteristic of relative value while also inflecting on growth and margins.

P14 sees ~90%+ upside potential in less than 2 years in this NETWORKING STOCK TO RULE THEM ALL.

Disclaimer: Nothing posted by P14 Capital should be considered financial advice. The author of this post holds a long position in BDC. Please consult a financial advisor and conduct your own due diligence before making investment decisions.

Happy Reading!

Belden Inc.

NYSE: BDC | 06/22/2026

Thesis Summary

Belden (BDC) is priced like a cyclical legacy cable company despite its shift into higher-value connectivity, active networking and IT/OT solutions. The business is repositioning across industrial automation, smart buildings and broadband, with a serviceable TAM that is due for rapid expansion. The market is still underwriting BDC as an LSD-MSD top-line grower even though the business is exiting the post-COVID destocking cycle with improving demand across all three core verticals.

The main overhang is the RUCKUS acquisition, which raises leverage near term and looks expensive at ~13x projected 2026 adjusted EBITDA versus BDC’s standalone multiple. In my view, the market is focused on the headline multiple and missing what the deal changes. RUCKUS adds a faster-growing, higher-margin active networking asset with 48K+ customers, 32% Y/Y revenue growth in 2025, 60%+ gross margins and 20%+ adjusted EBITDA margins in the first full year of ownership. It also gives Belden the active wireless and switching layer needed to sell a more complete network architecture from the industrial edge to the enterprise campus. The overhang is the opportunity because RUCKUS changes the growth, margin and strategic profile of the combined business.

The core business is also inflecting before giving credit to RUCKUS or the call options. Automation is shifting from destocking recovery to a longer upgrade cycle driven by reindustrialization, labor scarcity and higher automation intensity. Broadband is past the inventory correction and entering a multi-year upgrade cycle tied to higher speeds, fiber density and BEAD-funded deployments. Smart Buildings is no longer a CRE proxy, with demand shifting toward healthcare, hospitality, education, institutional retrofits and data center gray space. Across the business, Belden is moving from component sales toward solutions wins that combine passive infrastructure, active hardware, software, security and support.

The margin thesis follows suit with the mix shift. Traditional wire, cable and connector sales carry gross margins in the 36%-37% range, while full solutions can command gross margins in the 50% range. RUCKUS should push consolidated gross margins to ~40% post-close and help move solutions mix toward 30% of revenue by FY28. I model consolidated GAAP gross margins expanding 445 bps over the next 3 years and EBITDA margins expanding 473 bps, with most of the step-up arriving in 2027 as RUCKUS is fully integrated.

Consensus looks stale or overly conservative. FY26 consensus implies only LSD BDC-only growth if one quarter of RUCKUS contribution is included, while 2027 and 2028 estimates appear disconnected from the recovery and acquisition contribution. I model revenue inclusive of RUCKUS at $3.16B in FY26E, $4.1B in FY27E and $4.5B in FY28E, implying a 2026-2028 CAGR of +18.7%. Physical AI, edge computing and localized inference are additional upside paths because each requires secure, low-latency local networking infrastructure that Belden already sells.

Even after accounting for the new leverage, P14’s FY26E EV/EBITDA estimate is ~14.5x versus the comp set mean of 21.4x. If the base case is executed, I think BDC can trade at 15x FY27E EBITDA within the next two years as revenue growth, margin expansion, RUCKUS synergies and consensus revisions take center stage. Using management’s 2.9x net leverage target by 2027 and a steady share count, my base case price target is $235/sh, or +93% from current levels. In a bear case, I see ~20% downside to $100/sh.

Business Overview

Belden Inc. (BDC) was founded by Joseph Belden in 1902 and later incorporated in Delaware in 1988. Today, the company serves more than 48,000 customers across industrial automation, smart buildings, broadband and other mission-critical end markets. Automation Solutions is the largest revenue driver at 55% of FY25 revenue, followed by Broadband at 23% and Smart Buildings at 22%. Belden previously reported through Enterprise Solutions, also referred to as Smart Infrastructure Solutions, which included Broadband and Smart Buildings, and Industrial Automation Solutions. Effective January 2026, the company reports as a single segment under a unified operating model.

In simple terms, the company designs and manufactures the IT/OT networking infrastructure needed to move, secure and manage data across physical environments.

Automation Solutions sells rugged networking, connectors, software and cable into discrete automation, process automation, energy, transportation, warehousing, logistics and related industrial environments. Revenue grew +13.6% Y/Y in FY25 to $1.5B, reversing the declines seen in FY23 and FY24. Smart Buildings / Smart Infrastructure sells structured cabling, fiber and copper connectivity, racks, enclosures and active networking into data centers, healthcare, government, hospitality, commercial real estate and education. Revenue grew +7% Y/Y in FY25 to $586M, also reversing the prior two-year decline. Broadband serves broadband and wireless providers through brands such as PPC and Precision Optical, with revenue up +6% Y/Y in FY25 to $633M.

Belden is moving beyond component sales toward a solutions-driven model where customers buy integrated network architectures rather than individual products. This combines passive infrastructure, active hardware, consultative design, software, security and support into larger deployments. Belden Horizon adds the software layer by giving customers one interface for secure remote access, vLOTO-controlled machine approvals, automation tool connectivity, OT DataOps, edge application management, network orchestration and network health monitoring. Software turns Belden’s physical products into a managed IT/OT platform across secure access, machine data, edge applications, orchestration and network health.

Across the three core verticals, FY25 revenue came in at $2.7B, +10.3% Y/Y, reversing the declines seen in FY24 and FY23. Most revenue is recognized when control of goods transfers to the customer, while software and support revenue is recognized ratably over the contract term. The business is still short-cycle, with customers generally not contractually obligated to buy exclusively from Belden or purchase minimum volumes. The short-cycle exposure is part of why the market treats BDC like a cyclical component company, despite the mix shift underway.

Belden’s go-to-market strategy relies on a portfolio of specialized brands, including Belden, Hirschmann, Lumberg Automation, macmon, NetModule, PPC, Precision Optical Technologies, ProSoft Technology, Thinklogical, Tofino Security, CloudRail, Mohawk and West Penn Wire. Each brand addresses a specific part of the connectivity stack, creating a path to cross-sell as customers converge OT and IT networks. BDC uses a hybrid GTM model that combines direct consultative selling with a broad distribution network. Channel partners are the key route to market, but direct engagement is becoming more important as Belden sells more engineered solutions and gets specified earlier in customer designs.

Cable and wire can look commoditized from the outside, but Belden’s products are often engineered into environments where reliability, protocol compatibility, certification and installer familiarity drive purchasing decisions. This is especially true in industrial automation, data centers, broadband networks and regulated infrastructure. The company’s distribution model can be restrictive, but specification-led demand and a broad product portfolio help protect share. The U.S. accounts for 58% of revenue, with the remaining revenue spread globally.

Component manufacturing is lower-margin and exposed to input costs, especially copper, but Belden has historically passed through price increases with limited disruption. Gross margins have held in the 37%-38% range over the last three years, while FY25 operating margin was 11.6%. As the solutions mix rises along with acquisition synergies, management is targeting 25%-30% incremental adjusted EBITDA margins, supported by higher-margin active hardware, software, services and operating leverage.

Belden operates across a fragmented set of adjacent markets spanning industrial networking, enterprise and campus networking, structured cabling and physical network infrastructure, and broadband access connectivity. Its moat is built on switching costs, application engineering, channel/specification power, breadth and scale. In industrial automation, downtime is costly, protocol compatibility matters and rewiring is painful. In smart buildings and structured cabling, specification-in, installer familiarity and certification drive share. In broadband, operators care about reliability, standards compliance and cost.

The Customer Innovation Centers support this strategy by bringing customers into the design process earlier and helping Belden move from discrete product sales to engineered solutions. This supports higher wallet share, better stickiness and a stronger margin profile.

It is difficult to pinpoint a precise TAM given BDC’s broad exposure, but my estimate is that BDC currently serves a ~$40B serviceable TAM growing at an ~8% CAGR through 2035. Including RUCKUS, I believe the TAM expands to $50B-$60B.

Management’s capital allocation framework prioritizes organic reinvestment, strategic bolt-on M&A and shareholder returns. Belden completed 10 acquisitions during 2020-2025, deploying approximately $400M. The strategy culminated in the April 2026 announcement of the $1.85B RUCKUS Networks acquisition, which gives Belden the active wireless and switching layer needed to build a more complete IT/OT networking platform. With RUCKUS, management aims for solutions to reach 30% of total revenue by 2028, above the prior 20% target.

Belden has also returned more than $700M through repurchases since 2019, reducing the share count by more than 11% since 2021. Management historically targeted ~1.5x net leverage and reached that level in 2022 and 2023. The RUCKUS acquisition will raise leverage near term, shifting capital allocation toward deleveraging and temporarily pausing M&A and buybacks until leverage returns to target levels.

Belden is led by President and CEO Ashish Chand, who joined the company in 2002, and CFO Jeremy Parks, who has spent 15+ years in financial leadership roles at Belden. Management is aligned with shareholders through relative TSR, which represents 50% of PSU performance and is measured against the S&P 1500 Industrials Index over three years. Ownership requirements are also meaningful, with the CEO required to hold 6x base salary in Belden stock and other NEOs required to hold 3x. The 2025 PSU grants include a 1.5x share multiplier if adjusted EPS reaches $11.00 by 2028, tying long-term compensation to earnings growth and share price performance.

The core business thesis

BDC’s core business supports a large part of the thesis before assigning any value to RUCKUS and the “call options.”

Automation

In Automation, reindustrialization and reshoring are driving a multi-year need for industrial networking, rugged connectivity and secure plant-level data infrastructure. Manufacturing capacity is moving closer to the point of consumption, particularly in North America and Western Europe, and management has said that while customers may use different “friend-shoring” strategies, the reshoring element is a “100% definite yes.”

The macro data is beginning to support this view. After 3 years of contraction, the recent May ISM PMI reading of 54.0 marked the fifth straight month of expansion.

The important detail here is that Manufacturing Employment is the only major input still in contraction. For Belden, this is a favorable version of manufacturing recovery because production is coming back without the same recovery in labor availability, forcing manufacturers to increase automation intensity rather than rely on headcount.

7M manufacturing jobs have been lost over the past 15 years and are showing no signs of coming back. This means that the current “manufacturing renaissance” is decoupling production from human labor and shifting to automation.

Factory starts alone are a poor proxy for Automation revenue. Belden participates later in the capex cycle, when projects move from site preparation and structural framing into mechanical, electrical, plumbing, controls, networking and commissioning. FRED’s TLMFGCONS data shows manufacturing construction spending peaked at ~$240B in 2024 and declined to $186B in the latest April 2026 reading. The initial surge was driven by CHIPS Act and IRA-related factory buildouts; the next phase is the plant infrastructure layer needed to make those facilities operational.

This multi-year pipeline of factory footprints requires industrial Ethernet, secure network architecture, ruggedized connectivity and IT/OT infrastructure. Management has pointed to this trend across customer verticals since early 2025, with pipeline strength in pharmaceuticals, consumer goods, logistics and automotive.

Labor scarcity is the second leg of the Automation thesis. Reshored facilities need to run with fewer skilled workers, which turns automation and digitized productivity into a requirement. Management has framed this directly, noting that inflation is pushing customers toward automation because the U.S. does not have enough workers to manufacture all the products being reshored. The answer is higher output per worker through automation.

Compounding this issue is the macroeconomic reality of an elevated cost of capital. Industrial players face a complex trilemma: they must expand domestic capacity, navigate a labor shortage, and optimize returns on expensive borrowed capital. The only viable solution to defend corporate margins is to sweat existing assets harder by amplifying output per square foot through extreme automation.

Automation organic revenue grew MSD in 1Q26, while reported revenue grew +10.6% Y/Y. FY26 growth may come in below FY25 because last year benefited from the destocking recovery, but the vertical is shifting from inventory normalization to a longer automation and network upgrade cycle.

Broadband

Broadband has been one of the main reasons investors view Belden as a short-cycle cyclical. This perception was reinforced during the 2023-2025 downturn, when the business was hit by customer inventory correction and pauses in network architecture upgrades.

The downturn had two main causes. First, broadband customers over-ordered after COVID to protect against supply chain issues, creating excess inventory across the channel. In 3Q23, Belden saw a sudden procurement shift as major MSOs moved from restocking to inventory reduction and shorter lead times, driving a 28% sequential drop in orders. Second, deployments slowed in the back half of 2025 as MSOs dealt with technical interoperability issues across complex DOCSIS upgrade stacks, creating stops and starts in project execution.

By late 2025, customer inventory levels had moved back toward longer-term norms, and management confirmed in early 2026 that the correction phase was over. 1Q26 Broadband revenue grew +5.9% Y/Y despite lapping a +31% comp in 1Q25 and despite 1Q being seasonally slower. Management also expects improvement through the rest of 2026.

Belden is now entering a structural growth phase driven by a multi-year upgrade cycle for higher speeds and increased fiber density. Broadband has shifted away from bulk fiber and commodity cable volume toward higher-value access fiber and last-mile network upgrades. Fiber has grown from ~5% of Broadband revenue five years ago to ~50% today, with Belden focused on higher-finesse access fiber rather than lower-value trunking.

Aside from the cyclical recovery and shift in fiber, a significant catalyst for the Broadband business is the BEAD catalyst. The $42.45B Broadband Equity, Access, and Deployment program is the largest broadband infrastructure stimulus in U.S. history. With state allocations approved and field construction beginning, capital deployment should scale through the end of the decade. Belden is positioned to capture this spend because the leading BEAD recipients overlap with its largest long-standing MSO accounts. After the inventory correction, BEAD creates a multi-year demand layer on top of the fiber and DOCSIS upgrade cycle.

Smart Buildings

Do not let the networking label fool you; Belden is not a pure data center networking play. Data centers make up less than 20% of Smart Buildings revenue. Instead, this segment is a play on institutional property upgrades and data center “gray space”.

The segment was historically viewed as a proxy for non-residential construction and corporate office, but CRE now accounts for less than 10% of vertical revenue. Capital and engineering resources have shifted toward healthcare networks, hospitality campuses, education, high-density venues, data centers and localized enterprise edge installations.

Demand is increasingly tied to brownfield retrofits, regulatory pressure and operating efficiency. Hospitals, hotels, campuses and venues are upgrading existing infrastructure to automate workflows, improve security, manage energy usage and consolidate fragmented systems onto one backbone. This type of capex is less dependent on new physical square footage and more tied to the need to modernize existing assets.

Belden’s data center exposure is also more attractive than the headline mix suggests. The company focuses on gray space infrastructure, including facility power management, substation automation, building controls, security, specialized climate systems and liquid cooling networks needed to keep high-density AI clusters online. It avoids the most commoditized white space applications, including standard compute racks and high-volume cabling where pricing pressure is much higher.

A primary proof point is the 2Q25 multi-site solutions award with a tier-1 AI hyperscaler, where Belden is installing ruggedized industrial networking products across advanced cooling systems. Power and cooling availability are becoming the major bottlenecks for AI infrastructure, and those bottlenecks require the industrial-grade networking and control infrastructure Belden already sells.

Reported data center exposure likely understates the true revenue base because a portion of racks, wire and cable is sold through distributors that ultimately serve data center customers. Management has expanded the dedicated data center team by 2-3x over the last few quarters and disclosed that the active data center solutions pipeline is roughly 2x to 4x the current realized revenue base.

Solutions mix in the vertical was effectively 0% at the start of 2025 and reached a MSD % by 1Q26. As Belden moves toward its company-wide solutions target, Smart Buildings should see one of the stronger margin ramps because the offering increasingly combines active hardware, passive infrastructure, software and services.

The inflection already started to show in 1Q26, with Smart Buildings revenue up +21% Y/Y, aided by an easier comp and double-digit growth in data center-related revenue. FY26 growth should outpace the +7.4% seen in FY25 if retrofit demand stays healthy and data center gray space continues to scale.

The Solutions Transformation and a Unified Model

Industry experts have characterized Belden’s distribution as “very restrictive,” particularly in wire and cable, noting that Belden often limits direct access to a small number of approved partners to protect its highly specified “Belden Classics” products. This model has historically led to crossovers where unapproved distributors switch a Belden specification to a competitor’s product because they cannot procure Belden directly. This strategy is what has been putting a lid on top-line growth.

The unified operating model is designed to fix this constraint. Management has acknowledged that the prior segment structure created internal walls and limited cross-selling. The new model creates a single customer-facing structure and allows Belden to sell enterprise, broadband and industrial products through key distribution partners with a more complete view of the customer’s network needs.

CEO Ashish Chand noted that reaching a 15% solutions mix by 2025 involved “a little bit of brute force” because the company was not organized to service customers on a unified basis. The realignment is designed to move Belden from bespoke, one-off designs to scalable reference architectures that can be easily adopted across the entire channel. Management is focused on a balance of simplifying complex solutions so customers can adopt them more easily, which they believe will cause “adoption rates to go up dramatically”.

This move plays directly into ramping the solutions business, which in effect realizes higher revenue due to the bundling of products and services, as well as higher margins with a better mix of products including software and the ability to skip the distributor by selling to the end-consumer.

Belden has deployed over 250 solutions consultants who work directly with end-users to specify Belden products into their designs. Once specified at the design level, the high-margin active products (switches, software) automatically pull through high volumes of traditional passive products like cabling, regardless of which distributor fulfills the order.

IT/OT Convergence

IT/OT convergence is one of Belden’s most important growth drivers. Historically, enterprise IT networks and plant-floor OT networks operated separately. IT handled enterprise data, email, ERP systems and cloud applications, while OT controlled machines, sensors, assembly lines and physical processes through fragmented legacy protocols such as Modbus and Profibus.

Factories, warehouses, hospitals, utilities and transit systems now need real-time data moving securely between physical assets, edge devices, control systems and enterprise software. Without a converged IT/OT network, Physical AI, autonomy, predictive maintenance and edge computing cannot function at scale. A robot, grid asset or automated warehouse system cannot self-adjust in real time if the data is trapped inside isolated systems or routed through slow, fragmented networks.

This expands Belden’s wallet share. In a siloed network, Belden may only sell the passive layer, including cable, connectors and physical infrastructure. In a converged architecture, Belden can sell the full stack, including ruggedized switches, edge devices, security, network management software and data orchestration. A project that may have been limited to raw cable volume can become a larger engineered solution across the full data journey, from sensor capture to edge processing to secure cloud connectivity.

RUCKUS causing a Ruckus, the good kind

The RUCKUS acquisition is being priced as an overhang even though it accelerates Belden’s solutions transition. Leverage moves higher, but Belden also gains a faster-growing, higher-margin active networking asset with 48K+ customers, 32% Y/Y revenue growth in 2025, 60%+ gross margins and projected adjusted EBITDA margins above 20% in the first full year of ownership. At ~13x projected 2026 adjusted EBITDA, RUCKUS lifts the pro forma solutions mix from 15% to over 20%, pushes consolidated gross margins above 40% in 2026 and takes the combined EBITDA base to ~$650M with $360M+ of unlevered FCF.

Belden has recently secured $1.85B of debt financing from JPM at SOFR + 225 bps. The market is focused on the headline multiple and leverage step-up, while underweighting the asset quality and cross-sell opportunity inside Belden’s installed base.

Part of the skepticism comes from RUCKUS changing hands five times since 2016, from Brocade to Arris to CommScope to Vistance to Belden. The history can look like product weakness, while the better explanation is market timing, parent-company balance sheet pressure and a product portfolio better suited as part of a broader networking platform. RUCKUS was repeatedly caught inside parent companies with different strategic priorities.

The strategic limitation of RUCKUS as an independent asset is specialization. Its ICX switching line adds campus switching, but the portfolio lacks the broader enterprise stack of Cisco or HPE, including campus core switching, data center routing, SD-WAN, AI-driven security and unified licensing. Large enterprise buyers often prefer a single vendor across the network stack, which has left RUCKUS treated as a high-quality Wi-Fi point solution.

This is EXACTLY what a combination with Belden solves.

Belden gives RUCKUS the horizontal context it lacked. Belden already owns passive infrastructure, industrial connectivity, OT security, customer specifications and distribution reach. Adding RUCKUS gives Belden the active wireless and switching layer needed to offer a more complete active and passive network architecture from the industrial edge to the enterprise campus.

RUCKUS was first-to-market with enterprise Wi-Fi 7 through the R770 access point, placing the business at the front end of a multi-year upgrade cycle as enterprise networks move toward higher density, lower latency and more deterministic wireless. Peer commentary across networking already points to rapid Wi-Fi 7 adoption, with HPE citing +600% growth in Wi-Fi 7 sales from network modernization. RUCKUS also adds RUCKUS AI, an AI-driven cloud networking platform for assurance and optimization, along with a more advanced NaaS offering than Belden had on its own.

RUCKUS is primarily enterprise today, with hospitality and education as its two largest verticals. It also has exposure in automated warehouses and material handling, already overlapping with Belden’s industrial base. The overlap creates the first layer of synergy. Belden can combine its passive infrastructure with RUCKUS’ active wireless and switching portfolio on the enterprise side, while taking RUCKUS deeper into legacy industrial markets where wireless adoption is still early.

Discrete manufacturing offers the largest upside because most machine data still moves over wireline networks, while management expects wireless to move toward a 50/50 mix over the next 3-5 years and potentially become the majority of machine data traffic over time. Customers are already planning for the shift, but many lack one vendor capable of providing the full blueprint across passive infrastructure, ruggedized networking, enterprise wireless, cybersecurity and software management. With RUCKUS, Belden can become the blueprint vendor.

Across Automation, Smart Buildings and Broadband, RUCKUS adds capabilities directly tied to Belden’s existing vertical strategy. In Automation, it extends high-performance wireless into industrial environments needing low-latency connectivity for Physical AI, robotics, autonomy and machine-to-cloud data flows. In Smart Buildings, it completes the IT layer for healthcare, hospitality, education and large campus environments, moving Belden beyond fixed infrastructure into mobile, reliable hybrid networks. In Broadband, it brings carrier-grade wireless, 4G/5G and service-provider exposure, strengthening Belden’s ability to deliver integrated networks across campuses, venues and smart cities.

By 2028, management expects RUCKUS to help push solutions mix to 30% of total revenue. The importance of the acquisition is not only the revenue contribution, but the change in what Belden can sell. Belden moves from a strong passive and OT-facing supplier into a more complete active and passive networking provider with higher wallet share, stronger margins and a better position in the IT/OT convergence cycle.

Wireless networking upgrade cycles generally span 5-6 years. The shift from Wi-Fi 5 to Wi-Fi 6 lasted ~5-6 years from initial ratification to enterprise saturation. Wi-Fi 7, formally certified by the Wi-Fi Alliance in January 2024, should drive a 6-7 year enterprise upgrade cycle peaking around 2030-2031.

IDC data shows Wi-Fi 7 moving from less than 1% of enterprise dependent AP revenue in 1Q24 to 44.5% by 1Q26, making it one of the fastest-ramping enterprise wireless standards in its tracking. Adoption is being pulled forward by a maturing Wi-Fi 6/6E installed base, AI-powered enterprise applications with higher throughput and latency requirements, and broader availability of certified access points across price tiers. Wi-Fi 7’s Multi-Link Operation and 320 MHz channel support address these requirements directly, which is pulling forward buying decisions previously expected in 2027 or later.

Management expects HSD growth in 2026 from RUCKUS’ $687M 2025 revenue base, but the outlook appears conservative. Vistance WLAN revenue grew +12.9% Y/Y in 1Q26 before the asset was sold, while industry research points to 60%+ Y/Y Wi-Fi 7 market growth in 2026 and a 25%+ CAGR over the next 3-5 years. RUCKUS has MSD share behind Cisco and HPE, so share stability or share capture can drive growth above expectations.

Call Options

Physical AI

As AI moves into robotics and automation, Physical AI gives Belden another upside path beyond the core automation cycle. The global Physical AI market is estimated at $5.4B in 2025 and is expected to reach ~$61.2B by 2034, implying a 31.3% CAGR. Belden does not need to pick the winning robot, sensor stack or AI model. Its role is to provide the local network infrastructure required for these systems to operate safely and reliably.

Physical AI shifts machinery from fixed automation to adaptive real-world decision-making. Humanoid robots, computer-vision picking systems and autonomous mobile robots require uninterrupted telemetry from cameras, LiDAR, sensors and force-feedback loops. These assets make constant mechanical adjustments based on real-time inference, so packet loss, jitter or latency can cause immediate physical failure, including machine collisions, missed picks or dropped freight.

Industrial facilities deploying these systems need more than standard connectivity. They need active infrastructure capable of Time-Sensitive Networking, deterministic data movement and plant-wide clock synchronization. This pushes customers from basic cable and connector purchases toward full IT/OT solutions, including OpEdge hardware, industrial firewalls, ruggedized switches and Belden Horizon for data orchestration. The same factory footprint can shift from passive products carrying ~37% gross margins to solutions wins with gross margins above 50%.

Physical AI therefore expands Belden’s wallet share while limiting platform-specific risk. Whether the winning system comes from a robot OEM, warehouse automation vendor or industrial software company, the facility still needs a secure, low-latency network layer underneath it. Belden gets paid on the infrastructure required to make the physical AI deployment work.

Edge Computing and Localized Inference

Edge computing moves processing and storage closer to the data source, while localized inference runs AI models directly on local devices. The need is simple: industrial customers cannot send every machine-data stream to the cloud, wait for processing, then route the answer back to the plant floor. Management has already called this “super expensive,” and safety-critical environments require microsecond decision-making with local control.

Token costs, bandwidth limits and open-source model restrictions should also push more inference toward the device or local edge server. Research estimates the global edge computing market will reach $140B by 2030, implying a 32% CAGR, while Edge AI is projected to grow at a 21% CAGR as connected devices scale toward a projected 1T IoT connections by 2030.

Belden is positioned to benefit through both hardware and software. On the hardware side, the company provides ruggedized switches, gateways and time-sensitive networking infrastructure required to move machine data locally with low latency. On the software side, Horizon acts as the orchestration layer for edge devices, allowing customers to run containerized applications, detect anomalies locally and manage distributed edge networks without routing every decision through the cloud.

The edge opportunity reinforces the same core shift across Belden’s business. More data created at the machine level means more need for secure connectivity, local processing, network orchestration and software-enabled infrastructure.

Revenue Build

Management’s long-term target is MSD annual revenue growth, or 4%-6%. In my view, this is conservative and does not reflect the revenue opportunity over the next 2-3 years. There is far more upside than downside given the drivers across the core business, RUCKUS and the call options.

Consensus is even more conservative, projecting +4.6% revenue growth in 2027 and +3.8% in 2028. Either estimates have not been updated for RUCKUS, or the sell-side is waiting for the deal to close before fully reflecting the pro forma revenue base.

The RUCKUS acquisition is expected to close in 2H26. Financing has been secured, leaving regulatory approvals as the primary remaining step, which I expect to clear given the size and structure of the industry. Assuming a 4Q26 close, BDC would receive one quarter of RUCKUS contribution. Management expects HSD growth from RUCKUS’ $687M 2025 revenue base. Using +8% Y/Y growth implies ~$742M of annualized revenue, or ~$185M per quarter. I use $180M to account for seasonality, with 2Q and 3Q typically stronger.

From what I see, FY26 consensus may not be fully updated for RUCKUS, or it may simply be overly conservative on the core. FY26 consensus revenue is $2.97B, +9.4% Y/Y. If consensus includes one quarter of RUCKUS, or ~$180M, implied BDC-only growth is just +2.7% Y/Y. Unless the sell-side is seeing something I am not, this looks too low for a business coming out of destocking with improving demand across all three verticals.

Under the assumption consensus is still BDC-only, the 2027 and 2028 numbers look even more disconnected. Consensus projects +4.8% Y/Y growth in 2027 and +3.6% Y/Y in 2028. In my view, this is too low for a business in cyclical recovery with ramping exposure to automation, data center gray space, broadband upgrades and RUCKUS cross-sell. Consensus appears anchored to the post-COVID supply glut and destocking period, while also leaning heavily on management’s conservative long-term algorithm. As shown below, management has beaten top-line consensus estimates in each of the past 16 quarters.

The next 2-3 years should look very different from 2023-2025. Industrial customers need to automate in response to reindustrialization, inflation and labor scarcity. Smart Buildings customers need to retrofit facilities for IT/OT convergence, energy efficiency, security and data center gray space. Broadband is entering a multi-year upgrade cycle with BEAD as a funding catalyst. RUCKUS adds an obvious cross-sell opportunity across Belden’s enterprise and industrial customer base, while Physical AI, edge computing and localized inference provide upside beyond the base case.

My assumption is RUCKUS will be broken out as a separate business line under wireless networking. For the BDC core business, I model revenue growth of +10.4% Y/Y in 2026, +10.2% Y/Y in 2027 and +9.8% Y/Y in 2028, implying a 9.92% CAGR over the 2026-2028 forecast period.

Across the three core verticals, I project a 3-year CAGR of 9.7% in Automation, with growth ramping in 2027 and 2028. In Broadband, I project a 3-year CAGR of 6.2%, with growth highest in 2027 as BEAD begins to scale. In Smart Buildings, I project a 3-year CAGR of 14.3%, driven by retrofits, data center gray space and solutions mix.

This marks a sharp contrast to the cyclical irregularity seen in 2023-2025, when average growth was 4.7%. The market is still underwriting the last cycle, but supply is no longer overbuilt and demand is improving across all three verticals.

For RUCKUS, I assume management’s conservative HSD 2026 growth expectation holds. In 2027, I project growth of 11%+ Y/Y as cross-selling across BDC’s industrial and enterprise base starts to take effect. In 2028, I project +13% Y/Y, reflecting further cross-sell ramp and the solutions transformation moving closer to management’s 30% mix target.

All in, revenue inclusive of RUCKUS comes out to $3.16B in FY26E, +16.3% Y/Y, $4.1B in FY27E, +30% Y/Y, and $4.5B in FY28E, +10.5% Y/Y. This implies a 2026-2028 CAGR of +18.7%.

A smart reader like yourself could point to pass-throughs and FX, but these have contributed at most ~2% over the past four years, including 2025. The disconnect is not copper, FX or one-off pricing. The issue is a likely misunderstanding of where the underlying verticals are in their cycles.

Margin Thesis and EBITDA Build

BDC’s margin thesis rests on mix, pricing, operating leverage and RUCKUS. IT/OT convergence is accelerating the solutions transformation because customers are buying integrated architectures across passive hardware, active hardware, software and services, rather than standalone cable, connectors or basic components.

Traditional product sales, including wire, cable and basic connectors, typically generate gross margins in the 36%-37% range. These products are often sold through distributors and contractors and carry more unit-price competition. Full solution sales, which combine passive products with active components like switches, routers and gateways, plus software such as Belden Horizon, typically command gross margins in the 50% range.

The shift also changes the pricing discussion. BDC engages with customers through a consultative process tied to operating KPIs, such as reducing plant scrap or improving hospital safety. This gives Belden more room for value-based pricing instead of competing through RFPs based mainly on unit price.

Expert calls have validated Belden product quality as best-in-class, reinforcing a TCO-over-price purchasing framework. Customers increasingly evaluate Belden’s solutions based on productivity gains and lifecycle savings, reducing price pressure even in inflationary environments.

Management targets 25%-30% incremental adjusted EBITDA margins, which translates to 50-75 bps of annual gross margin improvement plus operating leverage across SG&A and R&D.

