I have been following PSIX for a few years. In fact, I first owned the stock when it was on the pink sheets. Over the years, I’ve visited the company’s two manufacturing facilities in Wisconsin and spent a bit of time with management.

I am no stranger to the peculiarities of microcap management teams but my interactions with PSIX were really quite.. something. Anyone who has interracted with them knows what I am talking about. Unfortunately, most investors haven’t had the chance as they don’t answer the phone.

Largely for this reason, as well as lack of disclosure, we never quite got comfortable enough with the company for it to be a big position. My view was largely that the company was uninvestible whilst Dino was the CEO. I like Dino as a person but he was not fit to be a public market CEO. I had always considered that his exit would be a good opportunity to revisit the stock.

Well, Dino is now gone. Recently, the company announced Richard Hu as CEO. Richard Hu has quite an intriguing background. He was most recently running the Americas business for Borg Warner. For those who aren’t familiar with Borg Warner, it is a Tier 1 Automotive supplier.

In the Automotive Supply business, operations is the name of the game. It is a cost-driven business with constant, brutal pricing pressure. Operating efficiency and quality control are life and death for auto suppliers. Richard strikes me as a good appointment for a simple reason. PSIX, by all accounts, doesn’t have a demand problem. Its problems have all been operational. I am optimistic that Richard can bring some Fortune 500 operating rigour to the company.

The America’s unit rose margins over a period of declining revenues during his tenure. It’s a stretch to attribute that to him, but it is interesting insofar as that would generally only be achievable for a company with a serious operating efficiency culture.

This set-up is a particularly well-known tale for smallcap investors: A new CEO, who was previously running a regional business unit at a large industrial company comes to a small company and makes it a serious company.

What remains to be seen is whether or not PSIX gets serious about investor engagement. The company’s investor comms are horrific. However, maybe this is changing. They have atleast strarted doing sellside earnings calls (if you could call them that.) In June, they were hiring for an IR person. One would hope that an announcement is around the corner. And you’d have to hope that the new CEO cares about getting the stock up, which shouldn’t be a stretch.

What does the company need? Well, if Richard or the new IR person happen to read this, this is what you need:

1. A CEO who cares about educating the street on their opqaue business and an IR person capable of dispelling innacurate short reports/conspiracy theories about Chinese connections.

2. An investor day. Discuss the enclosures business and talk about your strategy in the gas engines business.

3. Disclosure about the size of the data center business relative to oil and gas. Disclose it going forward so that people make proper forecasts.

4. Proper guidance and proper communication with the street so that estimates are accurate.

Now let’s talk about recent results. Q2 results were a huge step in the right direction. Operating issues seem to have been somewhat dealt with, and the 27% gross margin proves that the company is capable of making serious profit. Commentary on 2H sales recovery was comforting. Management walked the street back from expecting 27% every quarter, but they did say that 25% should be achievable, which is still way better than 21-22% than investors seemed to be forecasting when the stock was $28.

So, the operating issues are hopefully past the worst and revenues are recovering. 2027 looks set to be a strong growth year. I do believe that the company has significant demand and that their issues are operational, not demand-led. I could be wrong on that, but have no reason not to believe it is true.

I also think that the opportunity on the engine side could be significant, but it’s just too opaque to make any serious forecasts about right now.

The other big positive is that the balance sheet is in a much better place today. By the end of the year, the company should be firmly net cash.

At a 25% gross margin, I estimate that the company has ~$80m of earnings power of the current run-rate sales, putting the stock at ~10x P/E.

If it can return to growth in 2027 and get the gross margins back into the high-20s with some serious operating improvement, it’s not hard to see how we could get to $100m+ in FCF for 2027.

What multiple should the stock trade at? At one point, investors put PSIX at 30x P/E. If they’re able to grow-up and become a serious company with serious IR, investor comms, guidance and disclosure and prove that they can grow again, I don’t see why it couldn’t get back to 20x.

So, I see an upside case for $100+ for the stock if things start to go right. The good news is that 2Q suggests things are already starting to go right. It’s hard to see all too much downside from here given the starting valuation and the clean B/S. Of course things can always go wrong - maybe we go back to the mid-$20s if things go haywire. Down 30% doesn’t seem to bad if we’re playing for a 2-3x.

If you think there’s a 50% chance that they get their act together, then it seems like a good bet. I think there’s a >50% chance so I own the stock.

Disclaimer

Not investment advice, do your own DD, I could own/buy/sell stock or other financial instruments at any time, etc.

This post was written in Late February - I never posted it due to poor trading liquidity. Two quarters have since passed - I haven’t updated the post. 1Q results today were confirming and the trading liquidity has improved, so I am posting it. I am still quite bullish at $12.

I recently bought a small position in LEAT. The name is probably familiar to MicroCap investors from the 2020/21 era. Liquidity is poor and it is OTC listed so it will remain small. I don’t think that much needs to happen to retrace its $30 highs. The 3Q earnings miss seems to have been a timing issue and I expect that 4Q25 will be strong with momentum continuing into 2026.