RUCKUS adds a higher-margin asset to an already improving core. RUCKUS generates gross margins above 60% and should push consolidated gross margins to ~40% immediately post-close. RUCKUS also generates adjusted EBITDA margins above 20%, compared with Belden’s 16.9% adjusted EBITDA margin in FY25. The combination accelerates the solutions mix toward 30% of revenue by FY28. Belden also gains scale synergies from selling RUCKUS into its industrial customer base, while durable Wi-Fi 7 demand should limit pricing pressure for at least the next few years.

In my view, consolidated GAAP gross margins expand 445 bps over the next 3 years, with most of the improvement driven by RUCKUS integration. I assume RUCKUS gross margins hold steady at 61%, in line with management’s comment, “RUCKUS consistently delivers gross margins above 60%.” For the core business, I model 67 bps of expansion in Smart Infrastructure, including Broadband and Smart Buildings, and 25 bps in Automation. Most consolidated gross margin expansion arrives in 2027, the first full year of RUCKUS ownership.

EBITDA margins should inflect under the unified model and RUCKUS combination. I model 473 bps of EBITDA margin expansion over the next 3 years (458 bps of adjusted EBITDA margin expansion), with most of the improvement in 2027. This aligns with management’s expectation for 20%+ RUCKUS adjusted EBITDA margins in the first full year of ownership and its 25%-30% incremental adjusted EBITDA margin target.

Valuation and Risks

Due to the overhangs mentioned above and Belden’s “industrial” label, the stock has consistently traded at a discount to networking peers. This has also reinforced consensus conservatism. As they say, price drives sentiment.

The acquisition overhang is turning into a catalyst, while Belden’s industrial label, which capped the multiple during the post-COVID destocking cycle, should begin to help as reindustrialization and upgrade cycles accelerate. An inflecting core business, revenue and margin upside, RUCKUS synergies, the industrial + networking + AI thematic, likely beats versus stale consensus estimates, and the call options from Physical AI, localized inference and edge computing create a setup for multiple expansion.

BDC currently trades at a steep 10.3x EBITDA discount versus the mean of the comp set on consensus forward estimates. Even after accounting for the new leverage, assuming the full $1.85B was borrowed as of 1Q26, P14’s FY26E EV/EBITDA estimate is ~14.5x, still a considerable discount versus the comp set mean of 21.4x.

If the P14 base case is executed, I think BDC trades at 15x 2027 EBITDA at some point within the next two years. The likely catalysts are earnings upside versus consensus, a reflexive move in price driving sentiment, and a broader rotation toward industrial automation and Physical AI. Once the market realizes 2026-2028 does not look like the post-COVID 2023-2025 period, the multiple should re-rate closer to peers.

With the new debt raise, principal debt would be ~$3.1B assuming no repayments. Using management’s 2.9x net leverage target by 2027, net debt would decline to ~$2.6B. Assuming a steady share count, the P14 base case PT on 15x FY27E EBITDA is $235/sh, or +93% from current levels, to be achieved in under two years.

Risks

Unlike high-multiple networking stocks caught in the momentum trade, Belden offers relative value in a market where momentum can unwind quickly. The multiple has more room to expand in response to earnings upside and catalysts, while the lower starting valuation should provide better downside protection in a sector-wide selloff. In a bear case, I see ~20% downside to $100/sh.

Quick rise in copper prices - A parabolic rise in copper would create an initial margin headwind before pricing actions flow through. Belden can pass through price increases with relative ease, but a rapid move in copper typically pressures margins near term. I do not expect a parabolic move given the strengthening dollar.

Macro/rates - BDC’s customers are capex intensive and sensitive to rates. A rate hike would pressure customer spending and increase BDC’s interest expense given the SOFR-linked debt financing. This would also weigh on the stock because leverage is one of the current overhangs.

Integration risk - Management has a strong capital allocation track record, and RUCKUS has clear synergies with BDC’s core business. Still, integration risk needs to be monitored because failure to execute would weaken one of the main drivers of the margin thesis.

No repurchases - Share repurchases will be temporarily paused until deleveraging targets are met. This removes an incremental buyer of shares near term, but projected FCF growth from the core business and RUCKUS’ high FCF conversion could allow buybacks to return in 2027.

Conclusion

Belden is a market leader across verticals entering a growth and margin inflection, with a highly accretive acquisition being priced as an overhang. The market and consensus are still stuck in the cloudy past while missing the bright green skies ahead. I am long with a core position at $112/sh.

1Q26 ER was thesis-changing for me in a more bullish direction. The print, guide, management commentary, and P14 research highlight several underappreciated drivers that could materially improve expected returns from here on out.

In this report, I will cover how 1Q26 compared to my thesis, what changed after the update, and the new aspects of the Qnity story that I believe the market is still missing.

For new readers, I suggest reading the original pitch and 4Q25 update linked below before diving in.

Note: Going forward, given how dynamic the market is and the fact that readers want new long single-stock pitches as well as screens on specific themes, I will only write full earnings updates on names where a) something notable has changed in the thesis for good or b) the stock has sold off and the thesis needs to be revisited.

This allows me to focus on putting out new and actionable ideas instead of regurgitating the same old ones. Earnings updates where a and b do not apply will be provided in short-form note format via chat. The status of each single-stock pitch is provided on the P14 website linked below.

Disclaimer: Nothing posted by P14 Capital should be considered financial advice. The author of this post holds a long position in Q. Please consult a financial advisor and/or conduct your own due diligence before making investment decisions.

Subscribe now

Q - Qnity Electronics, Inc.

NYSE: Q | 05/19/2026

Qnity delivered a solid 1Q26 print with a beat and raise across the board. Revenue came in at ~$1.32B, +17.6% Y/Y, beating consensus by ~+3.5%, while NGEPS came in at $1.08, +33% Y/Y, beating consensus by +17.4%.

More importantly, as predicted by yours truly (P14), management raised its conservative FY26 guidance by a significant margin:

• Revenue: raised to $5.225B - $5.375B, +11.5% Y/Y at the midpoint vs. prior +6.65% Y/Y.

• AEBITDA: raised to $1.535B - $1.625B, +9.65% Y/Y and 29.81% margin at the midpoint vs. prior +5.48% Y/Y and 29.98%.

The stock obviously reacted well to this, +10% on the day of the report. Since then, along with broader semis and the overall market, the stock has taken a significant breather, erasing all of its post-earnings gain.

While market participants are continually searching for a top in semis, I don’t think we are quite there yet. In any case, if there is an actual top in semis, whether caused by a macro shock such as rapid inflation, a decel in economic growth, or even worse, a hyperscaler suddenly dialing down capex spend targets, then Qnity suffers alongside the entire industry.

I am of the view that semis begin their next leg up post-NVDA ER this Wednesday, provided there is no material escalation in the ongoing conflict that leads to rising oil prices and yields. Fundamentally speaking, the company has never been in better shape.

First, let us start with the positives of Q and how they reflect my thesis.

Positives

ST is outgrowing MSI through content per wafer

ST revenue grew +12% Y/Y to $722M, ahead of Qnity’s raised FY26 MSI wafer start assumption of MSD-HSD. This was driven by advanced logic, HBM, and strong CMP consumables demand. This is an important distinction because ST is no longer simply tracking wafer volume. As customers move toward 3nm, 2nm, GAA, and HBM, each wafer requires more process steps, higher CMP intensity, more advanced cleans, and more lithography materials. This is the core reason Qnity can grow faster than MSI.

The leading-edge roadmap is moving directly toward Qnity’s strengths

Management said 3nm volumes continue to scale, 2nm activity is becoming meaningful, and angstrom-era nodes such as 16, 14, and 10 are now the primary focus of customer R&D engagement. These transitions increase the importance of surface control, defect reduction, patterning, yield, and reliability, all areas where Qnity’s materials are critical. This gives Qnity a longer growth runway than the current 2026 guide implies because the company is already working with customers before these nodes hit full volume production.

POR wins increase visibility into future ramps

Qnity had a record year of Process of Record wins in 2025, with momentum continuing into early 2026. Management specifically called out new CMP materials across pads and advanced cleans, ArF and EUV sublayer wins, AI PCB board wins, copper solder and interconnect products, and progress across thermal pads, liquid gap fillers, and phase change materials. These are not one-off wins. They suggest Qnity is being built into customer roadmaps ahead of future volume ramps.

ICS is becoming the higher-growth piece of the business

ICS revenue grew +22% Y/Y to $593M, while advanced packaging, advanced interconnects, and thermal management grew more than +50% Y/Y. This is where Qnity’s upside looks increasingly underappreciated. AI compute is creating bottlenecks around package complexity, signal integrity, heat, and power efficiency. ICS addresses those problems, which gives Qnity a growth driver beyond front-end wafer processing.

PCBs are an underappreciated new driver

The AI PCB discussion was one of the most useful parts of the Q&A. Management said next-generation AI systems require more layers, finer lines, and tighter spacing to move data reliably across the system. This supports Qnity’s metallization portfolio. Management also noted that Qnity shifted R&D toward this area during the 2023 downturn, which means the company is now benefiting from earlier product and customer work before the market fully recognizes the opportunity. This also gives Qnity fixed and variable cost leverage on PCB-related growth.

Margins are improving alongside growth

Adjusted Operating EBITDA grew +22% Y/Y to $411M, with margin expanding >125 bps Y/Y to 31.3%. ST margin was 36.4%, helped by manufacturing efficiency and favorable mix. ICS margin was 28.5%, up 280 bps sequentially, driven by volume leverage and favorable mix. The faster-growing areas are not pressuring profitability. Advanced packaging, interconnects, thermal, CMP, and advanced-node demand are improving both growth and margins.

Demand is broadening across AI, memory, and mainstream recovery

Qnity is seeing strength beyond hyperscale AI. Mainstream logic utilization improved from the mid-70s last year to the high-70s and low-80s in 1Q26. DRAM utilization moved from the mid-80s at year-end 2025 into the high-80s, with potential to exceed 90% in 2H26, while NAND also improved. Management also said premium device exposure is limiting consumer electronics risk, and AI infrastructure is more than offsetting softer pockets of demand. This gives Qnity multiple demand drivers across advanced logic, HBM, memory utilization, mainstream recovery, and premium devices.

Customer and capacity announcements

The NVIDIA collaboration, Apple American Manufacturing Program inclusion, Delaware facility opening, and Taiwan expansion all support the same point. Qnity is becoming more embedded in customer roadmaps as materials matter more to performance, yield, reliability, signal integrity, power efficiency, and thermal management. Delaware’s first line is already operational and in customer qualification, while Taiwan should be fully operational in early 2027. These investments also support critical CMP materials, one of the fastest-growing areas in Semi.

Negatives

China memory is less direct than I originally assumed

This was the clearest negative from the Q&A. I viewed China memory expansion as a potential tailwind given CXMT/YMTC capacity growth and Qnity’s local-for-local footprint. Management’s answer was more limiting. They said most of Qnity’s China semiconductor exposure is mature logic, not advanced memory, because memory usually converts faster to advanced technologies and Qnity is not selling into the most advanced technologies in China. They also said they do not have many in-depth conversations with the China memory market. This does not kill the China localization angle, but it weakens the specific China memory upside case.

ST had a restocking benefit in 1Q26

ST revenue grew +12% Y/Y to $722M, but the quarter included a $20M inventory restocking benefit, particularly in mature nodes, after customers managed inventory carefully in 4Q25. Management said this pattern was similar to 1Q25. The underlying ST demand was still strong, especially advanced logic, HBM, and CMP consumables. Excluding the restock benefit, ST grew +9% Y/Y, still materially outperforming wafer starts.

ST is expected to be ~flat sequentially in 2Q26

Not a major issue if ICS continues to carry sequential growth, but it does limit the near-term acceleration story for ST.

ICS margins face normal seasonal mix pressure in 2Q26

ICS had a strong 1Q26, with margin of 28.5%, up 280 bps sequentially. However, management said 2Q26 will include a slight product mix headwind, especially in ICS, as the business shifts into the consumer electronics season.

Middle East-linked cost pressure embedded in the guide

Management included ~$20M of raw material, energy, and logistics headwinds for the rest of 2026 tied to the Middle East conflict. This partially drove the -17 bps change in the midpoint FY26 margin guide relative to prior.

GAAP earnings are messy post-spin

Adjusted numbers were strong, but GAAP net income fell to $162M from $199M Y/Y and GAAP EPS fell to $0.72 from $0.92. The biggest issue is the standalone cost structure after the spin, including $61M of interest expense versus $0M last year, higher SG&A, and transformation costs. I think the market sees through this, but in a prolonged semis selloff we could see Qnity get clubbed into the “overvalued on GAAP P/E” basket.

New thesis elements

Qnity is exposed to ALL-THINGS AI

This point needs to be repeated, and I think the picture below provides a brief view of the number of areas Q is exposed to.

Exposure to INTC and Qnity’s importance in the domestic supply chain

Intel is quite literally the most sought-after semis asset in the U.S. It is now a play on national security, and absent a semis cycle bust, it should continue to hold up well.

The 18A ramp is the key read-through for Qnity. Intel 18A uses RibbonFET and PowerVia, with PowerVia improving standard cell utilization by 5–10% and ISO-power performance by up to 4%. Fab 52 in Arizona is reportedly designed for 18A and more advanced nodes, with potential full-ramp capacity of ~10,000 wafer starts per week. If this potential succeeds, Qnity gets another domestic leading-edge growth driver tied to reshoring, AI, and national security.

Qnity’s exposure comes through the materials consumed in the process flow. In the front end, Qnity sells CMP pads, slurries, post-CMP cleans, advanced cleans, and lithography materials. More advanced wafers require tighter surface control, more CMP steps, better defect reduction, and more precise patterning. In the back end, Qnity has exposure to advanced packaging, interconnects, metallization, bonding, and thermal materials.

Another element is Qnity’s role in an increasingly important domestic supply chain. The State Department said EXIM approved a direct loan of up to $10B for Project Vault, while CSIS describes the program as a $12B critical minerals backstop where manufacturers make long-term commitments in exchange for access to specified materials during market disruptions. Photoresists, CMP slurries, high-purity copper systems, and HBM-related materials all depend on reliable access to specialized minerals and chemicals.

While it is not confirmed that Qnity is a recipient of this program, Qnity is the type of upstream materials supplier that could benefit from this policy direction given its role in the domestic supply chain. U.S. policy is now moving beyond fab construction and into the upstream materials layer Qnity depends on. China has used mineral supply as a strategic lever for years, and Project Vault is part of the U.S. response to reduce that exposure.

Qnity’s Newark, Delaware facility is directly tied to this. The company opened a 385K sq ft facility to expand CMP materials production for AI, HPC, and advanced semiconductor demand, including a new manufacturing line for CMP pad components used in wafer fabrication. This gives Qnity domestic capacity in one of the most important recurring consumables categories for advanced chips.

Updated Value Drivers

Revenue Build

Key changes include a step-up in FY26 revenue growth to reflect guidance and the improving thesis. I think Q outperforms its midpoint FY26 revenue guide by 118 bps to reach +12.7% Y/Y, driven by strong growth in both ST and ICS. I expect ST to continue its course in 2027 as consumer electronics pressure dissipates as memory pricing potentially moves lower, and industrial and automotive end markets recover.

I also believe memory production is set to rise and bottlenecks should ease in 2027, which is a tailwind for consumables businesses like Qnity. All in all, I continue to expect ICS to be the incremental growth driver for the near term. FY26 total revenue is expected at +12.7% Y/Y, FY27 at +12.3% Y/Y, and FY28 at +9.9% Y/Y.

EBITDA & IS Build

Key changes include a small step-up in margins across the board. FY26 EBITDA margins are expected at 30.3%. I believe Qnity outperforms its midpoint margin guide for 2026, driven by operating leverage from a recurring revenue consumables business. As volumes increase, the company absorbs fixed costs more efficiently, scaling EBITDA faster than the top line. The $100M run-rate EBITDA benefit from the multi-year transformation plan set to be completed by 2028 assumes minimal to no contribution in FY26, leaving room for upside. A shift to more advanced nodes should be a pricing tailwind for Qnity, and I expect the $20M geopolitical headwind to be passed on to customers with relative ease.

All in all, FY26 margins are expected at 30.3%, 2027 at 31.0%, and 2028 at 32.0%.

Valuation and updated PT

Based on the ramping thesis and new thesis elements that position Qnity as a key player in a domestic supply chain that is shifting its focus toward materials, I think Qnity deserves a 25x multiple on FY27E EBITDA (current: 18x).

The AI trade has been all about spotting the next rotation, which I think consumables will be next. In my view, memory is still a commodity, and with all commodities, higher prices always correct with higher supply.

For Qnity, recall that this is a volume-based business. Qnity’s revenue is fundamentally tied to wafer starts and fab utilization rates, rather than the market price of memory chips. When this supply crunch ends, assuming demand is still intact, Qnity should print money.

Under my base case, I expect Q to trade at 25x FY27E EBITDA. This assumes stable debt and strong cash generation once capex calms down from the currently elevated 9% of revenue. 25x FY27E EBITDA implies 44% upside from current levels to $209/sh.

My bear case is $95/sh, -35% from current levels.

While the announcement was largely framed around securing domestic supply of transformers and other grid equipment, the broader message was clear. The U.S. electrical grid can no longer afford delayed investment. A large-scale grid rebuild is now a national priority.

By now, everyone reading this knows electricity demand is growing at the fastest pace in years, driven by datacenters, electrification, and manufacturing reshoring. This demand is colliding with a severe lack of supply, with the grid increasingly unable to respond. Median interconnection wait times have stretched to ~5 years, while much of the equipment currently powering the grid dates back to the 1960s and 1970s.

The bottom line is simple - the GRID NEEDS AN UPGRADE! Per JPM, the upcoming investment cycle could drive ~$1T of grid-related investment in the U.S. alone over the coming decade.

Among the most obvious beneficiaries of this cycle are transformer and grid equipment suppliers. Some freebies for you: 2 names I have been highlighting for weeks are American Superconductor Corporation (AMSC) and Forgent Power Solutions (FPS). AMSC shares are up ~65% since my energy storage screen was released just 3 weeks ago. FPS calls are up 100%+ since entry alerts were sent to subscribers.

Less discussed are the second-order beneficiaries. These are the businesses enabling grid expansion, modernization, engineering, and interconnection work behind the scenes. Today’s pitch focuses on one of the clear beneficiaries of the coming grid rebuild cycle, with an accelerating datacenter business and growing exposure to utility infrastructure spending.

Temporary pressures have driven multiple contraction just as the underlying business is entering a period of TAM growth unlike anything seen in years. Trading at just 1.3x FY26E Sales and ~14x EBITDA, this capex-light, net cash business offers ~70% upside under conservative assumptions in less than 2 years. In a bull case, a 2x return is achievable. My bear case is 20% downside. Q1 earnings after-hours today are likely the first catalyst for a re-rating.

Happy Reading!

Disclaimer: Nothing posted by P14 Capital should be considered financial advice. The author of this post holds a long position in the stock discussed. Please consult a financial advisor and/or conduct your own due diligence before making investment decisions.

Willdan Group, Inc.

NASDAQ: WLDN | 05/08/2026

Given the lack of time ahead of earnings today post-market, this pitch is going to be short and sweet and will follow a different format than others.

Business Overview

Willdan Group (WLDN) was founded in 1964 as a provider of engineering and planning services for public agencies in California. The company transitioned into a holding company structure and completed its IPO on the Nasdaq in November 2006. It operates today as a nationwide professional engineering and consulting firm focused on energy and infrastructure solutions. The company operates through 2 reporting segments. Energy contributes ~85% of revenue while Engineering & Consulting accounts for the remaining ~15%. WLDN effectively acts as a critical intermediary between utilities, government agencies, and commercial customers.

WLDN generates revenue primarily through fixed-price, time-and-materials, and unit-based contracts, alongside software licensing revenue. Contract duration varies by customer type but is generally characterized by long-term visibility. Utility contracts are typically multi-year in nature, often spanning 3-5 years. Typical projects are divided into 3 phases. First is the upfront policy, data analytics, and software phase, where WLDN utilizes proprietary software such as LoadSEER to provide utilities with grid mapping and long-term load forecasting capabilities. Second is technical engineering and design, where the company develops the electrical engineering plans and infrastructure blueprints required for upgrades. Third is program management, which represents the majority of contract value, where WLDN manages the entire project lifecycle from customer outreach and energy audits to coordinating subcontractors responsible for hardware installation.

Contract revenue, or gross revenue, reflects revenue prior to pass-through subcontractor costs. This reached $681.6M in FY25, +20.5% Y/Y. Net revenue reached $364.8M, +23.1% Y/Y, driven by +17% organic growth and ~6% acquisitive growth. Management typically enters each year with ~75% visibility into annual sales, which rises to 90%+ by early summer. ~75% of annual sales are considered highly recurring. Long-standing relationships with utility customers enable a “land-and-expand” strategy where upfront consulting and software deployments eventually lead to larger long-term program management contracts. Customer concentration is elevated, with the top 10 customers accounting for ~55% of revenue. Key long-term clients include LADWP, Southern California Edison, Con Edison, and PG&E.

The Energy segment is the primary growth engine of the business. The segment provides audit, program design, grid optimization, and infrastructure resiliency services, while more recently expanding into power solutions tied to datacenter growth. This segment leverages WLDN’s proprietary software and analytics capabilities to support large utility customers in modernizing aging infrastructure. A substantial portion of the company’s ~$1B funded backlog, which management states provides ~1.5 years of revenue coverage, is tied to the Energy segment. Importantly, ~89% of backlog is funded through regulated utility budgets or non-discretionary government spending. Because these contracts are often funded via municipal bonds or state-mandated ratepayer programs, backlog collection is viewed as highly resilient and predictable. Energy segment contract revenue grew 22% Y/Y to $576M in FY25 while net revenue increased 28% Y/Y to $265M. Pass-through costs primarily consist of payments to direct subcontractors.

The Engineering & Consulting segment provides specialized civil engineering and municipal financial consulting services, primarily to public agencies. This segment carries a somewhat higher margin profile due to its labor-intensive technical service mix. Contract revenue grew 14% Y/Y to $106M while net revenue increased 12% Y/Y to $100M in FY25.

WLDN operates within a highly fragmented industry where competition varies by geography, technical specialty, and project scale. Competitors range from small regional engineering firms to large multinational infrastructure consultancies. WLDN’s moat is its upstream consulting-led model, which positions the company early in the utility planning process before physical infrastructure deployment begins. This advantage is reinforced by proprietary software such as LoadSEER, which effectively becomes the forecasting and planning layer utilities rely on to model future grid demand and resiliency requirements. Switching costs become significant once utilities integrate these systems into long-term planning workflows.

Additionally, WLDN increasingly functions as a staff augmentation partner for utilities and municipalities that lack the internal capacity to manage increasingly complex regulatory and infrastructure mandates. To consolidate share within this fragmented industry, management is pursuing an acquisitive growth strategy alongside organic expansion opportunities. In FY25 alone, the company completed 3 acquisitions.

It is difficult to precisely size WLDN’s TAM. What matters is that the business is directly indexed to the accelerating need for electrical grid investment. After more than a decade of stagnant load growth, the grid is now facing a historic surge in electricity demand driven by AI datacenters, manufacturing reshoring, and broad electrification trends. One of the largest long-term tailwinds for WLDN is LoadSEER itself, which may ultimately prove to be the company’s most valuable asset. Utilities increasingly require advanced forecasting tools capable of modeling 10-year grid expansion scenarios and identifying where existing infrastructure may fail under rising demand loads. LoadSEER effectively acts as a predictive radar system for utilities attempting to navigate this transition.

At the same time, interconnection wait times across the U.S. grid have become unsustainably long. Reducing these bottlenecks will require massive investment into transmission, substations, transformers, and distribution infrastructure. This directly benefits WLDN’s engineering, consulting, and program management capabilities, particularly as utilities increasingly outsource execution capacity.

The overarching TAM driver here is grid resiliency. The recent presidential determination under the Defense Production Act formally recognized aging and constrained electrical infrastructure as a national security issue. Per JPM, the U.S. alone could invest ~$1T into grid infrastructure over the coming decade. WLDN is positioned as a clear beneficiary of this investment cycle.

WLDN’s balance sheet is pristine. The company exited FY25 with a ~$15M net cash position for the first time in over a decade. Profitability also reached record levels as Adjusted EBITDA grew 40.2% Y/Y to $79.5M, resulting in a 21.8% Adjusted EBITDA margin, above management’s long-term 20% target for the first time. FCF continues to inflect higher, reaching ~$71M in FY25, +11% Y/Y, supported by margin expansion, working capital efficiency, and minimal maintenance capex requirements.

The company is led by an experienced and stable executive team. CEO Mike Bieber assumed the role in late 2023 after serving as President and brings over 30 years of infrastructure and environmental consulting experience, including nearly two decades at Tetra Tech. He works alongside CFO Kim Early, who has more than 40 years of financial leadership experience. Oversight is anchored by Chairman Tom Brisbin, whose 50-year industry career includes founding PRC Environmental Management and previously serving as COO of Tetra Tech before leading WLDN for 16 years.

Incentives are closely aligned with shareholders through a combination of Adjusted EBITDA and organic revenue growth targets. A DSO cost-of-capital modifier further adjusts compensation based on receivables collection efficiency, with management targeting ~70 days. More than 50% of CEO compensation is equity-based, making stock performance a major driver of executive net worth.

What Brings This Opportunity

It should be obvious by now that WLDN is directly exposed to one of the strongest emerging infrastructure themes in the market. So why has the stock collapsed ~45% from its January 2026 highs?

The first reason is simple. The market completely miscategorized the business. Beginning in January, shares sold off aggressively alongside broader consulting and IT services names following the endless rollout of new Claude models and rising fears around AI disruption. Once large consulting firms such as Accenture started breaking down, investors saw “consulting” attached to WLDN and immediately assumed the company was another AI loser.

The second and far more important reason for the sell-off was management’s conservative FY26 guidance, which triggered a major multiple reset. After delivering +17% organic net revenue growth in FY25, management guided FY26 net revenue growth to just 7%–11%, implying $390M–$405M of revenue. While management framed this as a normalization toward its long-term 7%–8% growth algorithm, the market interpreted it as a sharp slowdown for what had previously traded as a premium multiple growth stock.

Adding to concerns, management disclosed that contributions from the record ~$330M LADWP contract would ramp more slowly than expected in FY26. Investors immediately feared execution risk or potential contract issues despite no evidence suggesting the contract was at risk.

The largest overhang, however, was EPS guidance. Management guided FY26 Adjusted Diluted EPS to $4.50–$4.70 versus the record $4.89 achieved in FY25. On the surface, this appeared to imply declining earnings despite continued growth. In reality, the pressure is almost entirely tax-related rather than operational. During FY25, WLDN benefited from an unusually large 31.4% effective tax benefit driven primarily by Section 179D deductions tied to energy-efficient building design projects. Under the recently passed “One Big Beautiful Bill Act,” these incentives are scheduled to expire in June 2026. As a result, management is guiding toward a materially lower ~10% tax benefit in FY26, creating an extremely difficult comparison that masks the continued growth of the underlying business.

Technicals likely worsened the decline. Following the 4Q25 earnings sell-off, shares consolidated around the ~$70–$73 range before breaking below support on 04/27 and quickly falling toward ~$65 despite no material news. The move appeared largely technical in nature, potentially driven by systematic selling or a large holder exiting the position. Since then, the stock has stabilized and recovered back above support heading into earnings today.

Why Buy

First, the market is fundamentally misunderstanding the business. WLDN continues to trade as if it belongs alongside AI-disrupted consulting firms despite operating in a completely different category. The company’s core value proposition is tied to physical grid modernization, utility engineering, forecasting, and project execution. This is a people-intensive, relationship-driven business that relies on decades of accumulated operational expertise and subcontractor coordination.

The company’s proprietary LoadSEER platform is also far more likely to benefit from AI than be disrupted by it. LoadSEER has already evolved beyond traditional econometric forecasting models and now incorporates machine learning techniques, spatio-temporal sequence prediction, and graph convolutional networks designed to model modern grid volatility. As utilities increasingly struggle with datacenter load growth, electrification, and intermittent renewables, forecasting tools become more important, not less. More importantly, software itself is not even the primary revenue driver. The bulk of WLDN’s economics comes from engineering design, program management, and implementation services.

Second, management has a long history of sandbagging guidance. The setup today looks very similar to FY25. Following the 4Q24 earnings report, management guided FY25 net revenue growth to just 8%–11% while implying only high-single-digit organic growth. What actually happened? Net revenue grew +23% Y/Y while organic net revenue growth came in at +17%, massively ahead of the initial guide.

I believe the current FY26 guidance has a high probability of proving similarly conservative. The company currently has a funded backlog approaching ~$1B, which management states represents roughly 1.5 years of revenue coverage. Even under a simplistic straight-line conversion assumption, that implies potential annualized backlog realization of ~$667M. On top of this, disclosed PR contract wins have already reached ~$188M YTD in 2026, while the large LADWP contract is expected to begin contributing meaningfully starting in 2Q26.

Aside from its core utilities and government agency business, the datacenter opportunity is ramping faster than expected. A key part of Willdan’s strategy is shifting its customer mix toward higher-margin commercial technology clients, particularly hyperscalers and datacenter developers. Through the APG acquisition, the company now provides upfront electrical engineering and planning services for datacenter projects, an area management expects to more than double in FY26. Commercial revenue reached 11% of total revenue by the end of FY25, with datacenter-related work projected to approach ~20% by the end of FY26. This business carries structurally higher margins and faster sales cycles than traditional municipal projects.

The tax normalization issue weighing on EPS should also prove temporary. While the expiration of Section 179D incentives creates a difficult FY26 comparison, this drag should normalize by FY27. Meanwhile, stronger-than-expected revenue growth combined with operating leverage from margin expansion, fixed-price contracts, and higher-margin commercial work should support upside to consensus EBITDA and earnings expectations.

Also not included in my projections is the recently announced Burton Energy Group acquisition. Burton provides data-driven energy management, procurement, and efficiency services to commercial customers across more than 60,000 monitored sites in the U.S. The business generated ~$103M of contract revenue, ~$15M of net revenue, and ~$7M of EBITDA in FY25. Management expects the acquisition to be accretive to FY26 adjusted EPS.

I ultimately view EBITDA as the cleanest way to assess WLDN’s true earnings power, particularly given the temporary tax distortion impacting EPS comparisons.

Revenue Build

Management’s long-term framework targets 15%–20% total revenue growth alongside high-single-digit organic growth. I project contract revenue growth of +14.8% Y/Y in FY26 to ~$783M, slightly below consensus expectations of ~$786M. Beyond that, I project +12.8% growth in FY27 and +11.7% in FY28.

Management has discussed a long-term objective of reaching $1B in contract revenue. Given the structural tailwinds tied to domestic grid modernization, utility infrastructure investment, and datacenter-related electrical engineering demand, I believe this target is achievable by 2028–2029, if not earlier should management continue pursuing acquisitions aggressively.

On net revenue, I continue to expect growth to outpace contract revenue as subcontractor efficiency improves. Investor concerns intensified in late 2025 after subcontractor costs rose to 48.5% of contract revenue during 4Q25, weighing on margins and creating fears around operating leverage. Management’s stated objective for FY26 is improved cost absorption, which effectively means shifting more work toward internal engineering and technical staff rather than relying on external subcontractors for higher-value services.

Utility advisory work and program management contracts generate substantially stronger net revenue conversion than construction management-related activity due to lower pass-through content. As the mix shifts further toward utility optimization, software-driven forecasting, commercial datacenter work, and fixed-price engineering contracts, I expect continued leverage on pass-through costs.

As a result, I project net revenue growth of +15.7% Y/Y in FY26, followed by +13.7% in FY27 and +11.9% in FY28.

EBITDA Build

The primary drivers of margin expansion are mix-related. WLDN continues shifting toward higher-margin commercial and datacenter customers while also increasing the percentage of fixed-price contracts within the overall portfolio. Both trends should continue supporting operating leverage over the next several years.

The commercial datacenter business in particular carries stronger economics than traditional municipal and utility contracts due to faster project timelines, lower bureaucratic friction, and higher-value engineering requirements. At the same time, fixed-price contracts allow WLDN to capture incremental efficiency gains internally rather than simply passing labor costs through to customers.

Beyond gross margin expansion, corporate G&A should also see meaningful leverage. As the company scales toward its long-term multi-billion-dollar contract revenue opportunity, the supporting corporate infrastructure does not need to grow proportionally with revenue. This should allow overhead costs to increase materially slower than the top-line over time.

I project GAAP EBITDA margins of 10.2% in FY26, modestly below consensus expectations of ~11%, followed by expansion to 11.7% in FY27 and 13.1% in FY28 as mix shifts, operating leverage, and internal labor utilization improve.

Valuation

Under my base case, I value WLDN at 20x FY27E EBITDA. This implies modest multiple expansion from the current ~18x TTM EBITDA multiple, driven by the market increasingly recognizing WLDN as a direct beneficiary of the grid upgrade investment cycle. Similar thematic re-ratings have already occurred across names such as AMSC and FPS.

The ~15x contraction from the peak ~33x TTM EBITDA multiple appears excessive in my view given the stability and continued improvement of the underlying business.

Applying a 20x multiple to my FY27 EBITDA estimate, while assuming continued cash generation, moderate acquisition-related debt usage, and ~5% and ~4% share dilution in FY26 and FY27 respectively, results in a base case price target of ~$126/share, implying ~66% upside from current levels. I believe this target can be achieved within the next 2 years.

Under a bull case scenario, upside approaching 2x current levels is achievable should the datacenter business ramp materially faster than expected and commercial mix continues improving margins ahead of my model assumptions.

The largest risk to the thesis is a delay in utility program implementation. This occurred previously in 2022 when results were impacted by slower California IOU program ramp-ups and delayed project timing. Given the current condition of the U.S. grid and accelerating electricity demand growth, I view a repeat of that environment as less likely today.

Competition is another risk. The accelerating grid modernization TAM could eventually attract larger engineering and consulting firms into the space. My view is that WLDN’s utility relationships, software integration, operational expertise, and subcontractor ecosystem provide a meaningful moat, though this remains worth monitoring.

Near-term earnings volatility is also possible. It would not surprise me to see a softer-than-expected 1Q26 given that the large LADWP contract is expected to ramp beginning in 2Q26 rather than contributing immediately.

Under a bear case scenario, I estimate downside of roughly 20% to ~$60/share. Technically, a decisive break below the key ~$70 support level would likely be the first major warning sign that the thesis is deteriorating.

Conclusion

I view the current risk/reward as highly attractive. WLDN is currently a 4.5% position in my portfolio with an average cost basis of $74.28/share. I will decide whether to add or reduce exposure following management commentary and the Q1 earnings release tonight.

Temporary technical pressures are highlighting Mr. Market’s ignorance. This is a rare opportunity to capture a fundamental re-rating, where the upside reflects a business’s future turning from coal to diamonds.

Should my base case play out, I see 50%+ upside in the near term (6–12 months) and ~100% over a 2-year holding period, while the bear case points to ~10–20% downside.

Disclaimer: Nothing posted by P14 Capital should be considered financial advice. The author of this post holds a long position in the stock discussed. Please consult a financial advisor and/or conduct your own due diligence before making investment decisions.

Hallador Energy Company.

NASDAQ: HNRG | 04/24/2026

Thesis Summary

Hallador is a vertically integrated IPP and fuel supplier operating in Indiana, with exposure to accredited capacity, wholesale energy, and coal. The company is transitioning from third-party coal sales to contracted energy generation following the 2022 acquisition of the ~1,080 MW Merom plant, supported by low-cost fuel from its nearby Oaktown Mining Complex. Merom currently runs at ~65% utilization, leaving ~350 MW of idle capacity, with a ~$1.3B forward sales book through 2029 and meaningful uncontracted exposure.

The stock is down ~-16% YTD and ~-30% from ATH due to a few reasons. The largest is the failed exclusivity agreement with a data center developer in 2025, which reset expectations around a near-term IPP. Management has maintained a bullish tone, but no deal has been signed and Mr. Market is now sick of waiting. A second overhang is a 60-day planned maintenance outage beginning May 2026, tied to prior reliability issues, which drove a 4Q25 miss and weak 1Q26 guide. Additional pressure came from a recent share lockup expiry and broader selling in coal-linked names during the recent broader energy selloff.

At $15.79/share, the market assigns no value to a data center PPA and minimal value to the 515 MW natural gas expansion.

The core thesis does not require either catalyst. As legacy contracts roll off, HNRG benefits from higher pricing in a tightening MISO Zone 6 market with a ~4.7 GW capacity deficit. Capacity revenues step up at ~2x prior pricing through 2026–2028, adding ~$86M of incremental revenue, while ~90% cost control drives strong operating leverage.

On top of this, two catalysts add upside. A ~$6.5B, ~2.1 GW data center is being developed directly opposite Merom, where interconnection constraints make HNRG a likely supplier. The 515 MW gas expansion increases capacity by ~50%, extends asset life, and enables long-term PPAs. Both are not reflected in the current valuation.

Recent developments have been largely ignored by the market, though the sell-side is beginning to adjust, with an average PT of ~$27.75 that still assigns limited probability to a signed data center PPA.

Under the P14 base case, which excludes a signed data center IPP and assumes only core business execution with progress toward both catalysts, I see a re-rating to 10x FY27 EBITDA (minimal expansion from the current 9.5x TTM), implying ~48% upside to ~$24/share. In a scenario where a data center IPP is confirmed and the natural gas expansion continues to advance, I see multiple expansion to 14x FY27 EBITDA, implying ~109% upside to ~$33/share.

Business Overview

Hallador Energy Company operates as a vertically integrated Independent Power Producer and fuel supplier primarily in Indiana. The company evolved from its legacy roots as a bituminous coal producer into an integrated energy provider following the October 2022 acquisition of the Merom Power Plant. This shift moves HNRG up the energy value chain from a commodity price taker selling coal to third parties to a utility-scale producer of electricity and accredited capacity. In 2024, the company formalized this transition by changing its SIC code to electric services, aligning the business with dispatchable generation and expanding power market margins.

Operations are organized across Electric Operations and Coal Operations, anchored by the 1,080 MW Merom Generating Station and Oaktown Mining Complex. In FY25, Electric Operations generated $310.7M of revenue and $85.4M of segment EBITDA. Coal Operations contributed $148.7M in third-party revenue and $20.1M in segment EBITDA. HNRG generates revenue across three primary products including accredited capacity, wholesale energy, and thermal coal. A ~$1.3B forward sales book through 2029 provides multi-year revenue visibility. For 2025, ~71% of energy production was contracted, leaving 29% exposed to spot pricing.

HNRG’s commercial strategy is built on long-term utility relationships with customers such as Vectren, Duke Energy, and Hoosier Energy. Hoosier Energy is the largest customer and accounted for 23.4% of total revenue in 2025. Electric sales are concentrated, with three customers representing 89% of delivered energy revenue.

HNRG operates in MISO Zone 6, covering Indiana and parts of Kentucky, where a structural supply deficit is emerging. Demand is driven by the need for reliable baseload power tied to AI buildout, industrial onshoring, and large-scale data center investment. MISO projects an average 4.7 GW generation shortfall during summer seasons through 2031, and NERC has designated the region as High Risk for supply shortfalls as early as 2028. This backdrop supports sustained demand for dispatchable generation as large load customers prioritize reliability.

While HNRG competes with coal producers such as Peabody Energy and Alliance Resource Partners, its model benefits from vertical integration and infrastructure ownership. The company controls ~90% of variable energy costs by supplying its own fuel at cost. In December 2025, MISO accepted HNRG’s application for a 515 MW natural gas expansion at Merom through the ERAS program. HNRG benefits from high barriers to entry and a competitive advantage with flexibility for future energy diversification (renewables, etc.)

HNRG maintains a disciplined capital allocation approach, reinvesting over $230M into the business since 2022 while reducing leverage. Total bank debt declined to $30M by year-end 2025, with total debt of $38M including finance leases as of 4Q25. In March 2026, the company closed a new $120M 3-year senior secured credit facility. Management alignment is strong, with officers and directors owning ~17% of the company. Recent leadership changes include CFO Todd Telesz, appointed in June 2025, whose background in power cooperatives supports the shift toward long-term energy contracting.

Core Business Thesis

Excluding a likely data center PPA and the ERAS natural gas expansion, HNRG’s core electrical and coal operations are positioned to generate excess returns in the years ahead.

The core thesis for Hallador Energy (HNRG) for 2026–2029 is driven by a step-up in pricing and margin as legacy low-priced contracts roll off and the company captures a tightening supply backdrop in MISO Zone 6. Even without a data center PPA or ERAS gas expansion, the base business is transitioning toward a higher-margin IPP model, supported by ~$1.3B of contracted forward sales that provide multi-year visibility.

Electric Operations

Electrical sales consist of two components, delivered energy and capacity revenues. Delivered energy includes both contracted and uncontracted volumes. At its core, electrical sales are driven by MWh delivered as the volume metric and price per MWh as the price metric, with realized pricing a function of contract mix and market exposure.

Nameplate capacity for HNRG’s operations is 1,080 MW, which translates to ~6.0M MWh of potential annual generation at current utilization rates. Merom does not operate at full capacity due to periodic operational disruptions. Following the major maintenance outage in 1Q26 and the benefits from vertical integration with secured fuel supply, I expect realized generation to move closer to target levels post-2026. This should reduce reliance on purchased power, which has historically been used during maintenance periods or when market prices at the Merom Hub fall below variable production costs. As plant availability improves, internal generation should displace these purchases and expand margins.

Alongside higher MWh sold, my base case assumes pricing above the forward energy curve. The forward curve for the next six years provides a baseline but does not reflect tightening conditions in MISO Zone 6. The region faces a 4.7 GW summer capacity shortfall through 2031, while new generation projects face interconnection timelines exceeding seven years. Existing dispatchable assets such as Merom are among the few near-term solutions to meet incremental load. This supply-demand imbalance supports bilateral contracting at pricing above the wholesale curve. I assume a 1% premium to the curve, which I view as conservative given the scarcity value of existing capacity.

Delivered energy revenue is expected to decline -9% in 2026 due to the planned outage, followed by +10.7% annual growth in 2027 and +11% in 2028 as availability improves and pricing resets higher.

Capacity revenues, which compensate generators for maintaining available supply, are set to increase meaningfully over the next three years. HNRG has locked in capacity pricing at roughly 2x prior levels for substantially all remaining accredited capacity with an unnamed utility customer for 2026–2028. This contract is expected to contribute ~$86M of incremental revenue over that period. I model these contributions beginning in 2027, implying a ~$43M annual increase in 2027 and 2028.

Management has been clear that intermittent generation cannot provide consistent accredited capacity due to lack of dispatchability. Solar is typically accredited at ~5% of nameplate capacity and wind at ~15%, while coal-fired generation provides stable 24/7 output. Grid operators and large industrial users require this reliability during peak demand and stressed conditions, which supports elevated capacity pricing.

Capacity revenues for 2026–2029 are disclosed in the FY25 10-K. Incorporating the accredited capacity contract, I model capacity revenue growth of +6% in 2026, +53% in 2027, and -15% in 2028, reflecting the step-up and partial normalization thereafter.

Given the scheduled outage in 1Q26 and management guidance for a weak quarter, Electrical Sales are expected to decline Y/Y in FY26. I model -6% Y/Y in FY26, followed by +20% Y/Y in FY27 and +4% Y/Y in FY28, driven by higher utilization and improved pricing.

Coal Operations

As noted above, a large portion of coal produced by Sunrise is consumed by Merom through intercompany sales, which were ~47% in FY25. Third-party coal sales are driven by the deficiency in MISO Zone 6 and the ongoing Hormuz conflict threatening energy security globally. Countries are increasingly focused on diversification across renewables and dispatchable sources such as coal. As a U.S.-based producer, HNRG stands to benefit directly from this demand.

In late 2025, the administration announced $625M in funding directed at coal-fired power generation in the U.S. Management believes the company has projects that could qualify for grants from this fund. The administration has also explicitly called out coal as a power source for AI-related demand. There is not enough power to support data center growth, and coal provides reliable baseload generation at scale. The bottom line to take from this is that coal will be part of the generation mix for longer than current expectations.

Coal sales do not drive the core alpha of the HNRG thesis, but they serve as the backbone of its electrical operations as the primary fuel source for Merom. I project coal sales to grow at a steady-state ~2% annual rate.

These assumptions translate into a simple revenue build that implies total revenue growth of -4.5% in FY26, followed by acceleration on easier comps to +13.5% in FY27 and +3.3% in FY28.

Margins

At a high level, margin expansion is driven by operating leverage from a fully vertically integrated model that controls ~90% of variable costs. By owning Sunrise Coal as its primary supplier, HNRG reduces transport costs while limiting exposure to supply volatility and market pricing.

Accredited capacity revenues carry near-100% margin since there is no incremental cost to deliver capacity. Management targets $65M of annual capacity sales, a level that will be exceeded once the new capacity contract takes effect. This ~$65M effectively covers ~$60M of fixed operating costs, meaning the plant covers its cost base before any energy is sold. Beyond this point, energy margins, which management believes can exceed $25 per MWh, flow through to EBITDA with minimal friction. As generation scales toward the ~6M MWh target, additional variable cost leverage comes through. Running the plant at a steady baseload level is ~10% more fuel-efficient than intermittent dispatch with frequent start-stop cycles seen in prior periods.

Coal segment margins are supported by a structurally lower cost base following the 2024 reorganization, which repositioned the segment as a focused, at-cost fuel supplier to Merom. By idling higher-cost surface operations and consolidating underground equipment at the Oaktown Mining Complex, HNRG has reduced cash costs into the low-$40s per ton.

After a temporary impact in 2026 driven by scheduled maintenance, I expect EBITDA margins to recover and expand ~110bps to 23.3% by FY28, driven primarily by improvements in Electric Operations.

The key takeaway is that the core business is set up for excess returns through both revenue growth acceleration and margin expansion over the next three years. On this basis alone, HNRG is an attractive long with solid upside over the next two years. What follows is what makes this pitch really juicy.

Catalysts

Aside from the core business thesis, which in my view is not priced into shares, there are two major catalysts that act as free call options on HNRG, both with a high likelihood of playing out.

ERAS Natural Gas Expansion

HNRG is advancing its ERAS natural gas expansion, adding 515 MW adjacent to the Merom Power Plant, representing ~50% growth in total generation capacity. The project leverages a 100% owned ~1 GW interconnection and existing site infrastructure originally built for a third unit. This provides a clear speed-to-market advantage with a targeted 3Q29 completion, a ~3Y timeline versus 7Y+ for greenfield builds, while avoiding ~$300M in system upgrade costs.

This removes the terminal value overhang tied to a coal-only asset. By adding 515 MW of gas generation alongside the existing units, Merom becomes a dual-fuel facility. This extends the useful life of the plant beyond local coal reserve depletion, projected around 2035–2036, and keeps the site relevant as grid infrastructure for decades. The expansion is also required for the long-term PPAs management is negotiating with data centers and utilities. These counterparties require 10–15 year contracts, which cannot be supported under coal-only retirement timelines. With the gas build, Hallador can offer a 10+ year runway of reliable accredited capacity, positioning the asset to meet regional load growth through 2040. The project also aligns with a ~4.7 GW capacity deficit in MISO Zone 6 and increasing demand from ~2.1 GW of nearby data center development.

To support this catalyst, this year’s proxy and executive compensation plan explicitly includes the “execution of an interconnection agreement to facilitate the development of a natural-gas fired generating facility at the Company’s Merom Generation Station, and goals related to financing and contracting” as a target for bonuses. This language was not included in the 2025 proxy or prior compensation 8-Ks.

Management is clearly incentivized to execute on the key catalysts outlined in this thesis.

Datacenter IPP

This is not the first time HNRG has been in advanced discussions for a data center IPP. The company previously reached late-stage negotiations with a global developer in late 2024, which resulted in a non-binding term sheet and an exclusivity agreement effective January 2, 2025. The agreement included up to $5M of payments to Hallador through June 2025. Despite progress and capital deployment by the counterparty, the deal did not close as the exclusivity period expired without definitive agreements. Management pointed to the complexity of aligning multiple parties, including the utility and hyperscaler, and shifted to a non-exclusive process to capture improving market conditions. The stock declined ~20% on the announcement.

Primary research and regulatory filings indicate a higher probability of a data center deal by 1H26. This could be structured as a formal Data Center IPP or as a long-term capacity agreement with a hyperscaler through a utility sleeve, likely involving Hoosier given the existing relationship. Either structure represents a meaningful positive outcome for HNRG.

A ~$6.5B, 2.1 GW data center campus, Heartland Industrial Park, is under construction directly opposite the Merom plant, with construction expected to begin imminently, with some reports stating as soon as today (04/23).

Sullivan County is a strong location for data centers given the lack of zoning restrictions, but it is also energy deficient as MISO Zone 6 is constrained. New generation faces 7+ year interconnection queues, making it unlikely that greenfield supply will be secured in time. Existing interconnection capacity is needed.

What does that leave Heartland Industrial Park with? Merom has a strong chance of serving as an energy supplier. The site is expected to host a 2.1 GW data center, likely built by Fluidstack, a developer with backing from hyperscalers such as Google and recent funding through a record high-yield bond issuance.

Interestingly, HRNG’s inked accredited capacity deal would expire by 2028-2029, which aligns with the expected timing of the data center coming online.

Regulatory filings also indicate HNRG’s role in the supply chain. Recent filings list Hoosier Energy, Merom’s transmission partner and Hallador’s primary utility counterparty, as the submitting Transmission Owner, with explicit references to data center load.

In the 4Q25 earnings transcript, management highlighted additional competitive offers and confirmed that draft contracts have been exchanged with multiple parties for accredited capacity over a multi-year period. It is increasingly likely that HNRG will supply power to the data center being built opposite Merom. While it may not serve the full load, its ~350 MW of idle capacity, expiring capacity contracts by 2028–2029, full interconnection ownership, and existing relationships with Hoosier and other utility sleeves position HNRG well in this situation.

Base case PT

My base case excludes the impact of a potential data center PPA and the ERAS natural gas expansion. I typically do not model beyond three years, so my projections run through 2028. Over this period, HNRG is well positioned on its core business alone.

Execution in line with the core thesis warrants at least a 10x multiple on FY27E EBITDA, a modest step-up from the current ~9.5x TTM. This implies a base case PT of $23.64/share, representing 48% upside. The model assumes stable cash and debt and ~7% dilution from the recent share offering.

This base case does not include the financial contribution from the two main catalysts, but it assumes steady progress toward both. HNRG may not announce a data center IPP within the forecast period, but ongoing discussions support the setup. The ERAS natural gas expansion is assumed to progress without regulatory delays.

Bull case PT

Assuming confirmation of a data center IPP and continued progress on the natural gas expansion, HNRG could re-rate to ~14x EBITDA, which is at the low end of utility peers such as Constellation Energy, Vistra Corp, Talen Energy, and Entergy.

Revenue from a data center IPP would likely begin in 2028–2029 and would be priced above the forward curve. With capacity rolling off into that period, HNRG is positioned to allocate supply into higher-value contracts, supporting stronger margins.

While not explicitly modeled, current supply constraints in MISO Zone 6 suggest that unit economics for such agreements would exceed existing contracts. A 14x multiple on FY27E EBITDA implies 109% upside to $33.35/share.

Bear Case and Risks

Key risks include additional maintenance issues. A severe downside scenario would involve sustained operational disruptions at Merom or the Sunrise coal complex, which supplies the plant.

Another downside case would involve management confirming that no data center IPP discussions are ongoing. This appears unlikely, but it would remove a key source of optionality. The current valuation does not reflect a data center deal, and investor sentiment has already been impacted by delays and the absence of a signed agreement.

A separate risk is a regulatory delay or setback for the ERAS natural gas expansion. Policy risk also exists, particularly around the midterm elections, as current support for coal-fired generation may shift.

In a near-term downside scenario, HNRG could trade to ~$14/share, implying ~-12% downside.

Conclusion

The risk-reward is attractive given the strength of the core business and the potential for near-term catalysts, including a data center IPP announcement or updates during the 1Q26 earnings call.

I am long HNRG as a core position at an average cost of $15.70/share.

In the P14 view, nothing has changed. Investors must be willing to look through a conservative guide and accept temporary pain for future upside. In fact, the long thesis has strengthened.

Do not follow blindly. This is a small-cap Chinese ADR with its own unique risks and requires active trading. Treat me as the research guy.

Read the original thesis linked below for background before this update. Here, I dive deeper into what has driven the downside in shares and where future upside is coming from. My $18 PT in <2Y is UNCHANGED.

As always, if you have any questions or want to discuss stocks or any asset class, feel free to DM.

Disclaimer: Nothing posted by P14 Capital should be considered financial advice. The author of this post holds a long position in VNET. Please consult a financial advisor and/or conduct your own due diligence before making investment decisions.

VNET Group, Inc.

NASDAQ: VNET | 04/10/2026

Note: I have updated my model to reflect ¥ (CNY/RMB) figures only. The only conversion to spot occurs in my price target, where I convert future equity value to USD (as this is a US ADR). The spot used is RMB/USD = 6.8824 (source: FRED, 04/10/2026). In addition, I have shifted wholesale IDC pricing from annual ARR to monthly MRR for a smoother comparison, though it remains heavily influenced by annual net adds in capacity.

VNET delivered a small beat in 4Q25 with revenue of ¥2.69B (+20% Y/Y), beating consensus by 2.56%, and AEBITDA of ¥805M (+12% Y/Y), beating consensus by 4%.

Yet the stock dropped 9% the next day and has since been dragged lower, driven by the ME conflict. If not obvious already, the market clearly did not like the 2026 guidance:

• Revenue: ¥11,500M - ¥11,800M (+15.59% - +18.60% Y/Y), +17.09% Y/Y at the midpoint. This guidance met consensus estimates for 2026, but the market likely reacted to the Y/Y deceleration in growth.

• AEBITDA: ¥3,550M - ¥3,750M (+19.20% - +25.91% Y/Y), +22.56% Y/Y at the midpoint, implying margin expansion of 140bps Y/Y to 31.33%. Again, this met consensus and continues to show operating leverage in the business.

At first glance, it appears the stock has dropped due to the perceived revenue deceleration, alongside a broader selloff in China equities driven by the ME conflict. This is somewhat counterintuitive given China DC capex is ramping faster than expected.

Let us start by looking at the negatives and address the risks tied to the ongoing ME conflict, and how exposed VNET is as a China data center operator.

Negatives and Hormuz Risk

Starting off with the obvious, why project a decel in revenues if your order book is visibly accelerating? There are a few reasons.

CCP (China Conservative Party) guide - in 4Q24, management guided to FY25 revenue growth of +10–13% Y/Y whereas actual FY25 growth came in ~900bps higher than the guided midpoint of 11.5%, at 20.5% Y/Y. It is safe to assume some conservatism embedded in the FY26 guide, especially given VNET’s position and order book momentum.

Looking deeper, recall VNET’s revenue is a function of capacity utilized or the utilization rate (UTR) × price (MRR/MW). Capacity utilized depends on two main factors, one being the capacity VNET delivers on a yearly basis and two, the customer move-in rate.

Wholesale UTR dropped to 70.1% at the end of December 2025 compared to 74.3% in the previous quarter and 72.6% in 4Q24. It is crucial to understand that this is not due to customers not utilizing their data centers but rather a function of delivering a massive amount of new capacity (107MW) at the very end of the year, which initially dilutes UTR until customers move in.

Importantly, UTR for mature capacity is still within the healthy 90–95% range. New capacity is quickly becoming mature, as move-in rates are accelerating to a ~6–12 month window compared to ~24 months historically, driven by token demand and AI workloads in China.

Looking ahead to 2026, management expects UTR for the full year to come in at 70–75%, driven by an expected record MW delivery year.

The other piece of the revenue puzzle is pricing. This is difficult to isolate given the large swings in deliveries, but the pricing picture is more or less the same, stable to rising in the near to mid-term. TThis does not apply equally across the industry and is more relevant for VNET, given Ulanqab is a prime area for its natural cooling and proximity to demand centers, where VNET has secured meaningful land and power quotas.

The only other negative for me was the Y/Y contraction in EBITDA margin. Margins came in at 30%, +63bps Q/Q and -211bps Y/Y. The Y/Y contraction was primarily driven by one-off gains in 4Q24 where margin was inflated by a ¥87.7M disposal gain from the E-JS02 data center. Excluding this item, 4Q25 AEBITDA represented an increase of 39.3% Y/Y. Adj. operating expenses as a % of revenue also rose to 14.1% in 4Q25, compared to 10.2% in 4Q24, driven partly by higher agency service fees and personnel costs associated with the company’s expansion. Management noted that delivering a high volume of cabinets at year-end causes a scale-up of costs that hit the P&L immediately, while the offsetting revenue from customer move-ins ramps up over the subsequent ~6–12 months.

Other non-thesis altering negatives include no update on a potential HK IPO. This would meaningfully lead to a closure in the valuation gap between VNET and GDS, but management has yet to offer any update in this regard.

The Hormuz Risk

The way Chinese stocks have sold off much harder than US tech, it is almost like war is being waged on China rather than Iran. While one could make the “allies” argument, the focus should be on where China stands amidst all this.

China maintains strong relations with both Iran and Russia, two major oil producers. While flows from Iran are constrained but still moving to China, Russia remains a major supplier. Moreover, China has world-leading reserves estimated at 1–1.3B barrels, which at current consumption rates should last ~90–120 days. Combined with a well-established renewable energy ecosystem, China should be relatively protected in a ~3 month closure scenario.

So how does this affect VNET? Construction exposure is moderate. Data center build-outs are steel and concrete intensive. A closure through May drives up the cost of global shipping and refined products like diesel and naphtha. However, VNET has already contracted much of its 2026 pipeline (450–500MW target), so price locks likely mitigate near-term volatility in construction inputs.

As for power costs, VNET’s China operations are tied to a grid that is 60%+ coal and increasingly renewable. Electricity prices are regulated, and the government typically limits pass-through of global energy spikes to industrial users during a crisis, prioritizing AI and broader productive capacity.

Given insatiable demand, supply-side cost pressures can be passed through to customers, but the outcome depends on the duration of the SoH closure. My view is that China and thus VNET should be fine in a closure scenario through the end of May. Beyond that, risks increase, including potential demand slowdown driven by higher cost of living.

A more immediate downside risk to China ADRs is renewed tariff threats. The path for this is unclear given the May 14–15 presidential summit in Beijing and what appears to be an attempt to improve US-China relations.

In my view, this conflict has to end or meaningfully slow soon. When? I don’t know, but it needs to happen soon enough given the global impact. We have already passed the point where disruptions are feeding into inflation globally, and for that not to worsen, oil and other input flows need to normalize sooner rather than later.

Positives and Recent Developments

Unbeknownst to Mr. Market, 4Q25 and recent developments have strengthened the long VNET thesis.

Strong delivery guide - the real alpha in DC players to gauge future growth is expected deliveries, which in VNET’s case surprised consensus expectations of 400–450MW in 2026 with management guiding to 450–500MW. In my view, this is still conservative, given VNET has likely secured a blockbuster 500MW deal with ByteDance, which would be the largest hyperscaler deal in the China DC space. This was likely won due to VNET’s position in Ulanqab and its Hyperscaler 2.0 framework.

The ability to deliver ~400MW in ~6 months is an underrated advantage that VNET possesses and is something compute-hungry China hyperscalers are actively seeking. Management noted that traditional IDCs are shifting toward AI-ready data centers (AIDCs) to meet demand for large language model training and inference.

Management did not confirm this deal, and there is no historical precedent for such confirmations, but did note that an update on the 1Q bidding cycle would be shared in the 1Q26 earnings call.

Locked in Capacity - what gives further confidence in this delivery guide is that 150–156MW of the 2026 delivery plan is already locked in or pre-committed. The company entered 2026 with strong momentum, securing 135MW of new wholesale orders in 4Q25 alone across internet, cloud, and intelligent driving customers. In management’s words, Rotating President Xiao Liu stated the company is “very highly confident” in filling the remainder of planned 2026 capacity.

Massive Resource Bank - VNET has increased total wholesale resource capacity to 2.2GW as of the end of 2025, with 840MW held for future development.

Positive updates on financing - management has raised $150M in a private placement, completed a ¥6.36B private REIT listing on the Shanghai Stock Exchange, and issued a ¥860M green ABS, both at ~13–14x EV/EBITDA valuations, which VNET trades at an irrational discount to. The company also holds a domestic AAA credit rating, supporting attractive borrowing costs and access to science and tech innovation bonds. Overall, funding risk for the ¥10–12B 2026 capex plan appears manageable. There are still no updates on a C-REIT, but these private REIT listings are a step in the right direction.

DeepSeek V4 and domestic chip proliferation - V4 has not yet been released, with recent reports pointing to a late April launch. A delay may mute near-term stock reaction, but more importantly, DeepSeek V4 is expected to be the first frontier model running natively on Huawei silicon, specifically Ascend 950PR. A key dynamic here is that domestic AI chips are ~70% less power-efficient than Nvidia GPUs. Given VNET’s model charges based on power consumption, lower efficiency translates directly into higher revenue. In addition, higher power requirements increase cabinet density needs, driving incremental wholesale demand. A successful V4 release, especially if competitive with US models, would be a clear upside catalyst.

Interestingly, DeepSeek is looking to expand into Inner Mongolia…

Updated Value Drivers

Key changes to revenue build:

Wholesale IDC capacity service - I expect upside to management’s delivery target of 450–500MW, but without confirmation on timing of ByteDance-related capacity, I model 475MW added in 2026. For 2027, I conservatively estimate 300MW of deliveries.

Wholesale UTR - I take the midpoint of management’s 70–75% guidance (72.5%) for 2026 and model a rise to 80% in 2027, driven by move-ins and maturation of recent capacity.

Wholesale MRR/MW - I shifted from ARR to MRR/MW and now use the average of beginning and ending wholesale utilized capacity as the denominator. Pricing is difficult to isolate given swings in deliveries, so I project an -8% decline in 2026 to keep the revenue build grounded, broadly in line with consensus. This does not imply pricing pressure, but reflects the mix impact from new capacity. I hold MRR/MW flat in 2027.

Overall wholesale revenue comes in at +51.9% Y/Y in FY26 and +43.9% Y/Y in FY27.

Retail IDC revenues grow +2% in 2026 and +4% in 2027, driven by AI inference demand and new model deployment. Utilization and cabinet growth support this.

Non-IDC revenues held flat.

I see total revenue coming in at +18.9% in 2026, slightly above the high-end guide of 18.6%, and accelerating to 20.9% in 2027 driven by higher utilization.

Key changes to EBITDA build:

I expect margins to come in slightly above the high-end guide at 31.9% in 2026 and expand ~130bps in 2027 to 33.2%, driven by operating leverage and mix shift toward wholesale.

Base Case Valuation

Continue to expect a re-rate to 15x EBITDA on FY26 figures in <2Y, implying ~114% upside to ~$18/sh.

What takes shares to $18 includes successful China model launches such as DeepSeek V4, confirmation of the 500MW ByteDance deal in 1Q26, upside to consensus and management’s CCP guide, a C-REIT listing, what now looks like a slimmer chance of an HK IPO, and continued hyperscaler wins.

Shares would also likely respond to any sustained positive developments around the SoH. In the near term, a rally could be further supported by high short interest at ~13.8% (Seeking Alpha).

At the current valuation (~8.6x FY26E EBITDA), I would not discount the possibility of a PE take-private followed by a domestic relisting in China, given the valuation gap between infrastructure assets in China (14–15x) and the geopolitical discount applied to Chinese ADRs in the US. There is precedent here with Bain’s take-private of Chindata in 2023 for $3.16B, followed by a partial exit in January 2026 for $4B.

Risks and Bear case

Systemic risk via oil and a prolonged SoH closure.

Execution risk on the ByteDance deal, though given the pace of deliveries, I would discount this.

Chinese LLM risk, for instance if DeepSeek V4 materially underperforms.

China-US relations risk impacting Chinese ADRs. In my view unlikely given the upcoming conference, but still a risk.

A more structural risk is a slowdown in the global AI buildout. If newer models become materially more efficient and reduce compute intensity, demand for data center capacity could soften, leading to a repricing across DC and semi names, including VNET.

In a bear case scenario, I project downside to $7.44/share, -11.6%, driven by margin contraction to 29.2%.

Conclusion

I view the risk/reward over a ~2Y horizon as very attractive. I am long VNET but not at full position size given my shaky global macro view.

Out of the many themes that benefit from this trend, an obvious one is energy. Within energy, several subsectors stand to benefit. One that is underrated is energy storage.

What is Energy Storage

When people hear about this space, they often mischaracterize it as simple battery manufacturing. The reality, however, is far more complex, involving a multi-layered value chain and diverse applications.

At its core, energy storage allows energy to be moved across time and location, breaking the link between when power is generated and when it’s used. In essence, it transforms energy from a perishable commodity into a moveable asset.

As the grid evolves with the energy transition, storage becomes essential. Legacy systems rely on large turbines and fossil fuels to balance supply and demand. Today’s and tomorrow’s grid requires flexible, distributed storage to handle the surge in power demand and the variability of renewable energy sources. Storage is now a core part of grid infrastructure, industrial operations, and residential energy independence.

The Value Chain

Upstream - Raw Materials and Chemistry: The supply chain begins with the extraction and refining of critical minerals, primarily lithium, nickel, cobalt, and graphite. China continues to dominate this segment, particularly in graphite anode production and LFP cathode manufacturing.

The industry has pivoted decisively toward Lithium Iron Phosphate (LFP) chemistry for stationary applications. LFP is cobalt-free, lower-cost, and safer than traditional Nickel Manganese Cobalt (NMC) batteries, though it has lower energy density. LFP accounted for 93% of new stationary storage deployments in 2025.

Emerging chemistries, such as sodium-ion, are gaining traction as a hedge against lithium price volatility. Sodium-ion batteries do not rely on lithium, nickel, or cobalt and perform significantly better in cold climates, retaining 90% capacity at -40°C. While they currently struggle to match LFP’s cost at today’s lithium prices, mass production scaling could make them up to 20% cheaper than LFP by the late 2020s.

Midstream - Cell Manufacturing and System Integration: Manufacturing capacity in this vertical is heavily concentrated among a few major players. CATL holds a dominant 32% global share, followed by BYD and Korean leaders like LG Energy Solution and Samsung SDI.

Almost all of the top 10 midstream companies are ramping production in 2026, with estimates expecting 50%+ growth Y/Y for stationary energy storage. Stationary storage still only makes up ~26% (450 GWh) of top 10 player battery production in 2025.

While stationary storage is growing faster, battery production from the largest manufacturers is still primarily EV-oriented in 2025 (1,264 GWh), or 74%. EV battery production is expected to grow 20%+ in 2026E.

With the war in the Middle East, I expect a resurgence in EV adoption. A few players like BYD may be worth a look and I recommend checking out ’s thesis linked below.

Coughlin Cap

[Update] BYD (BYDDY)

Read more

5 months ago · 5 likes · Brian Coughlin

System integrators occupy a critical role between cell manufacturers and end customers. They design the overall system architecture, including thermal management and power conversion. This segment is expected to capture 25% to 30% of the total energy storage profit pool (McKinsey). Examples include Tesla and Sunrun.

Downstream - Software, EMS, and SaaS: The industry is seeing a shift toward SaaS models, which carry higher gross margins. Energy Management Systems (EMS) control system activity, including grid integration, state-of-health monitoring, and market optimization.

Global Energy Storage Capacity - Present and Future

Energy storage installations have reached an inflection point, driven by a convergence of declining hardware costs and aggressive policy mandates. Annual deployments grew 43% Y/Y to ~99 GW, lifting total global commissioned capacity to ~270 GW / 630 GWh by the close of 2025. Looking toward 2030, market projections indicate a sixfold to eightfold increase in total capacity. The International Energy Agency (IEA) Net Zero pathway suggests global storage must reach 1,500 GW by 2030 to support the tripling of renewable energy capacity pledged at COP28. In this scenario, batteries account for 90% of the storage increase, rising 14-fold to 1,200 GW.

China is expected to continue as the global volume leader due to aggressive policy mandates and a well-established renewable energy system. China alone accounted for 54% of energy storage installations in 2025. In second place is the U.S., where installations grew +53% in 2025 despite significant policy uncertainty and trade friction. California, Texas, and Arizona continue to lead U.S. deployments, together accounting for 74% of domestic capacity in 2025.

Demand Drivers

A good rule of thumb in thematic investing is to focus on sectors where TAM is expanding rapidly. Energy storage is one of the clearest examples of this.

The Grid

This is the largest driver and the statement below from JPMorgan is all you need to see. The current conflict only adds urgency around energy security in a grid that already needs a significant rebuild.

The U.S. power grid is a creaky relic valued at close to $2T, with most core infrastructure dating back 50 to 75 years. It was not designed for two-way power flow or extreme load volatility that we see today. ~70% of U.S. transmission lines and power transformers are more than 25 years old. With a typical design life of 40 to 50 years, a large portion of the system approaches structural risk post-2030.

To address these issues and integrate new capacity, the U.S. power sector requires $1.4T in capital investment from 2025 to 2030. Globally, annual grid investment must nearly double to over $600B by 2030 to support renewable buildout and rising AI-driven demand.

Renewable Penetration and Transport Electrification

As solar and wind take a larger share of the generation mix, the grid faces the “duck curve” problem: excess supply during peak production hours and a sharp ramp in demand when production falls. In markets like Spain, solar margins fell 38% below the returns needed to build new assets in 2025 due to midday saturation. Storage addresses this by capturing low-value energy and discharging during higher-value evening peaks. ~1/4 of all utility-scale solar plants commissioned in 2025 included co-located storage, with this segment growing at a 70% CAGR since 2020.

Transport electrification adds another layer of demand. Electricity consumption from EVs is expected to grow 2.4x between 2025 and 2030, adding ~20 GW of demand to the grid by 2030. Storage allows charging operators to take advantage of off-peak rates, often 50–70% cheaper, and supply charging sessions from stored energy. This peak shaving capability can reduce operating costs by 40–60% and help avoid long wait times for substation upgrades.

AI Data Center Infrastructure

The fastest-growing driver is data center demand. AI workloads require significantly more power than traditional cloud computing. A single ChatGPT query consumes ~2.9 Wh of electricity versus ~0.3 Wh for a standard Google search. Power demand is also more volatile, with rapid spikes that create voltage instability.

AI data centers require stable, highly reliable power that the grid alone cannot consistently provide. Hyperscalers are increasingly deploying Battery Energy Storage Systems (BESS) not just for backup, but to manage load variability and stabilize power delivery.

Within this market, a new segment is emerging: distributed Battery Backup Units (BBUs). These systems are installed directly in server racks to manage localized power fluctuations. Panasonic Energy, with ~80% share in distributed power supply, has effectively sold out production through fiscal 2029, with hyperscale customers pre-committing to over 80% of planned output to secure capacity for AI clusters.

Total Addressable Market (TAM) Projections through 2030

Utility-scale TAM: The primary growth engine, representing 82% of all 2025 installations. This segment is expected to grow at a 29% CAGR through the end of the decade. Annual deployments are expected to consistently exceed 150 GW post-2026, driven by renewable integration and capacity reserve requirements.

Commercial & Industrial (C&I) TAM: With increasing adoption in manufacturing and high-growth AI compute, the global C&I market is forecast to reach $60B to $100B by 2030.

Residential and Data Center TAMs: The residential market is expanding steadily, particularly in Europe and Australia, where high electricity prices and grid instability drive self-consumption. The residential TAM is projected to grow from $18.5B in 2025 to $35.8B by 2030.

The AI data center power management market, specifically BESS and UPS solutions, is estimated at $7B in 2025. As AI power consumption doubles to 980 TWh by 2030, demand for stable storage solutions will accelerate, making BESS as critical to data centers as server racks.

Unit Economics - Costs and Revenue Streams

Battery prices reset to the downside in 2025. Stationary storage pack prices fell to ~$70/kWh, down ~45% Y/Y, driven by Chinese overcapacity. Stationary cell production reached ~557 GWh, more than 2x global installations.

In utility-scale systems, cells account for >50% of hardware cost, making project economics highly sensitive to lithium and graphite. Distributed systems skew more toward enclosure and integration costs due to siting and system complexity.

The revenue model is shifting from subsidy-driven to market-driven, with a growing mix of merchant exposure.

Primary revenue streams

• Wholesale arbitrage: buy low, sell high; expected to be >60% of revenue in mature markets by 2030

• Ancillary services: frequency regulation, reserves; high margin but limited and saturates quickly

• Capacity payments: payments for availability during grid stress; expanding across regions

• Avoided transmission capex: defers upgrades to substations and transmission infrastructure

Pure merchant projects can generate mid-teens returns but carry higher volatility and financing costs. Contracted structures offer more stable cash flows with a lower WACC (project discount rate).

U.S. Regulatory Landscape

Concerns around a removal of storage tax credits are misplaced. Grid upgrades and data center power demand are accelerating, which supports continued policy backing. The OBBBA tightened compliance and sourcing requirements but left the core incentive structure intact.

Energy storage is positioned well within the current framework. Standalone storage retains access to Section 48E on the original timeline, while wind and solar face accelerated cutoffs tied to construction after July 4, 2026 or in-service dates beyond December 31, 2027. Section 45X credits for U.S.-made battery components hold at full value through 2029 and phase out by 2032. The Section 25D residential credit ended on December 31, 2025, shifting homeowners toward third-party ownership structures such as leases and PPAs that rely on commercial credits.

Policy is now centered on reshoring, with Prohibited Foreign Entity rules introducing strict eligibility thresholds. PFE status is triggered at ≥25% ownership by a specified foreign entity or ≥40% in aggregate, and also if ≥15% of debt is held by violating parties. These rules are already changing trade flows. China’s share of U.S. battery imports fell from 69% in 2024 to 40% in 2025. Domestic assembly capacity reached 69.4 GWh by end-2025, while stationary cell capacity is projected at 96 GWh in 2026. Supply is still constrained. BESS hardware costs have risen 56% to 69% since January 2025 due to tariffs and limited domestic cell availability. Early U.S. output is largely committed to hyperscale buyers, leaving smaller developers choosing between higher-cost compliant systems or lower-cost Chinese supply that forfeits a 30% to 40% tax credit.

The winners

All of the above should tell you that TAM expansion is inevitable and the puck is skating toward energy storage.

Now to the key question, who benefits most from this expansion?

Here are 5 that I think are well positioned to capture that growth…

FLNC - Fluence Energy, Inc.

Price: $13.30 | Market Cap: $2.45B | FY26E Rev Growth: 48.3% | FY27E Rev Growth: 20.4% | EV/FY27E Sales: 0.6x | EV/FY27E EBITDA: 19.7x | FY27E EBITDA Margin: 3.1% | YTD return: -33%

FLNC offers direct exposure to utility-scale battery storage through a model centered on system integration. The company generates ~96% of revenue from selling and deploying BESS systems, with a smaller contribution from services and software tied to its installed base. It has deployed ~6.8 GW and holds ~9.1 GW of backlog across 25+ countries.

The thesis is driven by strong end-market demand across grid reliability, renewable penetration, and emerging data center load. FLNC’s integrated platform and global footprint support its ability to win and execute large projects, but the industry is seeing declining ASPs as supply expands. EV-to-storage capacity conversions are increasing available cell supply over the next 12–24 months, bringing in additional competitors and putting pressure on pricing. Management is targeting 30%+ revenue growth, which requires continued share gains alongside a meaningful ramp in deployed MWh, raising the importance of consistent execution in a more competitive environment.

The recent selloff followed a weak margin print but did not change the underlying demand picture. 1Q26 revenue was $475M, up 154% Y/Y, with FY26 guidance reaffirmed at $3.2B–$3.6B of revenue and $40M–$60M of adjusted EBITDA. The company signed >$750M of orders in the quarter and exited with a record ~$5.5B backlog, which now fully covers the midpoint of FY26 revenue guidance. U.S. demand is accelerating, with >$500M of quarterly orders coming from the U.S., and total pipeline increased by ~$7B, or ~30%, driven primarily by domestic activity.

Data centers are an incremental driver that is not yet reflected in backlog or revenue. FLNC is engaged on ~36 GWh of data center opportunities, with only ~25% of ~34–36 GWh currently included in pipeline and none converted to backlog yet. Management expects initial conversions in late 2026, and many projects are outside the formal pipeline, which suggests additional upside beyond disclosed figures. The FLNC data center opportunity is still early, but the scale is material relative to current deployments.

AMSC - American Superconductor Corporation.

Price: $32.09 | Market Cap: $1.5B | FY26E Rev Growth: 32.4% | FY27E Rev Growth: 21.4% | EV/FY27E Sales: 3.9x | EV/FY27E EBITDA: 41.2x | FY27E EBITDA Margin: 9.4% | YTD return: +12%

AMSC supplies power equipment that helps customers move, stabilize, and protect high-power loads. The business is now heavily skewed to Grid, which was 85% of 3Q25 revenue, with the balance from Wind at 15%. It generates revenue from selling power control systems, grid interconnection solutions, and resiliency equipment into utilities, traditional energy, renewables, industrial customers, semis, and naval applications.

The core thesis is tied to rising demand for grid stability and power quality as loads become more volatile. 3Q25 revenue grew 21.4% Y/Y to $74.5M, gross margin reached 31%, and the company delivered its sixth straight profitable quarter. The 12-month backlog is now >$250M, and management guided 4Q25 revenue to >$80M, setting up another record quarter. Demand is broad-based, with traditional energy near one-third of shipments, renewables ~1/4, military and utility each above 15%, and materials including semis above 10%, which reflects exposure to multiple areas of incremental power demand.

The acquisition of Comtrafo adds large power and distribution transformer manufacturing for utilities and industrial customers across Brazil and Latin America, expanding AMSC’s footprint deeper into grid infrastructure. The business is expected to generate ~$55M of revenue in 2025 with ~$85M of backlog and >20% operating margins, and research points to ~20% revenue uplift from the deal with 25%+ growth over the next few years. This comes at a time when transformer supply is tight and grid capex is accelerating, which should support both volume and pricing.

There is also an emerging data center angle. AMSC delivered its first data center order in 3Q25, contributing ~5% of revenue or ~$4M, and this is expected to build gradually as hyperscale demand for power quality and interconnection increases. The opportunity is still early but adds to an already diversified demand base across semis, energy, and utilities. Don’t let the FY27E EBITDA multiple fool you, consensus estimates are still moving higher as the impact of Comtrafo, continued Grid growth, and incremental data center revenue flows through their models.

GNRC - Generac Holdings Inc.

Price: $194.09 | Market Cap: $11.4B | FY26E Rev Growth: 15.1% | EV/FY26E Sales: 2.6x | EV/FY26E EBITDA: 13.7x | FY26E EBITDA Margin: 18.6% | YTD return: +42%

GNRC makes generators and related power equipment for homes, businesses, and data centers. The business is reported in 2 segments, Residential and C&I. Residential includes home standby and portable generators and home energy products, while C&I includes larger stationary generators, switchgear, controls, and storage sold into commercial, industrial, and data center applications. In 2025, the mix was ~59% Residential and ~41% C&I.

The thesis is that GNRC is becoming less dependent on storm-driven residential demand and more leveraged to structural power demand growth. 4Q25 was the clearest example. Revenue fell 12% Y/Y to $1.1B because home standby and portable generator demand was weak in one of the softest outage environments in a decade, yet C&I product sales still grew 10% Y/Y on data center shipments. Management guided 2026 revenue growth in the mid-teens with EBITDA margins of 18% to 19%, including ~10% Residential growth and ~30% C&I growth. Commercial growth is becoming the main driver of earnings growth.

Data center backlog grew from $150M at 2Q to ~$400M at year-end and then to roughly $700M by investor day, with ~$300M set for 2026 delivery and the balance building into 2027 and 2028. The company also disclosed a non-binding notice to proceed from one hyperscaler that could become a $600M+ order in 2027, and management’s 2028 framework includes ~$1B of data center revenue with only partial hyperscaler contribution. Capacity is being built to support that ramp, with domestic large-generator capacity expected to reach roughly $1B to $1.2B by late 2026 and product scaling moving toward 3.5 MW to 4.25 MW units.

Shares sold off after investor day because investors wanted a signed hyperscaler contract and did not get one. The stock fell as much as 12% intraday even though the update was constructive. Backlog rose sharply, the 2028 framework called for $6.2B to $6.6B of revenue and $1.25B to $1.45B of EBITDA, and the company added more vertical integration through the Enercon acquisition, which strengthens its enclosure and switchgear offering and is expected to help margins.

NRGV - Energy Vault Holdings, Inc.

Price: $3.20 | Market Cap: $553.5M | FY26E Rev Growth: 26.3% | EV/FY26E Sales: 2x | EV/FY28E EBITDA: 16.6x | FY28E EBITDA Margin: 10% | YTD return: -31%

NRGV develops and deploys grid-scale energy storage systems for utilities, IPPs, and large industrial customers. Historically, most revenue came from battery storage projects built for third parties, but the company is now shifting toward owning and operating projects through Asset Vault, which should move the model from lumpy EPC-style revenue toward recurring tolling and power sales over time. Since 2022, the company says it has deployed 1.4 GWh of storage, generated >$550M of revenue, and >$90M of gross profit, with lithium-ion systems still driving most of the business.

2025 revenue reached $203.7M, up 341% Y/Y, Q4 revenue was $153.3M, gross margin improved to 23.6% for the full year from 13.4%, and Q4 adjusted EBITDA turned positive at $9.8M, but 2026 guidance is still wide at $225M-$300M and the bigger margin step-up is really in FY27 and beyond tied to Asset Vault scaling. In other words, the multiple here is more likely to rerate on revenue upside and backlog conversion than on near-term margin expansion. Management highlighted 540 contracted megawatts and said that alone supports $130M-$150M of delivery over the next 18-36 months, while Asset Vault still targets $100M-$150M of recurring EBITDA within 4 years across 1.5 GW.

Backlog ended 2025 at $1.3B, up from $0.9B at Q3, and the platform now includes five projects totaling 441 MW / 2.6 GWh after the 100 MW / 870 MWh New South Wales project announced in February. The company has also expanded into AI-related infrastructure through its sodium-ion partnership and 1.5 GWh supply agreement, and in late March it acquired the 175 MW McMurtre BESS project in ERCOT, bringing owned assets acquired, under construction, and in operation to 715 MW. The Texas project is expected to generate $15M-$20M of annual revenue over its life, or roughly $350M-$375M total, which helps explain why insiders have stepped in to buy the low-$3s. CEO bought 12,500 shares at ~$3.21 on March 31, the COO bought 12,340 shares at ~$3.08 on March 2, and a director bought 5,000 shares at ~$3.65 on March 25. The timing lines up with a sharp post-convert selloff, improving backlog visibility, and new owned-asset additions that strengthen the Asset Vault case.

This is the riskiest name on the watchlist. The stock is a microcap that comes with dilution risk. In February, the company upsized and priced $140M of 5.25% convertible senior notes due 2031, which hit the stock hard and came on top of an existing $25M equity purchase agreement disclosed in 2025. Management argues the balance sheet is now in better shape, with $103.4M of cash at year-end 2025, the $300M preferred equity structure for Asset Vault, and the convertible helping retire higher-cost debt, but the market is right to demand proof before paying up.

SHLS - Shoals Technologies Group, Inc.

Price: $6.82 | Market Cap: $1.1B | FY26E Rev Growth: 23% | EV/FY26E Sales: 2.2x | EV/FY26E EBITDA: 11.1x | FY26E EBITDA Margin: 19.6% | YTD return: -20%

SHLS provides electrical balance-of-system components used in utility-scale solar and increasingly battery storage projects. The products handle power aggregation, protection, and wiring across solar arrays, inverters, and storage systems, with a focus on lowering installation costs and improving reliability for developers. The business has historically been tied to large solar projects in the U.S., but expansion into BESS, OEM, CC&I, and international markets is starting to contribute more meaningfully.

Q4 2025 revenue reached $148.3M, up 38.6% Y/Y, supported by higher domestic project volume and early contributions from newer channels. The company added ~$175M of new orders in the quarter, bringing backlog and awarded orders to a record $747.6M, up 18% Y/Y. ~$603.4M of this backlog is expected to convert within the next four quarters, providing good visibility into 2026. Full-year guidance calls for $560M-$600M of revenue and $110M-$130M of adjusted EBITDA, which implies ~22% Y/Y growth at the midpoint.

Margins are the main issue. Q4 gross margin declined to 31.6% from 37% in the prior quarter and from 37.6% a year ago, while adjusted EBITDA margin fell to 20.4% from 24.7% Y/Y. Management quantified several headwinds: $2.1M of tariffs and logistics costs, $2.5M of incremental labor tied to new product rollouts, and $0.5M of added plant overhead. Pricing pressure and mix also weighed on profitability. More importantly, management reset expectations by removing the prior 40% gross margin target and guiding to a low-to-mid 30% range for the near term, with improvement pushed to 2027.

International revenue increased from less than $1M in 2024 to ~$13M in 2025, and international backlog reached $90M. The OEM segment grew 47% for the year, and CC&I demand continues to build. BESS is the most important incremental driver. Backlog and awarded orders for BESS reached $67M at year-end, up from $18M in Q3, with more than half expected to convert in 2026. Management indicated that current BESS bookings are driven by data center backup power and high-power applications, where reliability and speed to interconnection are critical.

Recent forecasts suggest U.S. utility-scale solar installations through 2030 could exceed 250 GW, well above prior expectations of 150-200 GW, driven by grid expansion and rising power demand from data centers and electrification. This level of deployment supports continued demand for balance-of-system suppliers even with pricing pressure in core products.

The stock reflects margin concerns more than revenue trends. Weak direct current bookings, pricing pressure, and the removal of the higher margin guide have driven multiple compression. At ~2.2x FY26E sales and ~11.1x FY26E EBITDA, SHLS now trades closer to industrial peers rather than higher-multiple clean energy names. Upside from here depends on stabilizing margins, converting backlog into revenue, and scaling BESS without further cost overruns. If margins improve as temporary costs roll off and newer segments scale, earnings can expand from 2027 onward while revenue growth remains intact.

That’s all for this week’s screen! Of the names listed above, I think AMSC, SHLS, and FLNC are the most attractive/suitable for starter positions (in that order). Feel free to let me know what theme you’d like me to cover next.

I have a pipeline of ideas I am sifting through for my next fat long pitch. A few include: UAMY, HNRG, SYM, SANM, NOK, and BWMN.

This pitch is geared towards the investor seeking exposure in a cheap, under-covered micro-cap (<$100M market cap) that serves one of the biggest bottlenecks facing AI and electrification: the widening Grid Gap.

Teaser

Entered the power-constrained data center market; catalyst for growth and a re-rating

New international deals signed; likely to ramp quicker due to the ME conflict

Long-awaited new products launching in 4Q25; expected to contribute to revenues in FY26

0.5x FY26E sales for accelerating revenue growth to 20%+

Aligned insiders (CEO is the largest shareholder)

Healthy cash balance and shares trade near tangible book value; good margin of safety

A surprise to consensus FY26 revenue expectations could drive 1.5-2x upside, any signs of margin consistency has the potential to drive multi-bagger returns

PPSI - Pioneer Power Solutions, Inc.

NASDAQ: PPSI | 03/20/2026

Thesis Summary

PPSI is a name that may bring back bitter memories for retail investors. The company has spent most of the past decade restructuring, with a history of margin inconsistency, recently compounded by concerns around future revenue growth and a perceived legislative overhang that is overstated.

The stock attracted significant retail attention in late 2024 following the divestiture of its slow-growing segment and the subsequent special dividend. Since then, it has largely been forgotten.

In the P14 view, now is the time for PPSI to shine. PPSI is scaling its e-Boost mobile EV charging platform across the U.S., with early traction in the Middle East and expanding into distributed power. The company is directly monetizing the grid gap, where mobile and off-grid solutions are required as utility upgrades lag EV adoption, with growing validation from large fleet operators and mission-critical end markets.

In 2026, PPSI begins generating commercial revenue from its new private power platform, PowerCore (formerly HOMe-Boost), targeting residential and light commercial markets. This meaningfully expands TAM and is positioned to be a key incremental driver of growth.

The launch of PRYMUS further expands TAM into distributed power and high-performance compute, addressing demand from data centers and AI workloads. Early traction would be meaningful given PPSI’s existing scale.

Despite these drivers, consensus implies only ~10% Y/Y revenue growth in 2026, which appears conservative given backlog, continued e-Boost momentum, PowerCore contributions, and optionality from PRYMUS. Margins have been volatile, but with improving mix, standardization, and operating leverage, margins should recover through 2026 with a path toward GAAP profitability by 2027.

Revenue growth drives valuation in this micro-cap. At ~0.6x FY26E sales and trading slightly above book with a healthy cash balance, a beat on FY26 revenue expectations in the upcoming 4Q25 print (<2 weeks) can drive 1.5–2x upside in < 1 year, consistent with prior earnings reactions. Any signs of sustained margin stability would support multi-bagger potential.

Business Overview

Founded in 1995 following the acquisition of a transformer business from Schneider Electric, Pioneer Power Solutions (PPSI) spent decades as a consolidator of electrical infrastructure manufacturing. In 2019, management initiated a major simplification strategy to pivot from low-margin heavy manufacturing to higher-growth, asset-light clean technology. This transition culminated in 2024 with the $50M sale of its Electrical Infrastructure (PCEP) segment to Mill Point Capital, leaving PPSI as a pure-play provider of mobile and distributed power solutions.

Today, PPSI operates through its sole Critical Power Solutions segment and generates revenue from a dual model of equipment sales/rentals and recurring services. Product revenue was $20.2M, +172% Y/Y in the LTM, and includes designing and assembling mobile EV charging and distributed power units. Revenue is generated through direct sales, short-term rentals, and long-term sales-type leases, with most equipment sold under project-specific contracts and delivered within defined timelines. Service revenue was $11.5M in the LTM, +38% Y/Y, and includes 24/7 maintenance, repairs, and remote monitoring across the installed base. Contracts typically range from one to five years, providing a recurring revenue floor. Revenue is seasonally weighted to H2, driven by municipal budget cycles and delivery schedules, with Q1 typically the lightest.

Following the 2023 pivot to an asset-light model, PPSI is scaling without heavy CapEx by leveraging contract manufacturing partnerships. This enables the company to focus on design, engineering, and high-value testing rather than fabrication.

The product portfolio is expanding to meet rising demand across end-markets. At the core is e-Boost, launched in late 2021 as a self-contained, high-capacity mobile EV charging solution. It addresses the “grid gap,” the lag between EV adoption and utility infrastructure buildout. e-Boost provides off-grid, immediately deployable charging, allowing fleets, municipalities, and enterprises to electrify without multi-year delays tied to grid upgrades or permitting. The platform has scaled from a single prototype in 2021 to 60+ deployed units by late 2025. Revenue from e-Boost grew from $0 to $22M in ~30 months. Cumulative charging sessions increased from 350 in 2022 to 60,000+ by late 2025, with over 10 GWh of off-grid energy delivered.

The initial use case for e-Boost was high-speed (Level 3 DCFC) charging for heavy-duty EV fleets. Key markets include electric school buses, where districts have already purchased vehicles but lack sufficient grid capacity for charging. Almost all wins from e-boost end-markets cite the same reason: utility infrastructure delays. These districts and companies already have the electric vehicles, but the local power company tells them it will take 2 to 3 years to upgrade the transformers. e-Boost allows them to start driving those EVs next month instead of in 2029. Additional markets include transit and municipal fleets powering buses, street sweepers, and refuse trucks in cities such as Portland and Long Beach. Adoption is expanding into last-mile logistics and emerging use cases such as robotaxis.

Since 2024, PPSI has expanded beyond EV charging into broader distributed power and resiliency markets, where charging is secondary to delivering mission-critical power. The company launched e-Boost Pure Power in 2024 to support large battery storage systems and hydrogen infrastructure by supplying off-grid power for rechargers and reformers. It has secured multiple non-EV orders since.

Recently, PPSI has also launched two natural extensions of the e-Boost platform. 1) PowerCore is a high-performance integrated system providing full backup power and EV charging for residential and light commercial users. It runs on a prime natural gas engine capable of 24/7 operation, unlike traditional backup generators with limited runtime. Target customers include high-income residential (>3,000 sq. ft.) and light commercial applications such as MRI centers, dialysis clinics, and small retail requiring uninterrupted power. Pricing ranges from $45K to $65K per unit. 2) PRYMUS is a mobile power platform delivering pre-engineered 1–10 MW energy blocks using a hybrid architecture combining multi-fuel generators with mobile battery storage and advanced switchgear. It targets AI edge compute and modular data centers, enabling deployment in ~6 months versus 2–3 years for grid upgrades. PRYMUS manages load variability and power spikes from AI workloads, including next-generation NVIDIA systems, providing stable behind-the-meter power for testing and scaling. Both PowerCore and PRYMUS launched in December 2025.

PPSI has pivoted its business model to capitalize on the "Grid Gap," a structural imbalance where the rapid adoption of EVs and the massive power requirements of AI compute have outpaced the ability of traditional utilities to upgrade infrastructure. Management expects this imbalance to persist for 5–10 years due to permitting and construction timelines that often exceed 2 years. The addressable market is supported by strong secular tailwinds across energy and data infrastructure.

The global EV market is projected to reach ~$803B by 2027. Charging infrastructure investment is expected to exceed $1T by 2040, with a need for ~120M chargers by 2030. The modular data center market, a key target for PRYMUS, is expected to grow at a ~17% CAGR to $164B by 2035. Global data center power demand is projected to increase 165% by 2030, creating a large opportunity for PPSI’s private grid solutions.

Given the EV focus, customer purchasing power is supported by federal and state incentives. Programs such as NEVI and the IRA provide grants and tax credits for fleet electrification and charging infrastructure. Currently, ~80% of PPSI’s orders originate from the West Coast, where policy support and funding are strongest. Management is expanding into the East Coast and international markets.

The often discussed bear case centers on potential removal of NEVI funding. However, much of the capital has already been deployed, and core customers such as municipalities and transit agencies have already committed to electrification. These entities have invested heavily in EV fleets, making charging infrastructure a requirement rather than a discretionary decision. In January 2025, the administration attempted to pull the plug on NEVI funding, which was quickly challenged by lawsuits from ~20 states.

Despite strong revenue growth, PPSI has faced margin volatility following the PCEP divestiture and has yet to achieve GAAP profitability. Gross margins declined to 9.29% in 3Q25. However, I argue that this is likely to turn positively (see later).

CEO Nathan Mazurek owns ~30% of the company, aligning incentives with shareholders, though capital allocation has drawn scrutiny. In January 2025, following the PCEP sale, PPSI paid a $1.50/share special dividend (~$16.7M total), which some viewed as tax-inefficient relative to buybacks or reinvestment. The company recorded a $35M gain on the PCEP sale but also dealt with prior restatements and material weaknesses tied to revenue recognition for over-time contracts in 2022–2023, which have since been resolved. PPSI is well-capitalized with $17.3M of cash and zero bank debt as of 3Q25, providing runway to scale PRYMUS and PowerCore.

The top-line thesis

The revenue thesis for PPSI centers on execution across a growing EV TAM by scaling e-Boost, success with PowerCore as the largest incremental driver in FY26, and expansion into modular data centers through PRYMUS.

Pillar 1: Capitalizing on the Grid Gap via e-Boost Execution

The core of the e-Boost thesis is the imbalance between ~24% CAGR EV sales and charging infrastructure growth requiring ~$1T of investment by 2040. PPSI is positioned as an early mover in mobile, off-grid charging, delivering solutions far quicker than utility upgrades. Beyond EVs, distributed power and adjacent verticals are gaining traction. Power availability is constrained and rising data center demand is driving electricity costs higher.

The landmark EV charging order was the $7.1M LAUSD contract for 25 units in 2Q24, completed by 3Q25. Another key win is the $10M multi-year SparkCharge award in 2Q25, expected to contribute ~$2M of revenue in 2025.

A strong indicator of momentum is the successful 2025 pilot with the world’s largest online retailer for last-mile delivery, likely Amazon. PPSI has been running this pilot since late 2024, now deemed successful. Before we get into the specifics, I want you to stop and think what this means for a $35M market cap company. This is massive market validation from the largest customer PPSI could possibly have.

The pilot involved short-term rentals during peak 2024–2025 holiday demand. This transitions into a direct purchase model in 2026, with expected orders of 5–20 units. The rented unit was a 240 kW e-Boost Mobile Max with Level 3 charging and dual Level 2 chargers. A comparable 250 kW unit sold to Long Beach, CA was priced at ~$725K/unit. Assuming ~$700K/unit, this implies ~$3.5M–$14M incremental revenue in 2026.

Momentum continues into 2026 with ~$1.6M of EV and electric bus orders through February. New verticals are emerging, including a recent win with Canada’s largest airport operator.

PPSI differentiates itself through mobility. A common bear case focuses on stationary charging growth, but PPSI offers both mobile and stationary solutions. The e-Boost pod functions as a stationary unit with optional mobility. Mobile solutions still dominate, accounting for ~80% of units sold.

Distributed power is emerging as another large opportunity. 3Q25 deliveries included a shopping center, condominium tower, and waste processing facility. A recent $700K order from a major U.S. fitness chain provides a 1 MW system for peak shaving (reduce electricity costs during high-demand hours) and “Island Mode” (keeping the gym running during a blackout). This is part of a potential rollout across 2,700+ locations, implying a ~$1.9B TAM from a single customer.

PPSI’s E-Bloc technology is seeing a surge in industrial sectors. The company recently secured orders to provide main power switchgear for one of the world’s largest titanium smelters building a new plant in West Virginia. One of the nation’s largest gas providers in the Western U.S. ordered multiple IPC (Electrical Buildings) to manage their own internal power distribution.

PPSI is winning large contracts from industries that cannot afford a single minute of downtime and grid independence is only rising.

Pillar 2: PowerCore as the Incremental Growth Engine for FY26

While e-Boost is primarily a project-driven business, PowerCore (formerly HOMe-Boost) represents a shift toward a standardized, higher-margin product that significantly expands PPSI’s TAM beyond fleet operators into residential and light commercial markets.

The product was initially targeted for an early 2025 launch but was delayed due to engineering refinements led by management, with a deliberate decision to remain in stealth mode until the final design met performance expectations. The product was officially launched in late 2025, with no contribution included in FY25 guidance, positioning it as a key driver of growth in 2026 and beyond. Manufacturing has been outsourced to a dedicated partner in Minnesota, allowing PPSI to scale production without incremental CapEx.

The rebranding to PowerCore aligns with the Distributed Energy Generation market, which is estimated at $432.4B in 2026 and expected to grow at ~13% CAGR through 2030. The commercial segment is projected to account for ~41% of the off-grid hybrid market, where uptime and island mode capability are essential.

PowerCore differentiates itself from traditional standby generators by utilizing a prime-rated natural gas engine capable of continuous 24/7 operation, while also integrating high-speed EV charging into a single system. This allows users to maintain full operational uptime while also supporting EV adoption without relying on grid infrastructure.

Initial feedback has been strong, with demand emerging from medical facilities such as MRI, dialysis, and cataract centers, as well as commercial bakeries and cement producers where even brief outages can disrupt operations. Existing EV customers are also potential cross-sell opportunities, as these users are evaluating PowerCore to maintain uptime while offering EV charging as an additional service to customers, particularly in retail and mixed-use environments.

Channel partners, including CleanSpark, generator dealers, and distributors, are targeting medical and small business segments that require more robust solutions than traditional backup systems.

For residential applications, PowerCore targets homes larger than 3,000 square feet, where customers are willing to spend $45K to $65K for high-end energy independence and reliability.

Pillar 3: PRYMUS and the Modular Data Center Frontier

The system can be deployed in ~6 months and is designed for modular data centers and containerized AI racks using next-generation NVIDIA chipsets. Units are pre-engineered in 1 MW to 10 MW blocks, allowing customers to scale power capacity alongside compute demand.

AI workloads exhibit significant load variability compared to traditional cloud infrastructure, and PRYMUS addresses this through a hybrid architecture that combines multi-fuel generators with mobile battery energy storage systems. This configuration allows the system to absorb rapid power spikes while maintaining stable output, which is critical for AI workloads. Next-generation chipsets such as NVIDIA Blackwell create highly volatile power profiles, with rapid transitions from idle to peak load that can stress traditional infrastructure and grid connections.

PRYMUS is not intended to power entire data centers but is designed for testing environments, where battery storage acts as a buffer for power spikes while generators provide steady baseline power. This makes it well-suited for semiconductor labs and OEM testing facilities that require isolated microgrids to validate performance without risking broader infrastructure.

The modular data center market is expected to exceed $164B by 2035, growing at ~17% CAGR, and PPSI’s ability to deliver modular 1–10 MW solutions within months addresses a key bottleneck in deployment timelines.

Supply of integrated mobile power solutions is limited, as incumbents such as Caterpillar and Cummins focus on large, permanent installations and lack the modular microgrid systems. PRYMUS is specifically designed for edge environments where grid capacity is constrained or unavailable.

While pricing has not been disclosed, comparable distributed power systems at ~1 MW have sold for ~$690K to $750K, suggesting PRYMUS installations will be multi-million dollar projects depending on configuration.

PRYMUS represents a meaningful incremental revenue driver beginning in 2026, and even modest adoption levels would have a material impact on PPSI’s growth profile given its current scale.

International expansion, the existing backlog, and revenue volatility

The lackluster +0.4% Y/Y equipment sales growth in 3Q25 reflects several timing-related factors rather than demand weakness. The comparison to 3Q24 was difficult due to a large one-time delivery that drove revenue growth of over +400% in that period, while approximately $2M of shipments were pulled forward into 2Q25, leaving a lighter delivery schedule in 3Q25. The completion of the LAUSD order also reduced incremental contribution, and diversification into distributed power deliveries were still in early stages at ~$700K.

Backlog growth has been uneven due to the company’s small size and reliance on large projects, where individual contracts can materially impact reported results. The mix has shifted from many smaller orders to fewer large contract, making the timing of deliveries more important to quarterly performance. As new verticals scale, backlog should become more consistent. As of 3Q25, backlog stands at ~$15M and typically converts to revenue within 12 months. I expect an increase to the backlog in 4Q25 given the expansion into new markets and e-Boost momentum. With the weak Y/Y compare vs 2025, this should optically bode well for Y/Y growth in 2026.

Perhaps one of the greatest achievements for e-boost validation is its first international MOU (ex-North America) with Savvy Charging Technologies in the UAE. Savvy is an established CaaS operator with strong relationships across utilities and government entities. The partnership includes a paid pilot of the e-Boost G.O.A.T. unit mounted on a Ford F-350 with a 40 kW Level 3 charger.

Following the pilot, the agreement outlines technology transfer and local manufacturing, with Savvy securing orders and producing units domestically. The pilot is expected to be delivered in 1Q26, with recurring revenue beginning in 2026 and scaling in 2027.

This expansion into the Middle East is significant, as the region is aggressively pursuing electrification targets, including 30% fleet electrification in the UAE by 2030. e-Boost provides a practical solution to bypass long grid upgrade timelines, which are a major bottleneck in the region. There is also an interesting perspective to take with the existing conflicts in the region (see later).

Revenue build

The top-line thesis translates into ~33% CAGR for equipment revenue and ~5.7% CAGR for services over the forecast period.

Management has guided to $27–29M in FY25 revenue, $28M at the midpoint or +22.4% Y/Y. I expect PPSI to meet this guide with equipment sales returning to strength in 4Q25. Looking ahead, P14 sees equipment sales as a 30% grower. PPSI’s TAM is expanding significantly and order momentum should continue to build into 2026. Services revenue is expected to grow at ~5% steady-state, as its lower-margin nature likely leads to incremental services being outsourced to third parties.

Consensus estimates for FY26 sit at just +10%, which in my view is very conservative given the existing backlog, the Amazon pilot translating into revenue, and diversification into distributed power with PowerCore, while e-Boost remains the core engine alongside upside from PRYMUS.

There are more paths to upside acceleration vs consensus than downside to a deceleration to ~10%. I expect an acceleration in total revenue growth from FY26E ~20% to FY27E ~21%, driven by these same factors and the persistence of electrification trends and the grid gap.

The margin thesis

Starting with gross margins, these have been volatile in recent quarters. The primary drivers are quarterly swings in equipment revenue and delivery timing, initial production costs on large contracts, strategic pricing on marquee projects, and execution variability.

Recent margin compression (e.g., 2.2% in 1Q25 and 9.3% in 3Q25) was largely driven by the 25-unit school district project. Initial units on large, complex projects carry higher costs as manufacturing processes are refined and workflows optimized. Management had also bid at highly competitive prices to secure this marquee customer and validate the technology at scale, which requires tight execution to avoid margin pressure. Reported margin weakness is also a function of mix, with a higher share of lower-margin projects or service revenue where subcontracting compresses profitability vs equipment sales. Execution issues on certain jobs, including the City of Portland project, also resulted in margins below targets.

For margins to recover, equipment deliveries need to drive a larger share of the mix relative to services. Management’s medium-term target is ~30% gross margins, supported by several factors.

Standardization and scale: The shift from one-off builds to standardized designs reduces labor intensity and improves unit economics as volume scales.

Shift toward higher-margin equipment: A greater mix of e-Boost and PowerCore drives margin expansion, as equipment carries higher margins than services.

New product lines: PowerCore is expected to have consistent pricing and improve margins with scale, while PRYMUS introduces high-ASP, MW-scale installations starting in 2026.

Expansion of recurring revenue: The international franchise model, including the UAE partnership, introduces high-margin licensing and revenue-sharing streams beginning in 2026. Leasing and Charging-as-a-Service are also higher return models with better asset efficiency than outright sales.

For 4Q25, management expects a rebound in gross margins as equipment becomes a larger share of revenue. Completion of the aggressively priced LAUSD order, incremental SparkCharge leasing revenue, and follow-on orders from the Amazon last-mile delivery program during the holiday season should support margins.

I model gross margins at ~13% for FY25, expanding to ~25% in FY26 and ~27% in FY27, trending toward the long-term target. This compares to consensus of 11%, 27%, and 29%, respectively.

At the operating level, I expect PPSI to reach GAAP operating profitability by FY27, driven by gross margin expansion and operating leverage across SG&A and R&D.

The Middle East Conflict and the PPSI Opportunity

PPSI is a net long-term beneficiary of the current conflict in the Gulf. Energy security is now a top priority, and the current oil shock is likely larger than COVID in magnitude. Oil remains the dominant input for global production, but across PPSI’s end markets, particularly EVs and distributed power, the second-order effects are constructive.

With oil prices spiking above $120/barrel in March 2026, the cost to operate gas-powered fleets has risen ~30% in a matter of weeks. The damage to regional supply suggests that even with a resolution, oil is likely to stay higher for longer.

For vehicle production, this reinforces the push toward electrification regardless of political backdrop. A renewed focus on EV production in North America is a direct tailwind for e-Boost, as infrastructure gaps persist.

On the grid side, higher energy prices and declining reliability increase urgency around electrification and grid upgrades. At the same time, these pressures accelerate demand for off-grid solutions, benefiting PowerCore and PPSI’s broader distributed power offerings.

The Savvy MOU comes at an opportune time. There is a high likelihood of faster-than-expected execution. Despite being major oil producers, Middle Eastern countries are facing localized disruptions from attacks on energy infrastructure, including drone strikes on refineries in Qatar and Saudi Arabia in March 2026. This has led to temporary fuel shortages in certain regions, which will take time to normalize.

In parallel, EV adoption in markets such as Dubai is accelerating, with Chinese EVs gaining share. This dynamic should drive increased government focus on electrification, creating a favorable backdrop for PPSI’s mobile charging and distributed power solutions.

Base Case Valuation

In my view, revenue growth drives the bulk of PPSI’s valuation. This is a small company operating across multiple rapidly expanding TAMs, and execution on top-line growth takes center stage.

At this stage in the business cycle, the market is less focused on the bottom line. This has been evident in prior earnings reactions. In 2Q25 (reported 08/14/25), revenue beat consensus by ~21% while margins compressed Y/Y but reaccelerated Q/Q, and the stock rallied as much as +83% to ~$5.7/sh before closing at $4.35 (+40%).

Given this setup, if indices remain stable and management guides comfortably above consensus FY26 revenue expectations of +10%, the stock is likely to see a sharp move higher, especially after the recent -42% drawdown from the 01/12 highs, which has not been driven by fundamentals.

Under the base case, assuming stable market conditions and a material upside surprise to FY26 guidance, I see 80%+ upside to $5.69/sh, or ~1.5x EV/FY26 base case sales, achievable within a year. If the stock approaches $5, I would look to exit, given the consistent failure to sustain momentum following prior earnings or guidance-driven rallies.

Longer term, if the company demonstrates gross margin stability and progress toward management’s LT targets, the stock could reach ~$8/sh, or ~7x EV/FY27 base case gross profit.

Bear Case and Risks

A prolonged Middle East conflict is a risk to equity multiples broadly. For PPSI, its small size adds volatility even if it is a long-term beneficiary. Higher oil prices also increase input costs, and the asset-light model could become a constraint if contract manufacturers face supply shortages.

Customer concentration risk is also material because of one-off large orders followed by recurring service revenues in later periods. In 3Q25, the top 2 customers made up 36% of revenue.

Execution risk is critical. Management must deliver on PowerCore without further delays and show progress on margins. I do not expect materially negative capital allocation decisions near term, as the CEO owns ~30% of the company and has meaningful personal exposure to the share price. He has publicly stated willingness to buy stock below $4, supported by recent purchases at $3.39/sh.

That said, I do not view PPSI as a long-term hold. The special dividend raised concerns around capital allocation, as it was tax-inefficient and disproportionately benefited insiders, which brings management integrity into question. At the same time, the CEO’s ownership likely keeps him focused on share price performance.

Dilution risk is present given the company’s size. Management indicated the PCEP sale helped avoid dilution, but if additional capital is required to support growth, an equity raise is a possibility.

Trading liquidity is another risk, so if you are an institutional investor, PPSI is not likely to fit your bill.

Competitive risk is present but manageable given the expanding TAM. Competition in the data center and distributed power market is intense, particularly for PRYMUS, although supply of integrated mobile solutions is limited.

In a bear case, I see the stock trading near tangible book value of ~$2.77/sh, implying ~-12% downside from current levels. The stock trades at 1.14x TBV at current levels.

Conclusion

I view the risk/reward as attractive at current levels and see PPSI as a compelling short-term (<1Y) long opportunity. The company is positioned for accelerating revenue growth across expanding markets, with a path toward margin stabilization. Initiating a position ahead of earnings (estimated 03/30) is attractive, although near-term downside risk tied to broader market conditions would certainly pressure the stock.

Disclaimer: Nothing posted by P14 Capital should be considered financial advice. The author of this post holds a long position in PPSI. Please consult a financial advisor and/or conduct your own due diligence before making investment decisions.

As is the case with many other recent P14 pitch earnings reactions, momentum quickly died and shares were sent reeling lower alongside a selloff in the indices. The selloff has worsened with the ongoing ME conflict (impacts addressed).

Today, Qnity trades at a lower price vs. the pitch date, but with materially better fundamentals, thus posing as another gift from Mr. Market at an attractive entry point.

Teaser

Largest pure-play electronic chemicals company trading at a steep discount to its peer despite offering higher growth and better margins

Meaningful and unique exposure to high growth areas of the AI buildout

Increasing AI inference on device is a structural tailwind for Q

Prime beneficiary of the China chip localization thesis

HSD/LDD total revenue growth with upside from high-growth areas, 30%+ EBITDA margins expanding through mix and automation

Risks exist with the ME conflict but Qnity is better positioned in the supply chain vs. peers

Speculative short-term catalyst for shares with GTC 2026 this week

More time in the public markets and beat-and-raises should overshadow the spin-off execution risk priced into shares, enabling a swift re-rate toward peer multiples in less than 2Y → 50%+ upside

Q - Qnity Electronics, Inc.

NYSE: Q | 03/13/2026

Since my initial pitch was relatively short, this earnings update will follow a slightly longer format. I will review both the thesis and how FY25/4Q25 results performed relative to it. Please read the original thesis for background.

Meaningful and unique exposure to the AI buildout

The core revenue generator for Qnity is its consumables-driven business model. ~90% of revenue is generated from unit-driven products either consumed during the manufacturing process (such as CMP slurries) or incorporated into the final electronic device (such as thermal materials and laminates). This is a recurring, volume-based business tied to customer utilization rates.

Qnity offers exposure to the AI buildout through two segments. First is the Semiconductor Technologies (ST) segment, which generated $2.64B in FY25 revenue, +7.8% Y/Y and 55.6% of total. This business provides high-purity materials for the front-end of the semiconductor value chain, specifically integrated circuit (IC) fabrication. In simple terms, ST supplies the materials used inside the chip factory (the fab) to build the silicon wafer.

High-level drivers here are increasing fab utilization and wafer starts. More specifically, ST is benefiting from the transition to 2nm and 3nm Gate-All-Around (GAA) architecture, the foundation of next-generation AI accelerators. AI chips require the most advanced nodes such as 3nm and 2nm logic and High Bandwidth Memory (HBM). These architectures require 3–5x more processing materials per wafer, particularly CMP pads, slurries, and additional lithography layers, compared with legacy nodes. This increases CMP intensity, meaning Qnity’s consumables wear out faster and drive higher volume sales.

Currently ~35% of the ST portfolio is tied to advanced nodes, expected to grow to 45–50% by 2030, driven by AI-led transitions in logic and HBM3/4. Q’s advanced logic and HBM business grew mid-teens in 2025.

All of this sounds good on paper, but within ST only ~20% of revenue is derived from the AI/HPC/Data Center end market. As a result, ST growth typically tracks MSI (million square inches) wafer starts, which management expects to grow MSD in 2026. This guide appears conservative because it does not incorporate the ongoing manufacturing and industrial recovery.

While ST may appear to be a drag on growth, its value lies in both AI exposure and its margin profile. The segment generates ~36% EBITDA margins. I urge you not to discount the diverse end-consumer exposure of the ST segment as this will be a tailwind for years to come (expanded below).

The Interconnect Solutions (ICS) segment provides AI exposure through the back-end of the semiconductor value chain, focusing on how complex chips are connected and how the heat they generate is managed. This segment is the primary driver of incremental revenue growth through its thermal and advanced packaging solutions. ICS grew revenue 12% in 2025, with EBITDA margins expanding 175 bps to 25.5%.

Three subsegments stand out and should continue to drive growth: advanced packaging, advanced interconnects, and thermal management. Each grew >20% in FY25. These businesses, particularly advanced packaging, which represents ~10% of total sales and growing, differentiate Qnity relative to peers.

Advanced packaging growth is driven by the increasing material intensity of AI chips. In traditional chips, such as those used in automotive or appliances, Qnity may only provide a simple laminate or single plating layer. In AI GPUs such as NVIDIA Blackwell or HBM stacks, the physical density of Qnity’s materials increases significantly. Thermal management growth is driven by the rising power consumption of AI clusters. Qnity’s materials improve heat transfer between the chip and the cooling system, supporting higher compute densities.

Increasing AI inference on device is a structural tailwind for Q

While AI/DC capex has been blowing past estimates, leaving investors scrambling for overlooked ideas across the supply chain, little focus is being given to the next major theme: the consumption of these applications. For massive investments in AI infrastructure to generate returns, AI must move beyond the data center and onto everyday devices. Increasing AI inference on-device is a structural tailwind for Qnity that should drive growth for years.

By 2030, analysts project that ~50% of enterprise AI inference workloads will be processed locally on endpoints rather than in the cloud. This shift creates a materials intensity tailwind across Qnity’s end markets, particularly premium consumer electronics and automotive.

According to Gartner (2026), AI PCs are expected to represent ~55% of the total PC market by the end of 2026, up from ~15% in 2024. These devices require a hardware content upshift. An AI-enabled laptop requires ≥16GB of RAM and dedicated NPUs (Neural Processing Units), increasing semiconductor process complexity. This drives ~3x more CMP (Chemical Mechanical Planarization) steps during wafer fabrication, directly boosting ST revenue per device.

On-device AI also increases localized heat generation beyond what traditional smartphone and laptop cooling solutions can manage. Qnity’s Laird brand is seeing rising demand for non-silicone thermal gap fillers, such as the recently launched 16 W/mK Tflex SF16, as premium device makers transition away from basic thermal grease toward engineered thermal materials.

Memory pricing is another dynamic affecting device design. DRAM costs are expected to rise ~130% by the end of 2026, pushing OEMs to concentrate product launches in premium AI tiers. Management notes Qnity’s consumer electronics exposure is concentrated in premium segments, so the expected volume impact from the memory shortage should be limited.

Prime beneficiary of the China chip localization thesis

While much of the market’s focus has been on the AI data center buildout in the West, Qnity is emerging as a key materials supplier tied to the China chip localization push. For Qnity, China is not a secondary market. It is a core growth engine that accounted for 33% of FY25 sales (~$1.57B) and grew +7.8% Y/Y despite broader macro headwinds.

Beyond China, Qnity is heavily exposed to APAC, where its largest customers operate, including TSMC, SK Hynix, and Samsung. The region represents ~79% of revenue and grew +9.6% Y/Y in FY25. All three companies are expanding semiconductor capacity. As these fabs come online over the next 1–3Y, Qnity should benefit through higher consumables demand, assuming utilization holds.

A core advantage is Qnity’s local-for-local operating model, which differentiates it from U.S. peers relying on centralized export supply chains. Manufacturing and R&D are located close to customers, improving service speed and supply reliability across APAC.

Turning back to China. Developments tied to the 15th Five-Year Plan make it clear that China is accelerating domestic semiconductor manufacturing. Export restrictions on advanced GPUs have pushed domestic alternatives from companies such as Huawei, reinforcing the push for local supply chains.

Memory is a major bottleneck in the AI buildout, and China is aggressively targeting this opportunity. A major tailwind for Qnity is the expansion of China’s domestic memory leaders, particularly ChangXin Memory Technologies (CXMT) and Yangtze Memory Technologies (YMTC).

CXMT is rapidly scaling capacity. In early 2026, the company announced plans to expand its Shanghai facility to roughly 2–3x the size of its Hefei base to capture rising global memory demand. The company is transitioning production from DDR4 to DDR5 and HBM (High Bandwidth Memory). YMTC is also expanding capacity. Higher wafer output directly increases demand for Qnity’s consumables.

Qnity’s local-for-local model fits this environment well. While the U.S. restricts exports of certain semiconductor equipment such as EUV lithography tools, Qnity supplies materials that fabs consume every day, including CMP slurries, pads, and cleaning chemicals. This also means that Qnity is shielded from China’s technology import restrictions and U.S. tariffs.

Recent investments highlight the company’s positioning in the region. Qnity committed $61.5M to acquire a research and manufacturing facility in Zhunan Science Park, Taiwan, indicating ramping demand across the APAC semiconductor ecosystem.

Capacity expansion to respond to pent-up demand

Qnity currently faces demand that exceeds available manufacturing capacity, particularly in high-growth areas such as advanced semiconductor nodes and advanced packaging. As a result, capex will rise to ~9% of sales in 2026, before normalizing to ~6% in later years.

More importantly, major capacity additions are beginning to come online. Compared with APAC, Qnity’s exposure to the Americas is modest at ~13% of revenue (+12.5% Y/Y). This should increase as semiconductor manufacturing continues to onshore to the U.S.

Earlier this month, Qnity opened a 385K sq ft facility in Delaware. The site is designed to scale production of CMP materials, supporting rising demand tied to AI and high-performance computing (HPC). This capacity should be accretive to revenue and margins as it targets higher-value materials and incorporates greater manufacturing automation.

Revenue build

The thesis above translates into the simple revenue build below. In both segments, volume is the core driver rather than price.

Management set a conservative FY26 guide after beating FY25 expectations. For FY26, Q guided to +4.54%–8.75% Y/Y revenue growth ($5.07B at the midpoint, +6.65% Y/Y). For Q1, they guided high-single-digit sequential revenue growth for the total company with margins similar to Q4. Momentum is currently led by AI demand, HPC, and advanced connectivity, with notable strength in ICS.

Revenue growth decelerating in 2026 appears unlikely. The thesis outlined above points to accelerating demand while new capacity is coming online. The guide also does not appear to include upside from CXMT and YMTC increasing memory production in China. Under my base case, I expect FY26 revenue growth of +9%, above the high end of the +8.75% guide. Looking ahead, I model +10.6% growth in 2027 and +7.9% in 2028.

Management said MSI (million square inches) wafer starts are a good indicator of overall demand and assumes MSI grows mid-single digits in FY26. The midpoint of guidance is anchored around MSI growth and the PCB market, both expected to grow mid-single digits, with Qnity expecting to outperform.

Base semiconductor demand should broadly follow MSI / wafer starts. Semiconductor revenue can grow faster due to content gains at advanced nodes, share gains, improved utilization, AI-driven demand, and higher materials intensity per wafer.

This dynamic played out in 2025. Semiconductor revenue grew +8% Y/Y, driven by +9% volume growth. Management attributed the increase to advanced-node content gains, share gains, and improved customer utilization.

For FY26, I model ST revenue growth of +5% Y/Y, roughly in line with MSI expectations. This assumption excludes any industrial or automotive recovery in 2026. In 2027, I expect +6% growth, supported by industrial recovery and rising on-device AI inference across Qnity’s end markets. For 2028, I model +4% growth.

ICS should be the faster-growing segment in FY26. Management said ICS should be stronger than Semiconductor during the year, and noted that Q1 demand strength is concentrated in ICS.

The key drivers are advanced packaging, advanced interconnects, and thermal management. Each grew >20% in FY25, and management expects strong demand in 2026 depending on how quickly customers bring incremental capacity online.

ICS growth is closely tied to HBM and memory capacity expansion. For FY26, I model +14% Y/Y growth, driven by HBM capacity expansions at Qnity’s largest customers. In 2027, I expect +16% growth as these facilities approach peak utilization.

Overall, I expect ~9% revenue CAGR over the next three fiscal years. Management’s 6–7% organic sales CAGR through 2028 is conservative to me and does not incorporate a cyclical industrial recovery, the acceleration in ICS, or the China memory capacity ramp.

Margins and FCF

Management has guided Adj. EBITDA margins of 29.48%–30.46% for FY26 (29.98% midpoint, +5.48% Y/Y). This compares with 30.31% in FY25.

Margin drivers

Commentary on gross margins has been limited. Expansion is driven by mix shift toward advanced nodes and higher-margin revenue sources such as Advanced Packaging. Capex investments should also generate economies of scale, while management continues to invest in facility automation that should improve productivity over time. I model gross margin expansion of +23 bps in FY26 and FY27, followed by +62 bps in FY28.

EBITDA margins should expand over time. FY25 saw +69 bps of expansion. Management has launched a multi-year transformation plan expected to deliver $100M of run-rate synergies by the end of 2028. The program includes upfront investment in automation upgrades and expansion of the local-for-local operating model, which should pressure margins slightly in FY26.

Qnity’s scale as a manufacturer provides fixed cost leverage, along with operating leverage across R&D and SG&A. Combined with the mix shift toward higher-margin revenue streams, I model GAAP EBITDA margins roughly flat in FY26, followed by +53 bps expansion in FY27 and +122 bps in FY28 as the transformation program matures.

Alongside margin expansion, management expects more favorable tax treatment over time (ENTG tax rate ~7.1% for reference). Management also authorized a $500M share buyback. I expect repurchases to begin in FY27–FY28, reducing share count by ~1% annually.

In my view, Qnity is capable of 20%+ GAAP EPS growth over the next 3-5 years.

FCF should be volatile in FY26 due to elevated capex tied to capacity expansion. In later years, I expect a strong recovery as capex normalizes and working capital stabilizes.

The Middle East Conflict

Like most players in the semiconductor supply chain, Qnity has exposure to disruptions in the Strait of Hormuz, particularly through noble gases (helium), specialty chemicals, and other energy-intensive inputs that could see short-term cost inflation.

I believe we are near the end of the strait disruption (I hope we are). If the closure were to extend >1 month, the impact would reach far beyond equity markets. Global supply chains would tighten and inflation would spread across basically everything.

In a prolonged disruption scenario, assuming AI capex from Qnity’s customers continues, the company has a $145M raw material buffer on the balance sheet. Any input cost increases would likely be passed through to customers.

Qnity’s local-for-local operating model and China presence could also provide flexibility. If disruptions persist, some supply flows could shift through Chinese facilities before reaching fabs and manufacturing sites across APAC. Energy-dependent countries such as Korea and Taiwan may also negotiate separate arrangements, similar to what has occurred with India.

If the conflict escalates, Qnity’s stock will likely sell off alongside global equities. My base case is that the disruption resolves within ~2 weeks.

Indirect catalyst with GTC 2026

As much as I’d love it, Qnity will likely not be mentioned by name during this week’s GTC conference. It is, however, a direct beneficiary of the likely increased wafer complexity being introduced in NVDA’s new products. Qnity should offer good thematic exposure this week.

Base Case

My original pitch argued that Qnity’s multiple should converge toward ENTG’s then ~25x EV/FY26E EBITDA multiple. Since the pitch, ENTG’s multiple has compressed and now trades at ~21x consensus FY26E EBITDA. Qnity currently trades at 16.9x my FY26E base case EBITDA.

If my base case plays out, I still believe Qnity should trade at ~25x EBITDA on FY26E numbers. Aside from the Middle East conflict, which could derail the thesis, management appears positioned to deliver beat-and-raise quarters, which should reduce the spin-off execution risk currently priced into the multiple. I see ~57% upside in under 2Y to ~$170/share.

This is a market-share leader with 30%+ margins, exposed to strong WFE demand and optionality from a recovering semiconductor market, with differentiated exposure to high-growth areas such as advanced packaging, unlike peer ENTG. I believe Q deserves to trade at a premium to peers and view current levels as an attractive entry point for investors willing to stomach the low-probability risk of a prolonged war.

My bear case is closer to the IPO price of ~$95/share.

In this pitch, P14 presents a Japanese semiconductor equipment component supplier that is at an inflection point.

Teaser

10x P/E for high-teens annual EPS growth over the next 2Y

1x EV/Sales for revenue growth accelerating into the teens

Serving a critical bottleneck in the AI buildout. Capacity is expanding (nearing completion) to respond to hungry demand

Solid capital allocation with a growing dividend that pays you while you wait

If you are interested in undercovered, cheap, and accelerating profitable growth in the AI semiconductor space, this one is for you.

255A - Gltechno Holdings, Inc.

TSE: 255A | 03/06/2026

Thesis summary

GLTechno (GLT) is a relatively newly formed small-cap that is a direct beneficiary of the AI capex buildout through its semiconductor segment, alongside rising demand from its analytical instruments business. I don’t see any mentions of this stock online and I may be the first to write about it. The lack of coverage is one of the reasons for its undervaluation, all while the business itself is quietly inflecting under the radar.

In classic Japanese fashion, management is extremely conservative with guidance, calling for only +3.3% growth for FY26 (March end) with an implied Q4 guide of -4.4% Y/Y. Their FY27 target as per their medium-term management plan is to reach ¥50B in net sales by FY27. In the P14 view, there is plenty of upside potential to this guide, driven by an inflection in semiconductor demand from the AI buildout while GLT simultaneously expands capacity meaningfully over the next two years.

Management projects a target operating margin of 15.5%, a target which I view as conservative given economies of scale as new capacity ramps and SG&A leverage. This is a high-quality industrial technology business already generating a ROE of ~11% and targeting >13% by FY27.

Due to its limited operating history following the business integration, the market is ignoring the underlying top-line growth potential and margin expansion. This ignorance is evidenced in the stock’s undemanding valuation at 10.4x FY26 guided EPS and ~1.08x tangible book value. This is a business set to see its top line inflect while EPS grows at a ~17.5% CAGR in the next two years as semiconductor demand rises meaningfully and capacity expansion begins contributing to revenue.

Management’s conservatism has set shares up for a violent re-rating, as seen in other Japanese semiconductor supply chain peers. A couple quarters of execution and earnings surprises should be enough for the market to recognize the earnings power embedded in the business, creating a credible path to a 2x opportunity within the next 12 months.

Business Overview

GLT is a newly formed holding company wrapped around three much older operating businesses. GL Sciences traces its roots to 1968, Techno Quartz to 1976, and GL Solutions was established in 2013. The combination was completed through a joint share transfer on October 1, 2024, with management framing the deal around 3 goals: seizing growth opportunities, improving management efficiency, and accelerating decision-making across the group. The stock went public on the TSE and closed at ¥2520 on its first trading day.

The group operates three segments: Semiconductor, Analytical Instruments, and Auto ID. The business mix has shifted toward Semiconductor in recent years as demand tied to semiconductor manufacturing equipment increased. On an LTM basis through 3Q26, Semiconductor generated ¥22,972M of revenue, representing 50.8% of total revenue, and ¥4,273M of EBIT with an 18.6% margin. Analytical Instruments generated ¥20,278M of revenue, representing 44.8% of revenue, and ¥2,090M of EBIT with a 10.3% margin. Auto ID contributed ¥1,994M of revenue, 4.4% of total revenue, and ¥85M of EBIT with a 4.2% margin.

GLT’s Semiconductor segment operates through its subsidiary Techno Quartz and focuses on high-purity quartz glass and crystalline silicon components used in the front-end of semiconductor fabrication. These parts function as jigs, rings, boats, and tubes inside high-temperature reaction chambers used for etching and Chemical Vapor Deposition (CVD). Because they are exposed to corrosive gases and plasma during wafer processing, the components degrade over time and must be replaced to maintain chip yield. This creates a recurring revenue profile, with ~70% of semiconductor sales coming from replacement demand. Demand has increased in the current AI-driven semiconductor capex cycle because advanced chips such as GPUs and HBM memory require more complex and precise etching steps, increasing the number and complexity of quartz components required per tool. Techno Quartz specializes in nanometer-level machining precision, which helps prevent wafer contamination in these environments.

The broader quartz supply chain is constrained. 60–70% of high-purity quartz raw material is sourced from a small number of regions, creating vulnerability to supply disruptions. While GLT does not control raw material supply, it operates one of the larger installed bases of semiconductor quartz machining centers. A key technical differentiator is its patented Quartz-to-Quartz Diffusion Bonding technology, which joins quartz components at the atomic level using heat and pressure rather than adhesives or melting. This allows the company to manufacture complex quartz structures with internal channels for gas or liquid flow inside semiconductor equipment chambers. GLT can also fabricate both quartz glass and crystalline silicon components, giving equipment manufacturers access to multiple critical materials from a single supplier.

The company competes with established quartz component manufacturers including Shin-Etsu Quartz and Ferrotec Holdings, typically focusing on high-mix, low-volume, high-complexity parts rather than commoditized quartz products. Large competitors dominate high-volume production of simpler tubes and standard parts, while GLT focuses on complex geometries and components used in advanced semiconductor processing equipment. Demand for these products currently exceeds supply, with production running near capacity. Management has therefore committed ~¥11B in capital expenditures to expand facilities in Fukushima and Yamagata through 2026–2027. The company estimates the broader semiconductor equipment market TAM at ~¥100T by 2030, while industry research suggests the quartz semiconductor materials market is growing at roughly 7–9% CAGR.

GLT’s customer base is concentrated among the global elite of Semiconductor Manufacturing Equipment (SME) makers. Its primary customer is Tokyo Electron (TEL), the world’s third-largest equipment manufacturer. Beyond TEL, GLT serves other major Japanese and international equipment players such as Screen Holdings, Applied Materials (AMAT), and Lam Research (LRCX). GLT was recently awarded by AMAT for outstanding performance.

The Analytical Instruments segment operates through GL Sciences and sells chromatography instruments, chromatography columns, sample preparation devices, and related consumables. Chromatography separates complex chemical mixtures into individual components so they can be identified and measured. GLT’s products are widely used across life sciences, pharmaceuticals, chemicals, environmental testing, food safety, and industrial analysis, with incremental demand now emerging from data center cooling and semiconductor-related chemical testing. 66% of segment revenue comes from consumables, creating steady replacement demand as laboratories replenish columns and other supplies. ~75% of analytical revenue is generated in Japan and ~25% overseas, with international markets gradually increasing their contribution. Industry estimates place the chromatography instruments market at ~$9–10B, growing at mid-single-digit rates, while chromatography consumables grow somewhat faster due to recurring demand from the installed base of analytical systems.

A key competitive advantage is Surface Deactivation Technology, which allows GLT to manufacture ultra-inert chromatography columns. If the column surface interacts chemically with the sample being analyzed, test results become unreliable. GLT’s technology minimizes these interactions, allowing samples to pass through the column without contamination or distortion. The company competes with large analytical instrument vendors such as Agilent Technologies, Waters Corporation, and Shimadzu. Those firms primarily focus on selling the analytical instruments themselves, while GLT concentrates on the consumable columns and components used across multiple instrument platforms. Its columns are designed to work across different manufacturers’ systems, allowing the company to compete on separation performance and reliability rather than equipment lock-in.

The Auto ID segment is the smallest business. Through GL Solutions, the company develops RFID modules, access control systems, parking management systems, and embedded identification products used in infrastructure, logistics, and security applications. Customers include infrastructure operators, system integrators, and industrial users that require asset tracking, authentication, or automated identification. The segment is small relative to Semiconductor and Analytical Instruments and operates at lower margins.

GLT operates manufacturing and sales locations across several regions. Production facilities are located primarily in Japan, with additional manufacturing in China and new semiconductor capacity under development in Vietnam. Sales operations extend across Japan, China, Europe, and the United States. Semiconductor customers are primarily global semiconductor equipment manufacturers, while Analytical Instruments customers include laboratories, pharmaceutical companies, and industrial testing facilities in Japan and international markets.

Capital allocation is focused on expanding semiconductor capacity while maintaining a strong balance sheet and shareholder return program. The company currently operates with a net cash balance sheet and targets a 30% dividend payout ratio. Management guides to a ¥111 dividend per share for FY26 (+3.7% Y/Y), implying a dividend yield of ~3%.

The top-line thesis

The top-line thesis for GLT in semiconductors is a picks-and-shovels exposure to the AI infrastructure cycle.

Tool complexity is increasing as advanced chips require more aggressive etching and deposition steps. These processes accelerate the wear of quartz rings, silicon boats, and other chamber components used inside semiconductor manufacturing equipment. Because these parts degrade in high-temperature plasma environments, they must be replaced periodically to maintain wafer yield. The result is a recurring demand profile for GLT’s components that scales with wafer fabrication intensity.

Semiconductor segment revenue is ultimately driven by three variables: tool shipments from semiconductor equipment manufacturers, fab utilization across the installed base, and GLT’s own qualified production capacity.

In FY25, management stated that the segment benefited from abundant order backlog and high factory utilization, which pushed both sales and operating profit above the initial plan. The semiconductor business is not purely tied to AI infrastructure. Exposure also includes consumer electronics, industrial electronics, and manufacturing equipment markets, which were weak during the recent semiconductor downturn. As those end markets recover while AI-driven spending continues to expand, demand across the semiconductor equipment supply chain should broaden.

This demand environment is already visible in order activity. Semiconductor orders reached ¥6.01B in 3Q26, the highest quarterly level in recent history, confirming that equipment manufacturers are increasing procurement.

Evidence of rising demand can also be observed from GLT’s major customers.

• Tokyo Electron (TEL) recently increased its FY2026 capital expenditure plan by 48%, targeting a record level of investment primarily focused on DRAM and High Bandwidth Memory (HBM) production. Management indicated in early 2026 that the company is effectively sold out of capacity, citing cleanroom constraints and parts procurement delays as current bottlenecks. This is a direct signal that suppliers like GLT are being asked for more than they can currently produce.

• Applied Materials (AMAT): has also pointed to large-scale AI adoption as a key driver of demand. In its FY2026 outlook the company highlighted strong growth in recurring services and spare parts, a segment that includes consumable components used in semiconductor tools.

The most definitive proof that GLT’s customers are ramping demand is GLT’s own financial behavior. A conservative Japanese small-cap does not spend ¥11B (22% of its market cap) on new factories unless the demand is guaranteed.

In Japan, GLT is expanding production capacity for fire-processed quartz products in Yamagata City and automating machining operations in Kitakata City, Fukushima. Management states that these projects are intended to strengthen the semiconductor supply system and restructure the production network to support increasing semiconductor demand. Overseas, the company is constructing a new semiconductor production facility in Ninh Binh Province, Vietnam. The project is designed to increase capacity, improve proximity to customers, reduce logistics costs, diversify supply chains in response to U.S.–China semiconductor restrictions, and allow faster response to market demand. Groundbreaking took place in December 2025, full construction began in January 2026, and the plant is scheduled to come online in phases beginning May 2026 and January 2027.

In essence, the demand is there, but GLT’s current production capacity limits revenue growth. Management expects semiconductor capacity to expand ~1.5–2x over the next two years, with the first phase of new capacity beginning to contribute to revenue April 2026.

Once fully ramped, the new quartz production capacity is expected to support >¥3B of incremental annual revenue.

The Analytical Instruments segment follows a different growth profile. Demand is less cyclical and more stable, with growth driven primarily by international expansion rather than domestic demand.

Management’s focus is increasing overseas penetration, where growth is stronger than in Japan. The U.S. analytical testing market is expanding ~7–9% annually, supported by biopharma R&D spending and strict FDA impurity testing standards. Asia ex-Japan is growing ~9–11%, driven by China’s domestic manufacturing expansion and India’s large generic pharmaceutical industry.

Historically, GLT has been viewed in the U.S. primarily as a supplier of chromatography columns. Management’s strategy now aims to reposition the business as a broader solutions provider. The company is expanding U.S.-based technical support, working to have its “inert” chromatography components specified directly within analytical instruments, and introducing automated sample preparation systems into laboratories facing technician shortages.

There is also overlap with the semiconductor industry. Analytical instruments are widely used in semiconductor chemical analysis, contamination monitoring, and process control. Expanding capacity and sales channels in overseas markets allows GLT to target this demand more effectively.

All in all, management estimates that new capacity expansion could support ~¥9B of incremental annual revenue once fully utilized. Note that this estimate was as of 10/24. Since then, demand related to AI capex has risen sharply, so it is reasonable to expect a comfortable surprise.

Revenue Build

Management guides to ¥44.7B of revenue in FY26, representing +3.3% Y/Y growth, which implies -4.4% Y/Y growth in 4Q26 (March quarter). In my view, this appears to be overly conservative given the current semiconductor backlog and the momentum in new orders entering the fourth quarter. In addition, management notes that Analytical Instruments revenue typically concentrates in the fourth quarter because many domestic customers operate on March fiscal year-ends. As a result, revenue recognition in both the Analytical Instruments and Auto ID segments tends to be back-end weighted toward fiscal year-end.

Looking at segment guidance provides additional context. Analytical Instruments sales are projected to grow +2.7% Y/Y to ¥20.5B in FY26, implying ~+3.7% growth in 4Q26. The larger discrepancy is in the Semiconductor segment. Management guides Semiconductor revenue to grow +3.2% Y/Y to ¥22.0B, which implies -17.2% Y/Y growth in 4Q26.

The reasoning behind this conservative semiconductor outlook is not entirely clear. Commentary in the 3Q26 release states that demand for personal computers, smartphones, and autonomous driving applications remains sluggish, though some markets are beginning to see higher device prices and supply tightness as demand tied to generative AI increases. Given the current backlog levels and the record ¥6.01B of semiconductor orders in 3Q26, it appears unlikely that revenue would decline as sharply as implied by the guidance.

One explanation is that management has set guidance conservatively to ensure the company is on track to meet its mid-term targets. In other words, the FY26 guide may reflect caution rather than operating conditions. Another factor is the difficult comparison with 4Q25, when semiconductor revenue grew +22.6% Y/Y, creating a high base. Even so, semiconductor equipment spending and AI infrastructure investment have continued to expand since then. On that basis, I estimate FY26 revenue growth of ~4.2% Y/Y to ~¥45B, modestly above management guidance.

The acceleration in revenue growth is likely to occur in FY27, when the majority of semiconductor capacity expansion becomes operational. New capacity in Japan and Vietnam is scheduled to come online between April 2026 and January 2027, allowing GLT to increase output for semiconductor components. Under this scenario, total revenue could grow ~13% Y/Y to ~¥51B, driven primarily by ~20% growth in the Semiconductor segment.

Looking ahead to FY28, Analytical Instruments should benefit from overseas expansion and new production capacity coming online, while Semiconductor growth continues but at a more normalized pace. Assuming ~12% Y/Y growth in Semiconductor and continued steady expansion in Analytical Instruments, total revenue could increase ~10.35% Y/Y to ~¥56B.

Throughout the forecast period, Auto ID is expected to remain a small but steady contributor, growing modestly and stabilizing at ~5% of total revenue.

The margin thesis

Gross margins are set to expand with economies of scale in FY27 and FY28. The initial expansion will appear limited due to accelerated depreciation. While raw material prices are increasing, demand for GLT’s products is strong and pricing power is already taking effect, mitigating margin erosion. The most important factor is the increasing mix contribution from the semiconductor segment. Second is utilization of new capacity. Quartz machining and semiconductor component fabrication have high fixed cost elements once equipment is installed. As new capacity in Fukushima, Yamagata, and Vietnam ramps between FY26–FY27, incremental volumes carry higher contribution margins because the fixed processing infrastructure is already in place. Management’s commentary repeatedly links margin improvement to higher factory utilization and stable production operations. There is also a product mix component. Semiconductor components are becoming more complex as advanced chips require additional etch and deposition steps. Higher-precision components and more difficult machining increase value per part. That supports pricing and margin stability.

All in all, it is reasonable to expect 30 bps expansion in FY27 and 50 bps in FY28 gross margins.

Gross margin expansion will directly flow down to operating margins, where SG&A leverage starts to kick in, offset by likely increased hiring in international markets to expand the analytical instruments business. SG&A costs include engineering staff, sales infrastructure, and administrative overhead that do not scale proportionally with revenue. As semiconductor capacity ramps and revenue increases toward the ¥50B sales target in FY27, these costs grow more slowly than revenue, allowing margins to expand.

All in all, I expect operating margins of 15% in FY26 (vs 14.94% guidance) and upside to management’s 15.5% FY27 target, with P14 operating margins coming in at 15.6%, with the full margin impact of capacity expansion visible in FY28 as margins expand to 16.6%.

Such margin expansion, along with a declining contribution from non-operating income and a stable 30% tax rate, leads to annual EPS growth of 17.68% in FY27 and 17.29% in FY28.

Base Case Valuation

Due to its small size and limited operating history as a public combined entity, the market is undervaluing GLT for its underlying earnings growth potential.

Key peers like Ferrotec (which itself is undervalued) and Shin-Etsu trade at premium valuations on their FY27E consensus EPS forecasts of 13.1x and 21.5x. This is notable as GLT is offering a similar forward revenue growth profile to Ferrotec (13% in FY27) and higher net income margins (11.1% vs 7%).

Should the base case play out and the stock start to get noticed, which is inevitably happening across the Japanese supply chain, P14 sees a comfortable re-rate to 17.3x FY27E EPS (the average of Shin-Etsu and Ferrotec), leading to a base case PT of ¥7,529, or +97% upside from current levels. I believe this is achievable in ~12 months.

Bear Case and Risks

Like most global equities, Japanese equities have taken a beating due to the current conflict in the Middle East. While GLT may experience second-order risks (logistics, uncertainty) in the event of a prolonged war, its supply chain is relatively shielded as most raw materials come from the Americas.

Competition risks exist, but this is a supplier’s market where supply is likely to be tight for the next few years as the industry responds to the boom in AI capex. GLT is relatively protected due to its specialty focus that requires scale and precision.

The major risk, which exists for all AI supply chain players, is a sudden stop in AI capex. While unwarranted at present, it is a tail risk.

In a bear case where growth fails to materialize and GLT underperforms its FY27 revenue and margin guidance, I see 18% downside to 8x FY27 bear case EPS, implying a bear case PT of ¥3,132.

Conclusion

P14 views GLT as a hidden beneficiary in the AI supply chain. It offers accelerating profitable growth and is due for a violent re-rating over the next 12 months, as seen with its peers. P14 has built a 5% position at an average cost of ¥4,100 and seeks to build this toward a 7–10% weight.

Disclaimer: Nothing posted by P14 Capital should be considered financial advice. The author of this post holds a long position in 255A. Please consult a financial advisor and/or conduct your own due diligence before making investment decisions.

NYSE: KLAR | 03/02/2026

After letting the feeling of “eternal bagholding” marinate a bit following a disastrous print and guide, P14 is back with a clear mind and his view on the future of an investment in KLAR.

I understand that KLAR’s performance (-57.8% since the pitch date) is probably giving my research a bad rep. Let this underperformance serve as a reminder not to blindly follow me. I will be wrong at times. Always supplement my work with your own research, and enter, exit, and size positions based on your own risk tolerance.

That said, in my view, I am not wrong unless I determine the thesis is clearly invalidated. My horizon is 2+ years, and I may sell, trim, or add to the position without prior notice. For example, if the opportunity cost of holding KLAR is high, I may take the tax loss, allocate to a better opportunity, and revisit as long as the thesis is still “alive.” The days of simply holding a stock because you have duration on your side are over.

On the bright side, here are some recent winners:

A special shoutout to this stock. It’s been weak over the past two weeks, but has two major catalysts converging in the next two weeks: DeepSeek and earnings.

Disclaimer: Nothing posted by P14 Capital should be considered financial advice. The author of this post holds a long position in KLAR. Please consult a financial advisor and/or conduct your own due diligence before making investment decisions.

Now, back to KLAR’s print and the path forward…

TL;DR

There are a few crucial positives in what appears to be a temporary sea of negatives.

Management needs to improve its guidance messaging and accept that it is no longer operating in a forgiving VC environment. TMD guidance is undoubtedly difficult to peg, but providing a range rather than a single low-end figure for the FY would have been far more sensible.

KLAR is “suffering from success” due to rapid Fair Financing growth.

KLAR has a significant wall of worry to climb. At its current valuation, with a market cap below cash and debt securities, the market is pricing the business as if its mere existence is a threat to its future. Upside hinges on 1) execution on Transaction Margin $, 2) fintech apathy abating, 3) the end of a senseless sell-off driven by inaccurate perceptions of CrEdIt DeToRiAtiOn, subprime lending, or whatever BS is circulating online, and 4) continued BNPL adoption.

I believe the lock-up risk is largely priced in and overstated.

Verdict: A reasonable entry point for new investors with a 2+ year horizon. If you are bagholding like I am and did not hedge into earnings, add only on strength. That likely means fintech apathy dissipates, the CEO stops shitposting about SaaS and AI on X, and/or the stock finds support into the over-feared 03/09 lock-up.

Read the original thesis:

The Q and FY guidance

KLAR reported a disappointing quarter at large. GMV beat estimates marginally while revenue of $1.08B beat consensus by 0.88%. However, management missed on executing on their most important metric for a second consecutive quarter: Transaction Margin Dollars (TMD), with a huge 5.8% miss ($372M) vs the stupidly tight midpoint of $390-$400M guide provided in 3Q25 ER.

Missing on the most important metric 2x in a row is bad enough, but management, in their natural clueless self, provided an FY guide that missed literally everyone’s expectations. KLAR only provided the LOW end of expectations across all key metrics:

• FY26 GMV: > $155B, or >21.2% Y/Y

• FY26 Revenue: >2.80% of GMV implying a >6bps Y/Y take rate expansion, or >$4.34B, +>23.7% Y/Y

• FY26 TMD: >1.04% of GMV or >$1.61B, implying a >7bps expansion as a % of GMV, or a >183bps margin expansion with >+30.1% Y/Y growth

• Adjusted Operating Profit: >6.9% of revenue, implying >+505 bps of margin expansion

Given the continued downdraft in shares, let us first address the negatives and other drivers of downside. I paid special attention to the balance sheet because I have seen too many stupid takes, especially from large accounts on X. If you are a bear, I’d like to congratulate you for being right, but it is most likely for the wrong reasons, not that it matters for now.

Negatives and drivers of downside

TMD weakness

Revenue - Processing and servicing costs - Provision for credit losses - Funding costs = TMD

It should be obvious that the one metric the market cares most about is Transaction Margin $. Despite beats on the top line and GMV, it is really TMD that takes the cake. For those unaware, the weakness in TMD is due to accounting mechanics related to credit loss provisioning that are natural when taking on longer-term financing loans (Fair Financing or FF). For every single loan given to a consumer, the process is the same, KLAR must book a portion of the new loan as a cost upfront before realizing the full revenue stream. What does this mean?

It means losses are OVERSTATED in the near term and appear amplified during periods of rapid loan growth. For the $2.5B in U.S. FF originated in 4Q25, KLAR booked $80M in upfront provisions while recognizing only $40M in revenue. With Fair Financing ramping at +165% Y/Y in 4Q25, short-term pressure is inevitable. The key point is that future gains from these loans will carry very high incremental margins, and profits should recover sharply once growth moderates. Recall that FF generates interest income over the life of the loan, along with merchant fees at the time of transaction (generally below 2%). The graphic below illustrates how revenue contributions build as the loan ages.

In essence, the market is being impatient regarding the margin contribution of these loans and skeptical of their credit quality (addressed below). Additional TMD pressure stems from higher processing and servicing costs as KLAR expands in the United States, where Visa and Mastercard take rates are structurally higher, as well as from the rollout of the Klarna Card, which creates upfront issuance costs. Funding costs have also increased due to forward flow agreements and fair value adjustments on loans sold, since Pay Later receivables are typically sold below book while KLAR retains the higher-margin merchant fee stream. Importantly, KLAR still maintains an industry-leading cost of funding supported by its wide deposit base.

The major valuation driver is TMD. Smart market participants understand that buying KLAR here is essentially an accounting timing bet, as margins recover a few quarters down the line when FF growth slows. In a sense, KLAR is “suffering from success,” as the margin swing could be meaningful when it occurs. That said, the harsh market reaction appears to stem more from a lack of confidence in management’s execution than from a deterioration in underlying economics.

Terrible guidance execution

This is some of the worst execution I have seen from a newly public management team. It should be common sense that TMD is what truly drives value, especially given the prior miss.

After missing its 3Q25 guide, management should have guided more conservatively for Q4. The miss becomes more understandable when you consider the 193% growth in December, as that late-period ramp created a larger-than-expected drag on results. Nevertheless, management should have anticipated that rapid U.S. expansion, combined with funding cost dynamics and mix shift, would pressure TMD optics and should have widened the range accordingly. Simply providing a broader range would have addressed much of this. They finally did so for 1Q26 with guidance of 11% to 26% Y/Y growth, but by then the credibility damage was done.

The real issue lies in the full-year guide. In my view, management is throwing the kitchen sink of conservatism at it as a self-inflicted response to two consecutive misses, which unfortunately drags shareholders down with it. Instead of providing a range for the year, they offered only the low end. While framing the low end as a floor reduces the likelihood of technical misses, the market does not reward floors. After two quarters of weak execution, investors care about forward visibility, and management has instead set an overly bleak tone. The figures provided should be viewed as bare minimum thresholds, and it is highly likely that beats occur across the board for the full year.

Addressing credit quality and the balance sheet

Most bears have shorted KLAR on the premise that the loan book is deteriorating and filled with borrowers who cannot repay. While bears have had strong performance, the underlying credit thesis is fundamentally wrong.

A significant driver of downside across lower-income fintech has been fear around private credit exposure, who are all (mostly) exposed to software financing and circular AI deals. Major private credit players such as Elliott, KKR, and Blue Owl are buyers of BNPL loans as well as having exposure to software and AI.

The prevailing market perception is that private credit players are underwater, which leads to the assumption that BNPL loans must also be underwater, and therefore KLAR is under threat. There is also constant chatter about subprime stress, with comparisons to a 2008-style scenario, which is not occurring in the context of BNPL.

Most people do not understand that a typical BNPL loan is structured and sold in tranches. KLAR often retains the senior tranche, while junior tranches are offloaded to buyers such as NelNet and, more recently, Elliott for Fair Financing. In addition, ~90% of Klarna’s funding comes from its own retail deposits, mitigating reliance on external credit buyers. The stress narrative around private credit is overstated, and Elliott has only recently ramped its purchases of FF loans. That suggests the buyer is not seeing incremental risk in longer-duration financing loans to the extent the market implies. Furthermore, these FF loans are being sold at a premium, reflected in gain on sale, demonstrating the value institutional investors assign to these assets.

The best proxy we have for Klarna’s underlying consumer health is provisions and delinquency metrics across Pay Later and Fair Financing, both of which show stable to improving performance. Provisions as a % of GMV declined 7bps Q/Q to 0.65%, despite significant FF growth, partly reflecting loan offloading. On the underwriting side, Klarna’s U.S. Gini score, a measure of model accuracy, reached 0.78 in 4Q25, up from 0.36 in 2019, indicating a materially improved ability to distinguish between good and bad borrowers as the platform scales. For comparison, VantageScore (a traditional credit bureau model) 4.0 scores around 0.43.

Delinquency metrics have risen modestly but remain within target ranges. Pay Later delinquency rates in both the U.S. and the Group showed a slight increase through the first half of 2025, consistent with a portfolio shift toward Direct-to-Consumer card-backed products, including the Klarna Card, Debit Flex Card, and One-time Card. Fair Financing 60+ DPD in the U.S. increased from 2.7% in Q1 2025 to 3.7% in Q2 2025, reflecting rapid origination growth rather than structural deterioration.

Overall, nothing suggests a blowout in KLAR’s BNPL credit quality, despite prevailing market sentiment. The company has largely been lumped into the broader private credit trade.

Moreover, the balance sheet continues to be pristine. By offloading FF loans through forward flow agreements, incremental stress is reduced while preserving Klarna’s industry-leading cost of funding, with deposits accounting for ~90% of total funding. KLAR ended the year with $5.3B in cash and equivalents (cash of $3.8B + debt securities of $1.5B), 106% of its market cap and up from $3.2B as of FY24. Net debt, when valuing KLAR as a neobank, calculated as Notes payable + (Consumer deposits - consumer receivables) + (Payables to merchants - settlement and trade receivables) - Cash and equivalents, was ~-$2B as of 4Q25, even better than ~-$1.75B in 3Q25.

Fintech apathy

Fintechs have been sold left and right for some time now. The primary reasons are weak employment data, soft consumer sentiment, and fintech serving as a funding source to short in order to go long the AI trade.

In the P14 view, we may be entering a rare period of a combination of fiscal expansion (OBBBA) that supports the consumer, alongside likely monetary easing in response to a weakening labor market. While headline sentiment data remains soft, I expect this to improve as rates fall. Consumer spending continues to show resilience, and fintech is primed for a comeback. When that occurs, given rising BNPL adoption, KLAR shares should benefit.

The dreaded lock-up

On top of the factors weighing on the name, the lock-up is approaching. This makes management’s FY26 guidance execution even more puzzling, given that the stock needs support heading into what many perceive as a trading liquidation event.

11x the float unlocks on 03/09, and that is top of mind for prospective investors. On one hand, I believe the buy side, including hedge funds and long-only funds, may welcome the additional liquidity because the fundamentals have not deteriorated. At its core, KLAR is largely an accounting timing bet once you look through the structural BNPL growth and healthy balance sheet.

On the other hand, the lock-up may already be priced in. This is not based on a mathematical model, and we could very well move lower, especially given recent geopolitical developments. Still, every trade requires a buyer and a seller, and it seems unlikely that VCs who entered late in the 2021 funding cycle would offload at steep losses driven more by weak execution optics than by a material deterioration in fundamentals.

Sequoia is the largest holder at 15%. Its board representative, Andrew Reed, has been seen buying shares of other bombed-out IPOs such as Figma. In my view, the lock-up risk is over-feared.

Positives

In a sea of negativity, there were several meaningful positives this quarter.

Neobank agenda success

Klarna appears to have sharpened its focus around a single objective: becoming a full-fledged neobank. The card has been a clear success, with active card users reaching 4.2M, doubling Q/Q, and Card GMV up +209% Y/Y.

The number of “banking consumers,” defined as users of deposits, cards, or term loans, doubled Y/Y to 15.8M. This cohort generates more than 3x the revenue per user compared to the broader consumer base, reinforcing the economics of deepening engagement.

Direct-to-Consumer (“DTC”) card-backed products, comprising the Klarna Card, Debit Flex Card, and One-time Card, delivered full-year 2025 volume growth of 59% Y/Y, significantly outpacing overall merchant checkout volume growth of 18%.

Fair Financing

GMV for interest-bearing Fair Financing products grew 165% Y/Y, reaching 193% in December, and now accounts for 12% of total group GMV. Distribution continues to expand and should further benefit from the Walmart partnership. Management noted that 20% of the global merchant base now offers Fair Financing, up from 18% in the prior quarter.

Scale and Network Expansion

Active Klarna consumers reached 118M, up 28% Y/Y, while the total merchant network expanded to 966,000 partners, up 42% Y/Y. Klarna added a record 115,000 merchants in Q4 alone, bringing full-year 2025 additions to 285,000.

The Q4 acceleration was primarily driven by the success of Klarna’s PSP integration with Stripe. Card adoption should create a reflexive loop, allowing Klarna to present merchants with a larger and more entrenched consumer base relative to peers, which in turn can support additional direct merchant wins such as Walmart. During the quarter, Klarna secured several high-profile partnerships, including LEGO, Emirates, Vinted, and StockX.

Merchant reach was further extended through expanded Apple Pay and Google Pay acceptance in additional markets.

Looking ahead to 2026, a full year of Walmart should be accretive to Y/Y growth. Management struck a slightly conservative tone regarding new PSP ramps, but default-on partnerships with Worldpay and JP Morgan Payments are expected to launch, alongside continued scaling of Stripe and Nexi. While these partnerships materially expand the addressable TAM, rollout and merchant activation are likely to be gradual.

All in all, the P14 thesis remains intact, and the FY guidance floor management has set, however unnecessarily cautious, appears achievable across all key metrics.

Updated valuation drivers

Revenue build

Key changes include downward revisions across the board. I believe these projections are achievable and broadly align with the floor management set in its conservative guidance. I anticipate revenue growth of +24.2% in FY26 versus management’s +23.7% low-end guide, and +18.9% in FY27 versus consensus estimates of +18.8%.

TMD & AOI Build

Key changes reflect lower margin expectations versus the prior update. FY26 TMD aligns with management guidance. Greater clarity on FY27 will depend on execution throughout 2026. My FY27 margin expansion assumption of +120bps may prove conservative, but it ultimately hinges on management delivering in FY26.

I believe the floor of AOI at 6.9% of revenue in FY26 is easily achievable. For all the CEO’s excessive commentary on SaaS, AI, and cost discipline, it would not be surprising if adjusted operating income exceeds that floor. My current estimates are 7.1% in FY26 and 10.7% in FY27.

Base Case and the path forward

Management has clearly spooked the market ahead of the lock-ups, which are being perceived as a liquidation event. In my view, much of the lock-up risk is priced in at current levels, though you never truly know. The stock may bottom as we approach the expiration date on 03/09.

Lock-up aside, the thesis is still alive. BNPL adoption is projected to rise for years, and KLAR, as the global market share leader, should benefit disproportionately. The company’s success with its neobank initiatives is notable and reflects a genuinely customer-obsessed culture. This is evidenced by Klarna’s consumer app consistently leading download charts versus peers. Over time, that neobank traction should allow Klarna to take share from players like AFRM as the market matures, simply because it is more deeply embedded in the consumer’s commerce ecosystem.

The market’s persistent concern around credit quality should dissipate with time, as it is not supported by the data. BNPL continues to demonstrate better credit performance relative to traditional revolving credit, and underwriting models are improving with scale and data. In the event of a broader credit tightening cycle, KLAR is arguably among the safest BNPL players given its deposit base and balance sheet strength.

Management’s conservative FY guidance suggests a longer-term orientation, at least implicitly. Profitable growth should be the focus in 2026 and beyond. If the executive team does not realize yet that Transaction Margin $ and a clear path to sustained GAAP/IFRS profitability are what drive valuation, then perhaps another career is suitable for them.

Public markets are unforgiving and fundamentally different from the VC ecosystem. Management needs to communicate accordingly. Rather than spending energy tweeting on SaaS, the focus should be on educating the market about BNPL unit economics, credit performance, and the margin trajectory. TMD volatility may be difficult to forecast quarter to quarter, but other areas of the P&L can be streamlined and communicated with more discipline.

From the P14 perspective, provided management executes, investing in KLAR at current levels is largely an accounting timing bet that should normalize over several quarters. The company is absorbing short-term margin pressure due to rapid Fair Financing growth, which should naturally moderate beginning in 2H26.

At current levels, KLAR holds ~108% of its market cap in cash and equivalents, including debt securities. With enterprise value at ~$3B, the stock trades at ~0.7x EV/FY26E sales and ~3x P/calendar year 2026 TMD, compared to AFRM at ~7x.

This discount is too wide, particularly given that KLAR represents a meaningful competitive threat to AFRM in the U.S. While one can argue differences in underwriting or margins, KLAR should experience similar multiple expansion once long-term financing growth stabilizes and margin optics improve.

For new investors, current levels represent a reasonable point to initiate a starter position. For existing investors, you should look to add on strength.

In my natural bagholding state, it appears prudent to push out my price target by a year or so. 2026 is likely to see lumpy TMD growth, especially in 1H26, before normalizing in 2H26. If the base case plays out by FY27, with an achievable 309bps TMD margin expansion vs FY25, P14 sees KLAR’s P/TMD multiple re-rating to 6–8x. At 6x FY27 TMD, the implied PT is $31/sh, or 133% upside from current levels. At 8x, the implied PT is ~$41/sh, or 211% upside. In the P14 view, this $31–$41 range is achievable within two years.

Note that this still implies a continued discount to AFRM, while also assuming AFRM’s multiple expands in 2027. I do not believe the industry faces a long-lasting structural threat that warrants such steep multiple compression across the board. This should begin to reverse as consumer sentiment improves during the cutting cycle.

Bear Case and Risks

The stock is trading at a market cap below the cash it holds on the balance sheet. That alone is close to a bear case. At this point, fundamentals are not fully driving price, so at peak pessimism, I would expect a downside scenario closer to $10/share.

The same risks outlined in the original thesis still apply. The primary risk management can control is execution.

Conclusion

KLAR is a timing and accounting bet. It trades at an undemanding valuation for a market share leader in a structurally growing market, and the lock-up risk appears largely priced in. I continue to hold my stake (5.6% at an average cost of ~$30) and will look to average down on strength, assuming the thesis still holds and the opportunity cost allows me to do so.

With its second earnings release (4Q25) since going public set for tomorrow (02/26), now is a good time to get familiar with an underrated AI capex beneficiary that combines industry-leading growth with strong margins.

Happy Reading!

Q - Qnity Electronics, Inc.

NYSE: Q | 02/24/2026

Given the lack of time ahead of earnings tomorrow pre-market, this pitch will be quick and sweet. I am only laying out the high-level thesis here. In the coming days, I will expand on this with my revenue and margin model. Let’s start with the business.

Business Overview

Qnity Electronics is a 2025 spin-off of DuPont’s ElectronicsCo, created as a standalone, pure-play electronics materials company with two operating segments: Semiconductor Technologies and Interconnect Solutions. The company manufactures process materials used in chip fabrication, advanced packaging, printed circuit boards, and thermal management. These products are qualified into customer production flows and scale with wafer starts and electronics unit volumes. FY24 revenue was $4.3B, split $2.45B Semiconductor Technologies and $1.89B Interconnect Solutions, with segment Operating EBITDA of $874M and $448M, respectively. Semi runs mid-30% EBITDA margins, Interconnect mid-20s depending on mix, supporting a consolidated margin framework of ~30%. Management targets ~6–7% revenue CAGR and ~7–9% EBITDA CAGR over the medium term. Pro forma net leverage is ~2.5x with a stated target below 3.0x.

Customers include leading global semiconductor manufacturers and electronics OEMs, with Samsung and TSMC among the largest accounts. Revenue skews toward Asia given global fab concentration, followed by the Americas and EMEA. The Semiconductor Technologies segment is largely consumables-based, with ~90% of the portfolio tied to products used in every wafer processed, directly linking revenue to utilization and wafer starts. End-market exposure spans smartphones, AI/HPC and data centers, automotive, industrial, and communications infrastructure. Qnity has meaningful consumer electronics exposure (~40% of sales), which introduces seasonality with stronger 2Q/3Q orders ahead of holiday production and softer 1Q/4Q periods. Advanced packaging and thermal solutions tied to AI and data centers have driven faster growth in recent quarters.

Competition is concentrated among scaled specialty materials players such as Entegris, Element Solutions’ electronics segment, and MKS Instruments. Qnity holds leading positions across several semiconductor process material categories and advanced packaging applications, with research suggesting ~15% share in its semiconductor processing chemical and ~10% in interconnects and advanced packaging markets. The business benefits from long qualification cycles, process integration with customers, and local manufacturing aligned with major fab hubs, which supports switching costs and share durability. Advanced packaging is a core growth vector across both segments, with exposure to high-layer-count PCBs, thermal interface materials, and heterogeneous integration tied to AI and high-performance computing. Segment leadership brings decades of semiconductor and electronics experience, with significant operating exposure to Asia where the majority of revenue is generated.

Quick Thesis

In Semiconductor Technologies, Qnity is well positioned to benefit from rising semiconductor materials content per wafer over the coming years. The company estimates materials content per wafer to grow at a mid-single digit CAGR through 2029 as node complexity increases across foundry, logic, DRAM, and NAND. A key driver is Chemical Mechanical Planarization (CMP), where Qnity has strong share. CMP represents ~50% of the Semiconductor segment, and Qnity is the industry leader in CMP pads. As nodes shrink and layer counts rise, CMP intensity increases, directly supporting above-market growth.

While consumer electronics exposure is viewed as cyclical, increasing AI inference on device should be a structural tailwind for years to come. Qnity also has the opportunity to expand its TAM to include other verticals as its products are used in several applications.

All of the AI-focused substacks have been harping on advanced packaging since the start of this year. Qnity is an underrated beneficiary in this realm. Advanced packaging is a differentiator for Q relative to its peers, and it has risen to ~10% of sales in recent quarters, spread across both the Semiconductor Technologies and Interconnect Solutions segments. As key AI chip and rack vendors roll out future roadmaps, it is becoming increasingly clear that higher power consumption and the greater need to integrate technologies like High Bandwidth Memory (HBM) closer to core GPU dies are driving demand for more sophisticated advanced packaging and substrate technologies.

The Investor Day revenue growth target over the near term (<3 years) is a 6–7% CAGR. This assumed CAGR is projected to be ~200bps higher than broader semiconductor materials growth. It is crucial to note that this growth does not assume an industrial chip recovery, which in my view makes the bar conservative given improving macro manufacturing data.

Leading indicators suggest upside to Qnity’s 6–7% 3-year CAGR. First, US manufacturing data is improving, as seen in the most recent ISM prints. Second, management has stated that customer utilization rates are improving. When you place that in the context of Qnity’s largest customers — Samsung (11.6% of sales), TSMC (~7%), and SK Hynix (~7%) — all experiencing strong demand, the revenue mechanics are straightforward. Higher utilization means Q sells more pads, slurries, cleans, lithography materials, and related inputs per wafer. These are not one-time equipment sales. They are recurring process materials used every production cycle.

Another tailwind is the China localization story. Unlike ENTG, Qnity’s China sales are largely sourced from non-U.S. production, limiting exposure to technology restrictions. With ~34% of net sales attributed to China and domestic chip production ramping, Qnity is positioned to capture incremental volume as capacity expands.

All in all, P14 sees upside to the 6–7% CAGR framework over the next three years, with my forecast for 2026 revenue at +8% Y/Y (vs +8.5% in FY25). We should hear more about this on the earnings call tomorrow (02/25). Big picture, with improving industrial data and key customers ramping production, Q is positioned to outperform consensus 2026 growth estimates (~+7.5% Y/Y). Recent commentary from peers like ENTG and ESI also points to acceleration in electronic chemicals.

On margins, management guides to a long-term consolidated EBITDA margin of ~30%. Gross margins are currently 46.3% LTM, and near-term expansion is reasonable as utilization improves, driving better fixed cost absorption. Mix is another lever, with advanced-node exposure carrying higher margins. Within Interconnect, AI-related packaging and thermal products carry better profitability than legacy consumer PCB exposure. If packaging and data center demand continue to outgrow consumer electronics, consolidated gross margin benefits. We are in a memory recovery phase, with strong DRAM and HBM demand improving both volume and mix. Operating leverage should begin to kick in as the company advances through standalone optimization, including SG&A rationalization and footprint adjustments. I project EBITDA margins to be ~30% in FY26.

Relative value

Q, despite being the largest pure-play electronic chemicals company, trades at a material discount to its closest peer ENTG on an EV/EBITDA basis. We should compare the two on EBITDA, not P/E, given differing tax rates (18% vs 7%).

Based on revenue growth, margin profile, and market positioning alone, Q deserves to trade closer to where ENTG trades. Mr. Market is applying two discounts. First, spin-off skepticism and a short public track record. That should fade after a few clean quarters of execution. Second, exposure to “lower quality” businesses such as printed circuit boards. Within that segment, however, mix is shifting toward higher-end products like packaging substrates and advanced packaging inputs tied to AI and data center builds.

With more time in the public markets, and if the base case plays out with consistent outperformance versus consensus, Q’s multiple should expand toward peer levels.

Risks include seasonality, as 4Q and 1Q prints are typically the weakest due to consumer electronics exposure, given that 2Q/3Q are the primary ordering quarters. Obvious risks include a fade in the AI trade and a hard stop in the manufacturing and industrial chip recovery. ~34% of Qnity’s sales are tied to China, which is vulnerable to potential disruptions from trade tensions, regulatory changes, or shifts in global supply chains.

Base Case

The stock currently trades at ~17.6x 2026 EBITDA, assuming Qnity generates ~$1.52B in EBITDA in FY26, which implies a ~30% margin on ~8% Y/Y revenue growth. Assuming stable cash generation, steady leverage, and that the multiple expands toward ~25x (in line with ENTG) on FY26 figures, this points to ~48% upside in under two years. The bear case is likely closer to the IPO price of ~$95/share, implying ~20% downside.

At the time of writing, P14 owns a 5.2% position at an average cost of $112.70/share. I will expand on this quick thesis following 4Q25 earnings tomorrow (02/26).

Disclaimer: Nothing posted by P14 Capital should be considered financial advice. The author of this post holds a long position in Q. Please consult a financial advisor and/or conduct your own due diligence before making investment decisions.

NYSE: GXO | 02/23/2026

With a 50% return since the pitch date on 01/17/25, it has been rewarding being a GXO shareholder, especially given that this was my first ever pitch on the Substack. For new readers, P14 now has a paid tier at $20/month or $220/year (best value), which gives you access to high quality pitches and thematic screens like the recent ones below.

Subscribe now

It always feels good when the market starts to align with your thesis, especially when conviction pays off because you were adding on dips along the way. In GXO’s case, the position reached 20% of my portfolio at its peak. I trimmed it to ~14% after the release, which feels questionable looking at the chart, I know.

The positive reaction to the Q4 print surprised me. My immediate takeaway was that the quarter was decent but below my own expectations, and I was frustrated by what I viewed as another conservative guide that I initially thought would weigh on the stock.

Fortunately, the market did not see it this way. Unfortunately, I trimmed my position too early, but GXO is still my largest holding, which I plan to carefully ride with the momentum. Notice the unusual yellow font on my BUY recommendation, that means something. I would not initiate a full position here, but I would trade the momentum while holding my core stake depending on the opportunity costs I see in the market. The thesis is still alive and well, and the market is now beginning to see GXO for its merits and call management’s bluff.

I am not a chartist, but excluding today’s 02/23 selloff, the stock is approaching a key resistance level around $67 per share. A clean break above that level likely opens a path toward ~$80. I wanted to clear that out first before diving into what actually drives the stock there.

Before getting into the quarter and thesis update, I strongly recommend reading the original pitch and the 3Q update below to understand how the thesis has evolved since 01/17/25.

GXO reported a decent 4Q25 with revenue meeting consensus at $3.5B, a slight beat of 0.8%, while Adjusted EPS beat by 5%, coming in at $0.87. The company posted in-line guidance:

• Organic Revenue growth of 4-5%

• Adjusted EBITDA of $930–$970M, +7.8% Y/Y at the midpoint, with 6.9% margin at the midpoint, +20bps expansion

• Adj. diluted EPS of $2.85–$3.15, +20% Y/Y at the midpoint

• Adj. EBITDA to FCF conversion of 30–40%, implying +28% Y/Y FCF growth at the midpoint

In an effort to not jinx the stock’s momentum, I will start by covering the negatives, or what I did not like this quarter, followed by the positives and developments to the thesis.

Negatives

Weaker than expected Q4 revenue growth

In my 3Q25 update, I expected 4Q25 organic growth of 7.3% and 5% for FY25. That was too optimistic given management’s prior commentary around flat to declining volumes. 4Q25 organic growth came in at 3.5% and FY25 ended at +3.9% Y/Y. This was below management’s midpoint guide of 5% for the year within the 3.5% to 6.5% range. Fortunately, consensus estimates were aligned with the print, so the stock did not react negatively to the miss relative to the midpoint.

The underperformance versus my expectations appears to have been driven by a deceleration in volumes tied to weaker general merchandise demand in continental Europe and the UK during Q4. The United States also posted flat revenue Y/Y. This likely reflects inventory depletion during the holiday season, as inventories were already thin entering Q4 after being front-loaded earlier in the year to avoid tariff uncertainty. Lower inventory levels translate directly into lower volume throughput.

Margin pressure

FY25 EBITDA met my expectations, but margins worsened slightly amid the Wincanton integration, which was flagged in 3Q25. More than half of the margin decline was due to gains last year that did not repeat, making comparisons tougher. While integration-related compression is understandable, there has been little commentary around multi-tenant warehouse utilization, which I had expected to support gross margins in the near term. Instead, the margin narrative has shifted more to other drivers, which I address below.

The return of tariff uncertainty?

On Friday, a SCOTUS ruling nullified some tariffs, leading to a rally in logistics stocks including GXO. The rally intensified after the White House announced a 10% global tariff, but over the weekend that rate was raised to 15%. The absolute rate is less concerning than the policy uncertainty. It is reasonable to expect names like GXO to experience short-term volatility as markets digest shifting trade headlines.

Conservative guidance

It appears that the market has called GXO on its bluff with the 2026 guide. I am glad conservatism worked out this time. Perhaps it was the messaging around other elements of the business that resonated, or certain components of the guidance that were viewed positively. GXO may now be priced as a 20% EPS grower, which alongside other factors could justify multiple expansion from its current ~21x P/F26E Adj. EPS.

The guide implies only a 55bps Y/Y acceleration to 4.5% organic revenue growth at the midpoint, which feels conservative given the tailwinds ahead. The caveat is that management continues to assume flat volume growth, which seems unlikely given improved tariff stability relative to 2025, inventory depletion at the start of the year that should drive replenishment and higher throughput, and a sharper focus on North America where manufacturing is reaccelerating. In management’s own words on flat volumes, “We just take a flat number for prudence. Just take it as prudence, nothing more than that.”

Margins tell a similar story, with guidance calling for only 20bps of expansion. Management attributes this to synergies being weighted to the back half of the year and to contract start-ups that carry lower margins before ramping to steady state levels.

Positives and developments

New business wins

FY25 wins totaled $1.1B, a third consecutive year >$1B. As of 4Q25, GXO has already secured $774M of incremental FY26 revenue, +23% Y/Y vs the comparable year ahead position last year of $627M. The sales pipeline stands at $2.3B at 4Q25, supporting continued bookings. Management expects churn similar to FY25, roughly 95% retention or MSD churn. Assuming conversion at that level, we are already tracking toward ~6% organic revenue growth before incremental upside from new wins.

A lot of chatter on automation

GXO is leaning further into a theme the market is beginning to reward. Recall that GXO was one of the first to deploy humanoids in a live operating facility and is considered a pioneer of robotics in 3PL. Management is zeroing in on this with GXO IQ, an AI powered warehouse operating system that improves labor planning, inventory movement, forecasting, and workflow management across large sites

Early pilots showed measurable gains in proactive replenishment and slotting, and the company plans to scale GXO IQ to more than 50 existing sites in 2026. This initial rollout covers roughly 4% of GXO’s 1,200 operations, but it targets the most suitable facilities first, with further implementation expected during future site openings. Management tied these tools directly to productivity improvement through tighter labor planning, better operational visibility, and replication of best practices across the global network, which are the core levers in a contract logistics model with thin margins.

On robotics, management emphasized scale and speed of deployment. By the end of 2026, they expect nearly 20,000 robots in operation and several humanoid pilots across all three regions. They believe physical AI and humanoids can expand the range of warehouse tasks that can be automated, increasing throughput per site and reducing reliance on variable labor. The stated objective is clear productivity gains that translate into margin expansion over time, stronger implementation performance on new contracts, and a more differentiated value proposition in high growth verticals such as life sciences, aerospace and defense, and data centers.

Penetration in high-growth verticals

Management framed the shift toward higher growth verticals as a deliberate reallocation of commercial focus toward end markets with faster demand growth, higher complexity, and stronger value based pricing. They highlighted life sciences, aerospace and defense, industrial, and technology focused on data center infrastructure as priority verticals, and tied this to the broader view that the contract logistics addressable market exceeds $500B. They also pointed to tangible traction in 4Q with meaningful contract wins in life sciences and multiple aerospace and defense wins, including expansions with Boeing and new business tied to BAE Systems and Thales, partly linked to the Wincanton acquisition.

In data centers specifically, they described the market as part of the rapidly expanding AI and cloud infrastructure ecosystem, and said they won five new contracts in the quarter, including multi region wins with a leading hyperscaler.

Upbeat margin outlook and Investor Day

While offering a conservative FY26 margin guide, management hinted at stronger long term margin expansion potential. Aside from fundamentals, this suggests a management team that understands messaging, unlike some other companies such as KLAR (4Q25 ER update coming soon).

In response to a question on near term margins alongside automation investments, the CEO said, “I can confidently say right now, we are aiming to deliver at margin levels at or better than our peer group.” More detail will likely come at the Investor Day later this year. Based on my estimates, the margin expansion opportunity exceeds 100bps over the next 3+ years, primarily driven by automation and GXO IQ, which management calls a game changer. Robotics should also contribute and supports the narrative, while multi-tenant warehouse utilization remains, in my view, an additional lever.

The Investor Day is shaping up as a major catalyst for continued re rating. It will be important to see how 2027 targets compare to those outlined at the January 2023 Investor Day, when GXO anticipated $17B in revenue and $1.6B in Adjusted EBITDA, implying a 9.4% margin by 2027. Those targets were set prior to the tariff disruption and the slowdown in the EU, so updated estimates will likely reflect a more tempered backdrop.

Free cash flows

The least conservative element of the guide is the increase in Adjusted EBITDA to FCF conversion from 29.4% in FY25 to a midpoint of 35% for FY26. This likely supported the positive share reaction as GXO is increasingly viewed as a stable, contractual logistics business with improving cash generation.

Updated Valuation Drivers

I am looking forward to the Investor Day whenever it is held this year. I would like management to provide more detail on the revenue model, specifically volumes and pricing, because modeling by industry today still feels like educated guesswork. Given the conservative guide and muted FY25 growth, I am now projecting 5.9% Y/Y revenue growth in 2026 followed by an acceleration to 6.3% Y/Y in 2027.

My upside versus management’s 5% high end organic growth guide for 2026 is driven by my view that volumes are unlikely to remain flat for the full year. Absent another tariff shock, with a combination of fiscal and monetary stimulus in the United States, it is difficult to see volumes not increasing. The same applies to what appears to be a stabilizing consumer backdrop in the EU.

There is also uncertainty around the timing of the $774M in incremental 2026 revenue and when those contracts ramp. For that reason, I am modeling growth below 6%, which is roughly where it would land if all of that signed incremental revenue converted at a 95% retention rate. Based on commentary, technology and consumer electronics appear positioned to improve, while industrial and manufacturing, including aerospace and defense, should continue their momentum.

I expect broader acceleration in 2027 as 2026 start ups mature, alongside the structural tailwind of increasing 3PL adoption and deeper penetration in North America, which is the largest 3PL market and the core automation opportunity.

I would highly appreciate if management becomes more granular with their revenue model, especially with the new hyperscaler contracts and what do they actually entail and if there are any volume or pricing tailwinds associated with such high-growth verticals.

You will notice that my forward revenue projections are a clear step down from the 10.9% and 10.8% assumed in 3Q25. I believe shareholders will gain more clarity at the Investor Day on how to think about GXO’s revenue growth trajectory.

On margins, I am aligning with management’s 2026 guidance, modeling 6.9% EBITDA margin, a 20bps expansion, driven by Wincanton synergies weighted toward 2H26 as well as contributions from automation and multi tenant warehouse utilization. I model a further 30bps expansion in 2027 to 7.2% as start ups mature and productivity gains compound.

These revenue and margin estimates are fluid with an upward bias as new leadership executes and commercial focus shifts further toward underpenetrated North America, where automation opportunity is greatest.

Given likely inventory restocking and higher volume throughput in Q1, I believe there is a reasonable probability management raises its FY guide at the Investor Day.

Base Case

With a ~20% return since the start of 2026, multiple expansion has begun to run its course. This has largely been driven by sector flows into logistics stocks, as highlighted in my logistics screen on 01/18, which GXO has clearly benefited from. Following what was a decent print and a conservative guide, though in line with consensus, it appears the market has looked through management’s conservatism and is now expecting upside to the guide.

In the P14 view, with a combination of fiscal and monetary easing this year, backed by a stimulated consumer through tax refunds, along with a risk-off rotation out of non-AI services stocks, flows into transportation and physical industries are likely to persist, which should continue to benefit GXO absent idiosyncratic issues. In addition, GXO is positioned to manage further tariff volatility as it expands its U.S. Foreign Trade Zone capabilities.

Even with lower projected revenue growth and margin expansion versus prior estimates, GXO’s multiple is likely to remain at or expand beyond the current LTM multiple of 14x. This is still below the 5Y peak multiple of ~21x (2021). Also note that GXO’s model treats lease expenses as pass-through to customers, so on an EV/EBITDAR basis of 5.9x, the stock still trades at a meaningful discount to the industry.

Applying the current 14x LTM multiple to FY26 base case projections, coupled with a 2% reduction in the float and a stable cash generation, implies an FY26 price target of ~$74 per share, or ~17% upside over the next year.

Bear Case

The usual risks still apply. More recently, tariff volatility is the most visible risk, though a potential mitigant is that we are in a midterm election year. Integration risk exists, as management has already cited delays in synergy realization flowing through to margins. Given the experience of the new leadership bench, I view the risk of a failed North American expansion as contained, but after the stock’s strong run this year, it would not take much to reverse sentiment.

A weaker consumer is another risk, though that is not evident to me in the near term. In a bear case scenario, assuming 3% revenue growth in the near term and 20bps of margin compression, GXO could trade down to ~$48 per share, implying ~24% downside.

Conclusion

Until the Investor Day, which could prompt upward earnings revisions, the risk reward is not compelling for initiating a full position or making a meaningful add. That does not change the fact that GXO is a quality business experiencing a gradual acceleration of profitable growth. If the base case plays out alongside improving macro data such as the recent ISM report, and if sector flows remain supportive, the stock price and thus the multiple have room to expand. GXO has not yet been fully grouped into the emerging robotics theme, and if that shift occurs, I would not be surprised to see the multiple expand toward 18–20x over the long term.

I continue to hold a 14% position at an average cost of $47.11 per share until I find the next compelling opportunity. Hint, it may be my next core pitch releasing this or next week!

Disclaimer: Nothing posted by P14 Capital should be considered financial advice. The author of this post holds a long position in GXO. Please consult a financial advisor and/or conduct your own due diligence before making investment decisions.

NYSEAM: ELMD | 02/13/2026

For the original thesis, please see below. This is a paywalled pitch, so all updates related to the stock, until the position is closed, will stay behind the paywall. This applies to all P14 paid pitches for the rest of earnings season. My next updates on GXO, DSP, and KLAR will be free, as those were free pitches. You can track the status of all pitches under “Core Stock Theses” on my website.

Teaser

I’ve seen many irrational post-earnings moves, but ELMD being up ~10% AH on the print, then dropping ~17% over the next two days before reversing ~+4% on Friday, despite wide beats on both the top and bottom line, likely ranks in the top 5 most irrational. And no, this is not a SaaS name that would warrant that reaction. This looks like a gift, IMO.

Accelerating top line with an expanding TAM; disease diagnosis growing double digits annually and now reaching an INFLECTION point

Double-digit revenue growth with operating leverage; profits are inflecting (op income +42% Y/Y in 2Q26, net income +40% Y/Y)

High-quality business with 78%+ gross margins & 18%+ EBITDA margins (EBITDA grew 32% in LTM, margins expanded 235bps)

EBITDA margins expanded 878 bps since FY23 and are expected to expand another ~ 400bps in N2Y, yet the stock trades at 11x EV/FY27E EBITDA

Revenue grew at an average ~15.4% Y/Y since FY23 and is expected to average ~15.4% Y/Y in N2Y, yet the stock trades at 2.3x EV/FY27E Sales

Buying back shares with excess cash (share count -3% Y/Y in 2Q26), supported by a new, larger repurchase authorization

Prime acquisition target for large medtechs and private equity

Similar to last quarter, when the stock sold off for a few days before climbing toward $30, ELMD pulled back again after reporting another record quarter, though this decline was more severe. The stock recovered modestly on Friday 02/13. In the P14 view, this is another opportunity for longs created by the inefficient micro-cap market.

ELMD reported a strong quarter, with revenue of $18.9M +16.3% Y/Y, beating consensus by 4.7%, and GAAP EPS of $0.32, topping estimates by 23%. Management does not provide quantitative guidance but reiterated its framework of double-digit top-line growth and expanding operating leverage over the coming quarters and for full FY26.

Before diving into the positives and negatives of the Q, here’s a quick thesis refresher.

• Valuation not giving credit for the runway. Trading at ~15.5x LTM EBITDA vs ~17.5x healthcare equipment, with market underestimating forward growth and margin expansion. LTM EBITDA multiple has now contracted since the original pitch despite continued outperformance.

• Proven compounder on sale. Stock is up ~113% in L5Y (vs 16.4% IWM, 74% SPY) but ~31% below the $35 ATH despite the fastest revenue growth in nearly a decade (+17% Y/Y in FY25) and record profitability.

• Early in multi-year growth cycle. Rising bronchiectasis diagnosis, greater pulmonologist and payer awareness, and increased attention from Insmed’s brensocatib expand the demand backdrop.

• Share gains vs larger competitors. Strong payer-relations effort to modernize coverage rules, a beefed-up sales org, and SmartVest taking share from larger, distracted HFCWO peers.

• Margin expansion story. New CRM and e-prescribe systems improve sales efficiency, prior manufacturing investments scale, and mix keeps gross margins in the high 70s with further operating leverage ahead.

• Disciplined capital allocation. Management is focused on the core business, avoiding value-destructive M&A, and is now buying back stock with excess cash.

• Takeout optionality. Last pure-play HFCWO player, mid-to-high-teens growth, U.S. focused, embedded with pulmonologists, attractive to large medtechs and PE; CEO’s Shore Capital background supports this angle.

• Compelling risk/reward. Base case on FY27 implies 82% upside with bear case downside of 15%.

Positives and recent developments

Top-line tailwinds intact

2Q26 Y/Y revenue growth of 16.3% represented 120bps of Q/Q acceleration and marked the 13th consecutive Q of revenue growth. This was achieved against a tough 2Q25 compare of +18.7% Y/Y.

The homecare segment carried the quarter, posting +18.4% Y/Y growth and accounting for 91%+ of total net sales. Growth was supported by disciplined headcount expansion, with direct sales reps increasing Q/Q by just 1 to 58. Management plans to grow the team to 61 by FY26 year-end.

What is most impressive is the sheer sales productivity being exhibited by the reps. What this suggests to me is that reps are being trained faster than anticipated and smartvests are being sold/prescribed by physicians quicker because of the inflection in diagnosis rates. 2Q26 productivity came in at a record ~$1.2M per rep, exceeding management’s target range of $1-$1.1M per rep.

Management cited increasing awareness of bronchiectasis among prescribing physicians. More clinicians are identifying patients who qualify for HFCWO, and when combined with targeted sales outreach and clinician education, the result has been faster patient identification and prescription flow.

TAM reaffirmed and bolstered

This graphic should tell you all you need to know. The bronchiectasis patient base is materially underserved. The U.S. population is aging, immigration trends are not offsetting that shift, and prevalence is rising meaningfully in older cohorts. Despite rising diagnosis, 58% of patients who qualify for HFCWO therapy not receiving it. Recall that receiving prescriptions is a tedious and unnecessarily long experience. These are CT-confirmed patients with chronic productive cough for 6+ months who have failed prior treatments.

Management has advanced disease education efforts through completed manuscripts using data from the NTM Bronchiectasis Research Registry. The analysis showed that clinically eligible patients who were not prescribed HFCWO were significantly sicker than those who did not meet criteria. In parallel, the 2025 “triple-down on bronchiectasis” campaign is focused on physician education, earlier identification, and accelerating the path to prescription.

Insmed’s Brinsupri is effectively acting as free marketing for airway clearance. For Brinsupri to be effective, HFCWO therapies like the SmartVest must be used first. Brinsupri is off to one of the strongest launches in the respiratory space, reaching $145M in sales in its first full quarter, highlighting the unmet need in BE treatment. Management also hinted at expansion into natural adjacencies, including targeting neuromuscular disease states.

Expanding payer coverage

Given the ~$10K price tag on a vest, ELMD’s model is dependent on insurance coverage and continued expansion of that base. The payer relations team executed 25 contracts in the first half of the fiscal year, all private pay and commercial, adding 2.9M covered lives to the 270M+ lives currently under contract that can access SmartVest.

Progress on the CRM and Smart Order e-prescribe

Management highlighted the Smart Order e-Prescribe solution, which is improving how clinics submit prescriptions to the fulfillment team. As noted prior, this digital platform replaces fax-based processes and ensures complete documentation, enabling faster delivery and improving clinic workflow. In Q2, over one-third of orders were submitted through the platform, reflecting strong adoption.

The CRM system launched in Q1 is already driving benefits, including improved field sales productivity, better market visibility, and smoother coordination with fulfillment. This is resulting in higher rep efficiency and faster order processing, positioning ELMD to continue taking share from Philips RespirTech’s InCourage and Hillrom’s Vest franchise.

Operating leverage and margin expansion

Gross margin expanded 30bps Q/Q and 70bps Y/Y to 78.4% in 2Q26. The primary driver was higher realized net revenue per device. This was supported by favorable payer mix, including strong Medicare growth, which management identifies as a top payer value given higher allowable reimbursement rates. ELMD also completed its manufacturing optimization plan at the start of FY26, which included restructuring the facility and redesigning the layout to improve efficiency and increase capacity for future growth. As prescriptions scale with higher sales headcount and rising disease awareness tied to increased diagnosis, incremental volume should further support margin expansion, particularly with pricing generally increasing in line with annual inflation. ELMD’s 99% U.S.-based manufacturing footprint limits tariff exposure.

A risk to monitor is shipping costs, while not broken out individually, spot rates are increasing rapidly. So this may affect margins, but given the critical nature of this product, any increase should be passed on to customers with relative ease.

Operating income grew 42.4% Y/Y! Operating margin expanded 340bps Q/Q and 360bps Y/Y to 19.2%. SG&A increased just 10% Y/Y to $10.8M, reflecting disciplined cost control as revenue scaled. An underappreciated aspect is the leverage management is seeing with its reimbursement team. Continual changes in this team has enabled the team to now process a higher volume of referrals without a corresponding 1:1 increase in headcount, which provides “terrific leverage in the P&L”. Sales productivity is increasingly showing up in leverage as management claims that their training process has improved significantly for new hires which allows new reps to hit full productivity sooner rather than later.

All of this naturally led to outsized net income growth of +40.3% Y/Y.

Capital allocation

In 1Q26, the board authorized a $10M share repurchase program. As of 2Q26, the company has repurchased 146K shares for $3.77M, leaving ~$6.23M remaining under the authorization. Management framed buybacks as an opportunistic tool based on cash position and share price.

Given the sharp post-print decline, accelerated repurchases in 3Q26 would be reasonable to expect. The presentation referenced potential inorganic acquisition opportunities, though that appears more long term. At current levels, prioritizing the buyback while continuing disciplined headcount expansion to target a significantly underpenetrated market would be the preferred path.

Negatives

Caution on sales productivity

This is likely the main factor behind the post-print downside. Management expects future sales productivity to normalize back into the $1.0–$1.1M range, versus the ~$1.2M achieved in 2Q26. I suspect the market is interpreting this as $1.2M being a near-term ceiling.

When hiring ramps quickly, there is naturally a lag before new reps reach full productivity. The expected normalization is primarily driven by this timing lag as the company hires toward its FY26 goal of 61 territories. New reps require time for training and to build clinical relationships before operating at full capacity. That said, management noted that the training program is reducing ramp time.

The market appears to be pricing in incremental headcount as a drag on operating leverage. In my view, nothing in the quarter suggests diagnosis trends or prescription volumes are slowing.

Caution on payer mix

Another contributor to the downside could be management’s comment that 2Q26 productivity benefited from favorable payer mix, including outsized Medicare growth, which they described as volatile and not guaranteed in future quarters.

On the regulatory front, little has changed. The Medicare fee schedule continues to indicate pricing indexed to inflation, HFCWO remains excluded from competitive bidding, and clinical audit activity appears consistent. A potential positive is the WISeR Pilot Program launched January 1, 2026, testing technology-enabled prior authorizations in six states. Another is the 2026 Medicare Advantage Course of Treatment rule, which requires that once an MA plan approves an HFCWO device for a course of treatment, the authorization remains valid even if the patient switches providers or plans within the year. This reduces mid-year reimbursement risk.

Volatility in the hospital market

After +51.75% Y/Y growth in 1Q26, hospital revenue decelerated to -9.4% Y/Y in 2Q26. This swing may have triggered quantitative selling by algos. Revenue was negatively impacted by management’s decision to prioritize shipments to high-demand homecare patients over hospital orders.

Hospital sales are capital equipment purchases with unpredictable timing and long sales cycles. Revenue can vary significantly quarter to quarter depending on health system capital budgets. Despite the quarterly decline, management continues to describe hospitals as a gateway to homecare and expects demand to rebound as systems replace aging equipment. For the six months ended December 31, 2025, hospital revenue grew 20.5% Y/Y.

Updated Valuation Drivers

Revenue Build

Not much has changed. I continue to believe consensus is too conservative at 13.3% Y/Y revenue growth in FY26 and 11.5% in FY27. For context, 13.3% in FY26 implies ~11% Y/Y growth in 2H26, which in my view is too pessimistic. With headcount ramping and early benefits from the CRM and e-prescribe platform flowing through, 2H26 growth should exceed that level. Hospital revenue should also begin to recover in 2H26.

For FY27, the law of large numbers is weighing on expectations. The market underestimates the size of the TAM and ELMD’s share gains. 1Q27 will be another education-heavy quarter with conferences and World Bronchiectasis Day. INSM will continue ramping Brinsupri spend, which acts as category marketing for ELMD. With a larger rep base and the operational tailwinds outlined above, I project revenue acceleration in FY27.

The only major adjustment to FY26 is reducing hospital revenue growth from 24% to 20% to reflect the lumpiness seen in 2Q26. I also modestly reduce the other revenue assumption. Overall, I project +14.6% Y/Y growth in FY26 and acceleration to +16.2% Y/Y in FY27.

EBITDA, EPS & FCF Builds

Key changes include raising gross margins to reflect manufacturing efficiency and pricing tailwinds. SG&A is increased to reflect rising headcount, while still expecting leverage. R&D is increased to reflect continued product investment.

FCF is likely to be lumpy this year due to negative working capital changes. Overall, I continue to anticipate operating leverage, with operating income growth comfortably exceeding revenue growth.

Base Case Valuation

I continue to expect ELMD to re-rate toward 20x EBITDA.

ELMD is in a league of its own in micro-cap MedTech. This is a business that should trade at least in line with the broader healthcare equipment market at 17.5x. Given superior margins, growth visibility, and a healthy balance sheet, multiple expansion relative to the market is warranted. At 20x FY27E EBITDA, coupled with stable cash generation and increasing buybacks, I see 82% upside to $44 per share in under two years.

Given the hit to the share price, I see the odds of a buyout increasing, with an acquisition PT on FY26 numbers pointing to 43% upside.

Risks and Bear Case

The core risks from the original thesis and 1Q26 ER are intact. We are now hearing rumors of documentation tightening by commercial insurers, but these are unlikely to have a major impact given how stringent the approval process already is. Risks are inherent in MedTech due to reliance on insurance coverage, and this is a risk I am willing to stomach. In a bear case, I see ELMD trading near ~$21 per share, ~15% downside from current levels.

The thesis is chugging along, and the main valuation driver, operating leverage, is showing no signs of slowing. Multiple expansion is warranted with consistent surprises to consensus. At current levels, P14 holds a 6.4% position at an avg cost of $24.25 per share. I will look to add on further weakness.

Disclaimer: Nothing posted by P14 Capital should be considered financial advice. The author of this post holds a long position in ELMD. Please consult a financial advisor and/or conduct your own due diligence before making investment decisions.

I am admittedly late, as many names are already up >30% since late last year. That said, there is still room to run, with a select group of stocks offering added torque.

Much like my logistics screen, where my targeted list of 5 SMID-caps has already generated an average 11.8% return since the pitch date on 01/18/2026.

The drivers of chemicals stock returns are similar in some respects. At its core, it all comes back to a recovering industrial economy. But chemicals always have more layers. I will walk through the broad thematic setup first before diving into the names positioned for further upside under the paywall.

Cyclicals and market mechanisms

Let’s get the basics out of the way. If there were a textbook definition of cyclicals, chemical companies would fit it perfectly. A key reason the sector has rallied since the new year is this cyclical nature and how it responds to shifts in market flows.

In reflationary environments, cyclicals tend to outperform. Chemical companies are largely price takers, so when indicators like PPI and ISM bottom, that is typically when cyclical industries like chemicals begin to wake up.

Valuation floors

The current awakening has followed a deep and prolonged downturn. Several major chemical companies such as DOW saw their multiples and share prices crushed, with a -41.7% return in 2025. For the industry, 2025 was largely a lost year, with ample supply and historically weak demand compounded by tariff uncertainty.

During sector-specific drawdowns like the one we recently saw in tech, capital often rotates toward deep value areas offering cyclical recovery. Chemicals have emerged as one of those beneficiaries.

Demand drivers

Demand drivers for the chemicals industry are set to return in spectacular fashion. The major demand drivers are the “big two”:

Manufacturing

You have likely seen the headlines. The ISM manufacturing index just posted its strongest reading since August 2022. ISM manufacturing rose to 52.6 from 47.9, backlogs increased for the first time since 2022, exports improved, and inventories declined sharply, signaling sustained demand.

A major contributor here is tariff stability. Much of last year’s weak PMI data stemmed from uncertainty around tariffs. With the year turning over and midterms approaching, markets have largely begun to look through tariff noise. Supply chain managers now have greater visibility than they did in 2025.

Inventories were depleted by the end of the year, so it may be the case that the ISM manufacturing index is “inflated” to start the year. Even so, the outlook for 2026 and 2027 looks materially better than 2025, with reshoring effects increasingly showing up in manufacturing demand.

Rate cuts are also on the horizon, with weak employment data to start the year and inflation appearing relatively stable despite recent hot PPI data . In my view, pricing the incoming Fed chair Kevin Warsh as an outright hawk based solely on historical commentary is a silly assumption. Policy will be dynamic, and political pressure is only rising.

Hot PPIs are generally good for Chemicals, given that they are the de facto raw material in manufacturing and are price takers.

Housing & Construction

Housing is the largest end-market for chemicals (PVC, paints, insulation, adhesives). Forecasts from Fannie Mae and the ACC point to housing starts being relatively flat in 2026 at roughly 1.3M–1.36M units.

Mortgage rates are the key bottleneck for buyers. These are driven by the long end of the curve, which reflects Fed policy as well as expectations for inflation and growth. That said, mortgage rates are also influenced by supply and demand dynamics. The administration has become increasingly supportive of homebuying as a midterm election strategy. Recently, GSEs Fannie Mae and Freddie Mac have been authorized to purchase $200B of MBS. While this represents a fraction of the $9T+ US MBS market, it is a small step in the right direction to stimulate mortgage demand. There are other ways the mortgage market can be stimulated. On this topic, I recommend listening to this brilliant podcast featuring Joseph Wang on Monetary Matters:

Additional speculative upside comes from affordability-focused homebuilding initiatives. Recent reports indicated that builders have floated proposals to construct up to 1M homes, representing >$250B in housing investment.

Other demand drivers

Electrification is another key tailwind. Data center capex is expected to surge again in 2026, with chemical usage benefiting both during construction and in cooling and operations.

Less discussed is the acceleration in vehicle electrification and the resulting demand for batteries. Global EV battery demand is forecast to grow 31.7% Y/Y in 2026 to 1,187 GWh, up from 901 GWh in 2025.

Consumer staples packaging provides an additional defensive demand driver, typically offering a floor even during economic slowdowns.

The weakest end market is agriculture. Elevated fertilizer prices have pressured volumes as farmers tighten budgets amid tariffs and inflation. Rabobank expects 2026 to be another year of subdued consumption, though stabilizing natural gas prices and the ramp of green ammonia projects could support a volume recovery into the 2027 planting season.

Supply side dynamics

This is where stock selection matters most and where you want to distinguish between the good co’s and the bad co’s. A simple and key distinction you can make is specialty versus commodity chemicals.

As with most physical industries, China is always in the picture. Over the past two decades, Europe’s share of global chemical production has dwindled to 13%, while China’s has risen to ~45%. China’s push for self-sufficiency has driven massive capacity expansion. Between 2020 and 2025, global ethylene capacity increased by 40M tons, with ~70% built in China. The result is China exporting excess commodity chemicals such as polypropylene into Southeast Asia, Africa, and Latin America, depressing global prices. While there is speculation that EU plant closures could tighten supply, in my view, this is unlikely to offset the scale of Chinese overcapacity.

We are hearing reports of the Chinese government finally becoming aware of the wrath their investment cycle has bestowed upon the Europoors and the rest of the world. That being said, this has been discussed multiple times, and recently constructed capacity is just going live. So, even if you hold the CCP true to their word, any easing of oversupply will likely occur a few years down the line.

US chemical stocks are the most shielded from Chinese supply, given domestic consumption and production, with trade tariffs offering protection. Even so, as a generalist, I prefer to avoid incremental commodity exposure through chemicals and rather stick to metals.

From a P14 perspective, commodity chemicals are best avoided unless there are idiosyncratic factors that justify upside. These are generally low-margin businesses that exhibit 2x the cyclicality of their specialty peers.

The industry is perpetually restructuring, with ongoing plant shutdowns and consolidation. Japanese ethylene producers and Korean petrochemical firms are now coming under pressure, largely due to the Chinese supply glut. Oversupply is unlikely to abate without a sharp rebound in global demand. That said, with demand recovering at the margin, my view is that investors would likely still do well owning diversified players.

Due to ongoing consolidation and M&A, many players in the industry are heavily leveraged. This creates risk for dividends if investment cycles restart in response to improving demand.

This is why I prefer higher-margin specialty chemical businesses with durable demand and niche positioning. In my view, the true upside lies in smaller specialty players that also offer buyout optionality as consolidation accelerates.

A key supply-side cost driver is rising crude prices alongside lower natural gas prices. In industrial chemistry, the “magic formula” for North American profitability is a wide oil-to-gas price spread. Most NA facilities use ethane as a feedstock, which is derived from natural gas. When gas prices are low, raw material costs are low. In contrast, much of the rest of the world relies on naphtha, which is derived from crude oil. When oil is expensive and gas is cheap, NA producers manufacture the same products as global competitors at a lower cost. Natural gas prices recently experienced a historic speculative rally, but prices have since cooled materially.

The case for flows

2026 is shaping up to be a year for the physical goods economy. We are seeing first-hand that in risk-off environments like this past week, sell-offs in SaaS and services stocks as well as high-momentum, high-beta AI stocks are finding their way into physical industries such as logistics, chemicals, and industrials. This coincides with large-scale fiscal stimulus via OBBBA and improving ISM manufacturing data alongside tariff stability.

This may be the year where big capex spenders are ignored and the beneficiaries come into greater focus, with cash flows taking center stage. This setup especially favors companies that completed their investment cycles during the downturn and can now benefit from fixed-cost leverage as demand recovers.

Valuation context - how high can we go?

You may be wondering, given the double-digit returns many chemical stocks have already posted YTD, whether it is even worth digging through hundreds of companies. Is there really more upside left?

For one, you do not need to. I have already identified 5 chemical co’s that, in my view, are best positioned and still yearning for a rip.

For two, yes, there is still plenty of upside. Similar to logistics, the last true bullish cycle for chemicals was post-COVID. However, multiples from that period are not the right reference point given the severity of the drawdown and the violence of the recovery. Beyond that distortion, the 2016–2018 period stands out as the last “normalized” bull market for chemicals.

The 2016-2018 cycle is widely viewed as the last healthy, pre-pandemic expansion, characterized by synchronized global growth. The US, Europe, and China all experienced expanding industrial production. NA producers benefited from the shale advantage, as new ethane crackers came online at a time when US feedstock costs were low and global chemical prices remained high, driving record margins. At the same time, strong demand for plastics and specialty chemicals tied to autos and construction outpaced supply additions, creating tight markets.

Given their cyclical nature, chemical stocks are typically valued on a forward basis. As a rule of thumb, cyclical stocks tend to price in a recovery 4–6 quarters ahead of the actual improvement. That is why we are seeing stocks move now. Specialty chemicals are slightly different in that their demand is more durable depending on end markets, but they are still cyclical.

Since 2022, we have effectively been in an industrial recession. With rate cuts on the horizon, deregulation, fiscal stimulus through OBBBA, and tariffs driving industrial reshoring whether one likes it or not, chemical stocks and other physical goods stocks are signaling that a recovery is coming. In my view, the recovery begins in 2H26 as rate cuts start to impact mortgage affordability and homebuying. The recovery will be stronger in 2027 as supply and end-markets demand stabilizes.

Back to multiples. During the 2016–2018 cycle, XLB peaked on 01/15/2018 at $31.64 before falling to $25.64 by year-end amid trade war 1.0.

At the peak, chemical stocks were trading at a meaningful premium on TTM P/E multiples relative to today. On a TTM basis, the potential expansion is ~9.9x. On a forward P/E basis, the implied expansion is lower at ~1.4x, largely because earnings revisions have not yet been priced in. We are still early in the cycle, and as the year progresses, earnings optimism should increase.

Within this, specialty chemicals stand out. The group currently trades at the steepest forward multiple discount versus 2018, at ~10.1x below prior cycle levels. On a TTM basis, multiples are ~6.1x lower.

All of this suggests that if the industrial economy does return in earnest, there is still significant room to run. You can see this simply by looking at the share prices of the largest beaten-down chemical stocks.

Which brings us to why you are here. Below are P14’s top specialty (4) and commodity chemical (1) picks. As usual, these are SMID-caps (avg mkt cap: $6.3B). They offer solid fundamentals and idiosyncratic growth catalysts in the specific end markets they serve. The SMID-cap exposure is also accretive in a cutting cycle, as the group historically outperforms during rate-cut environments.

Disclaimer: Nothing posted by P14 Capital should be considered financial advice. Please consult a financial advisor and/or conduct your own due diligence before making investment decisions.

Note: The watchlist is not ordered by top picks. Most stocks trade at reasonable forward valuations relative to larger peers. This is what I’d call a starting point. After doing more work, I will look to initiate where I find appropriate.

Q - Qnity Electronics, Inc.

Price: $98.22 | Market Cap: $20.6B | FY26E Rev Growth: 7.3% | EV/FY26E Sales: 4.5x | EV/FY26E EBITDA: 14.7x | FY26E EBITDA Margin: 30.4% | Debt/Equity: 18.5% | YTD return: 20.3%

Qnity Electronics is a pure-play electronic materials company supplying consumable chemicals used across semiconductor fabrication, advanced packaging, PCB build, and assembly. ~90% of revenue is derived from unit-based consumables tied to wafer and substrate volumes. Semiconductor Technologies accounts for ~57% of FY24 sales, driven by CMP pads and slurries where Q holds leading share. Interconnect Solutions contributes ~43% through advanced packaging, substrates, and thermal materials. These products are process-critical and embedded deeply in customer workflows, supporting pricing stability and high switching costs. Management guides to ~30% FY25E adjusted EBITDA margins and ~$1.4B in FY25E EBITDA based on pro forma metrics, reflecting strong operating leverage and durable cash generation.

Qnity’s portfolio is aligned with AI and semiconductor demand as materials content per wafer increases with node complexity and packaging intensity. AI and high-performance compute represent ~15% of sales across fabrication and packaging. Advanced packaging is approaching ~10% of revenue and continues to grow faster than the broader electronic materials market. CMP intensity, process steps, and packaging layer counts are rising across leading-edge logic and memory nodes, directly benefiting Q’s consumables mix. Revenue exposure is driven by wafer starts and board volumes, extending demand duration through the cycle. Management targets revenue growth ~200 bps above industry averages through share gains in fragmented markets and increasing exposure to advanced nodes, memory, and packaging.

While on a P/E basis the multiple might screen as expensive (~32x), Q should be looked at on an EBITDA basis because of its higher tax rate. Key peer Entegris trades at 23x EBITDA (vs 15x Q) for seemingly lower scale, lower margins and lower organic growth. In my view, the discount likely stems from the fact that Q is a recent spinoff that has “something to prove”. With DC capex only set to rise, a few more quarters of outperformance should broaden the investor base. As ownership shifts toward semiconductor-focused investors and earnings scale with AI-driven volume growth, valuation should move closer to comps.

ESI - Element Solutions, Inc.

Price: $29.94 | Market Cap: $7.2B | FY26E Rev Growth: 17.2% | EV/FY26E Sales: 2.8x | EV/FY26E EBITDA: 12.9x | FY26E EBITDA Margin: 22% | FY26E EPS growth: 16.7% | P/FY26E: 17.3x | Debt/Equity: 60.9% | YTD return: 19.8%

Element Solutions supplies specialty process chemistries used in electronics manufacturing, with earnings concentrated in its Electronics segment, which represents ~70% of revenue and EBITDA. The business sells consumable chemistries and surface treatments embedded in customer production processes across circuitry, assembly, and semiconductor packaging. In Q3 2025, Electronics delivered $465M of revenue with 7% organic growth, driven by Circuitry up 13%, Assembly up 5%, and Semiconductor up 5%. These products sit inside high-frequency production steps, supporting recurring demand tied to unit volumes across PCBs, semiconductors, and advanced packaging. FY25 guidance was raised to ~$545–550M of adjusted EBITDA and ~$1.45 of adjusted EPS, reflecting continued strength in Electronics execution.

ESI fits the physical economy and electronics recovery thesis through improving demand across both secular and cyclical end markets. Electronics exposure to AI and data centers is ~10–15% of segment revenue and is growing at >20%, supported by high-performance compute, advanced packaging, and data center infrastructure builds. The remainder of the electronics portfolio is leveraged to semis and PCBs tied to tablets, smartphones, laptops, and desktops. That demand base has been working through excess inventory from the work-from-home cycle and is positioned for normalization into 2026 as device refresh activity resumes. Management commentary points to improving PCB volumes and better utilization across China and Asia as contributors to the next phase of growth.

M&A is central to ESI’s strategy and is increasingly focused on electronic performance materials with higher margin and content density. The pending Micromax acquisition adds advanced circuitry exposure and is expected to contribute ~$40M of adjusted EBITDA on a full-year basis, with adjusted EPS accretion >5%. The EFC Gases transaction expands electronics-adjacent capabilities and deepens exposure to semiconductor and advanced manufacturing customers. Combined, these deals lift the Electronics mix, raise the earnings base into FY26, and support a path toward ~$660–670M of adjusted EBITDA. Alongside ongoing share repurchases, this supports a multi-year setup for top-line and EPS growth driven by electronics recovery and disciplined capital allocation. Multiple expansion is warranted as at 17x forward EPS, even minimal expansion towards the LTM 30x EPS suggests significant upside potential.

KOP - Koppers Holdings, Inc.

Price: $32.30 | Market Cap: $633.5M | FY26E Rev Growth: 2% | EV/FY26E Sales: 0.8x | EV/FY26E EBITDA: 6x | FY26E EBITDA Margin: 14% | FY26E EPS growth: 18% | P/FY26E: 6.8x | Debt/Equity: 188% | YTD return: 19.3%

Koppers is an integrated infrastructure materials company serving rail, utilities, and industrial end markets through 3 segments. RUPS supplies treated railroad ties, utility poles, and related services. Performance Chemicals sells wood preservation chemicals used in residential, industrial, and agricultural applications. Carbon Materials and Chemicals distills coal tar into carbon pitch for aluminum anodes and produces creosote and carbon black feedstock, with internal supply into KOP’s treated wood system. This vertical integration allows KOP to capture value across the full lifecycle of infrastructure wood products. In FY24, RUPS generated ~$943M of revenue, Performance Chemicals ~$652M, and CMC ~$498M, anchoring cash flow in replacement-driven infrastructure demand.

KOP fits the physical economy thesis through end markets tied to replacement cycles and deferred maintenance. Rail and utility infrastructure replacement is structurally required, with ~18–22M railroad ties and ~2–2.5M utility poles replaced annually in North America. Near-term results reflect cyclical pressure, but demand indicators are stabilizing as customer inventories normalize. Management has exited or completed capital-intensive projects, including the Railroad Structures divestiture and the closure of phthalic anhydride production, materially reducing future capex requirements. Maintenance capex is trending toward ~$50–60M annually, supporting higher free cash flow as volumes recover into 2026, with incremental contribution from utility poles, commercial ties, and aluminum-related carbon pitch demand.

Performance Chemicals has seen recent market share loss following management’s decision to protect pricing and margins rather than pursue lower-quality volume. In Q3 2025, Performance Chemicals revenue declined 18.3% Y/Y and EBITDA margin compressed to ~18%, reflecting lower volumes and higher input costs. Management disclosed residential volumes down ~5% and industrial volumes up ~2.5%, indicating stabilization within core end markets. Cost actions are in place and volume recovery is expected to drive operating leverage. With leverage manageable, debt maturities in 2030, and valuation at ~6x FY26E EBITDA and ~6.8x P/FY26E vs the industry average of ~14x, KOP offers a near-term earnings recovery and longer-term upside from improved free cash flow, capital discipline, and potential buyout interest at deep value multiples.

MATV - Mativ Holdings, Inc.

Price: $14.55 | Market Cap: $795.6M | FY26E Rev Growth: 2.2% | EV/FY26E Sales: 0.8x | EV/FY26E EBITDA: 7.3x | FY26E EBITDA Margin: 12% | FY26E EPS growth: 54.8% | P/FY26E: 13.9x | Debt/Equity: 275% | YTD return: 19.7% | EBITDA Interest coverage: 2.6x

Mativ Holdings is a specialty materials company formed from the 2022 merger of SWM and Neenah, supplying engineered materials embedded across industrial manufacturing supply chains. The company operates through 2 segments. Filtration & Advanced Materials sells filtration media, engineered films, netting, and membrane supports used in industrial process filtration, HVAC, water treatment, transportation, and medical applications. Sustainable & Adhesive Solutions sells tapes, release liners, labels, cable wraps, packaging substrates, and healthcare materials used in industrial, consumer, and construction workflows. Revenue is volume-driven and tied to production activity, with products positioned early in customer manufacturing cycles. On a TTM basis, MATV generated ~$2.0B of revenue and ~$210M of adjusted EBITDA, with SAS representing ~62% of sales and ~55% of EBITDA, and FAM ~38% of sales and ~45% of EBITDA.

Recent results show improving industrial conditions. In Q3 2025, consolidated sales grew 5% organically Y/Y and adjusted EBITDA increased 10% to $66.8M, marking one of the strongest quarters since the merger. FAM returned to growth for the first time post-merger, with sales up ~4%, driven by filtration demand across HVAC, water, air pollution control, transportation, and medical applications. SAS delivered ~5% organic growth with strength in tapes, liners, healthcare, and construction-linked products. Management commentary points to continued pricing discipline, improving volume mix, and lower manufacturing costs supporting incremental margin expansion into 2026.

The earnings setup is driven by operating leverage, portfolio focus, and declining capital intensity. Since the merger, MATV has realized $65M+ of cost synergies, reduced facilities from 48 to 35, exited engineered papers and tobacco exposure, and concentrated capital toward higher-margin growth categories such as filtration, specialty tapes, release liners, and medical films. Growth projects coming online through 2026 add ~$115M of incremental revenue capacity across filtration, liners, tapes, and medical films. Maintenance capex and working capital requirements are trending lower, supporting improved free cash flow and deleveraging ahead of debt maturities beginning in 2027.

MTX - Minerals Technologies, Inc.

Price: $72.77 | Market Cap: $2.3B | FY26E Rev Growth: 4.6% | EV/FY26E Sales: 1.4x | EV/FY26E EBITDA: 7.3x | FY26E EBITDA Margin: 18.4% | FY26E EPS growth: 12.5% | P/FY26E: 11.7x | Debt/Equity: 55% | YTD return: 19.4%

Minerals Technologies is a vertically integrated specialty minerals company that owns mineral reserves and processes them into functional additives and process technologies sold into industrial and construction end markets. The business operates through 2 core segments. Engineered Solutions serves steel, foundry, automotive, infrastructure, and environmental applications through refractories, metalcasting binders, and on-site services. Specialty Additives supplies ground and precipitated calcium carbonate, talc, and bentonite used in paints, coatings, drywall, roofing, flooring, paper, and packaging. The model combines owned mineral resources, processing assets, and customer-embedded operations, with revenue tied to production volumes rather than discretionary spending.

The earnings setup is driven by normalization across housing and industrial production. Specialty Additives has been pressured by elevated rates slowing residential construction, which reduced volumes in paints, coatings, joint compounds, and roofing applications. That pressure is easing as customer ordering activity stabilizes and inventory levels normalize, setting up improved volume absorption as housing activity improves into 2026. On the industrial side, Engineered Solutions is more stable, supported by steel production, foundry activity, automotive components, and infrastructure maintenance. MTX benefits directly from higher industrial utilization through refractory consumption, metalcasting binders, and process additives that scale with throughput.

Operating leverage is the key driver of upside. MTX owns mines and operates satellite plants co-located at customer sites, so incremental volume flows through at higher margins once fixed costs are covered. Capital projects are largely complete, with recent investments in process technology and environmental applications moving from build to utilization. That shift improves free cash flow and earnings visibility over the next several years. Optionality from a more protectionist materials environment exists through domestic mineral ownership and regulated infrastructure exposure, though MTX is not positioned as a direct rare earths producer. At ~7x FY26E EBITDA and ~12x earnings, the stock reflects depressed end-market conditions that are already improving, with earnings leverage skewed to the upside as housing and industrial demand recover.