I have not been able to speak with management so I am a little in the dark on this name. Given the small position size I’m okay with that. My conviction isn’t super high - but with a strong balance sheet and some business momentum, I particularly like the risk/reward. If the stock works, I think I can make 3-5x. If not, I don’t think that I will lose much.

Brief Overview and History

Leatt is a Motocross (MOTO) and Mountainbiking (MTB) gear manufacturer based in South Africa and the US. It has quite an interesting history. Founded in 2001, the company invented the first “Neck Brace” for motocross riders. The product went viral and Leatt sprung to global recognition in a very short period of time.

By 2010, the company had >$10m in revenues in Neck Braces, but Dr Leatt saw that other parts of the safety apparatus for Motocross were also lacking. Leveraging Leatt’s brand recognition for safety and innovation, he saw the opportunity to carve out a niche of selling premium, highly-engineered products for the safety-concious. The company’s motto was: “the business of selling safety.”

Since 2010, the company iteratively released body armour, knee braces, helmets, gloves, goggles and apparel across MOTO and MTB to become a full-body protection company. Leatt always brought products to market with a new innovation, such as ADF foams used in body armour (additional lightweight and comfort without compromising protection) and the C-Frame structure of their knee braces (less-bulky kit without compromising structural integrity).

What’s most intriguing about Leatt is that they seem to have built a repeatable muscle for R&D-driven, product-led-growth. The Leatt Lab in Cape Town has consistently created new, differentiated products that have driven Leatt from a single product into a full body protection company. The company grew from ~$14m of revenue in 2010 (all Neck Braces) to nearly $40m in 2020. All the while, Neck Braces shunk to just $5m as their early monopoly ultimately attracted competition. Today, Neck Braces are just ~5% of the company’s revenue. It’s quite an impressive track-record of product-led growth arcoss multiple categories.

Like most outdoors-exposed consumer companies, Leatt took a royal beating in the post-covid slowdown. Inventory gluts, distributor bankruptcies and aggresive discounting through 23/24 added insult to injury. All the while, Leatt continued to invest in the business, increasing its operating expenses as top-line capitulated. Unsurprisingly, the company went into the red and the stock is down 70% from its highs. It has largely been left for dead with dismal trading liquidity - a forgotten hero of the covid-era 10x club.

However, after studying the company somewhat, I believe that Leatt has a very bright future. Not only have end-markets bottomed and margins recovered, but multiple growth drivers put into place by management over the past few years appear to be reaping rewards. My industry work also suggests that the brand’s fundamentals have only strengthened throughout the downturn. They also seem to be on the cusp of a new growth driver from Adventure Riding with extremely promising early signs.

Below are the key points of my thesis.

1. The company remains on sound fundamental footing. Cyclical recovery is underway

My main goal was to determine if Leatt’s revenue decline has been purely cyclical or if the brand has suffered damange. I visited a handful of retail outlets in the US and Australia to chat with retailers about the company. None had anything but positive things to say about Leatt, chalking up their weak revenues to an industry downturn that has hit everyone.

Interestingly, Racer X magazine (large Motorcross publication) runs an annual survey of readers asking which gear they own. Between 2021 and 2024, Leatt grew its share in every single category that it was represented in the survey.

Leatt’s gross margins have historically been strong in the mid-40s, but took a hit during FY22-24 through a wave of excessive discounting to clear stock in a flooded market. This called in to question if the company’s historic pricing power would still hold. Encouragingly, gross margins have also shown a strong march back towards the mid-40s as the inventory clearance discounting has abated and fresh products are working their way through the channel.

Online reviews of the brand also remain extremely positive and they continued to win product awards at the coveted EuroBike conference in 2025. All of this is to say that there doesn’t appear to be any structural impairment of Leatt’s brand - an encouraging sign.

TTM reveues (better as sales can be lumpy due to seasonality and sell-in/out dynamics) for the US and International have both clearly bottomed and are on a recovery path, albiet still at emperically low levels as the consumer remains weak. The company appears well positioned to continue to benefit from a recovery in demand.

2. Increased spending appears to be reaping rewards

From FY21 to FY24, the company’s revenue fell ~40% whilst R&D and SG&A spending both increased ~40%. The investment has largely gone towards bolstering the sales team in the US, increasing D2C presence and fuelling the ADV product expansion. There are clear signs of these investments bearing fruit.

Firstly, sales in the US began to recover sharply in 2025. 2Q25 likely benefitted from a tarrif pull-in, however the trend on a TTM basis is clear. Leatt has installed an almost entirely new sales/marketing team in the US over the last ~12months and the investment appears to be working.

The D2C strategy is also bearing fruit for Leat. Whilst still small, D2C revenues have more than doubled over the last 5 years whilst group revenue has fallen meaningfully. It has increased from ~5% to nearly 10% of revenue. Whilst still small, it is growing fast and as a size now that it can meaningfully move the dial for the company. D2C also comes it at appreciably higher margins.

The third large investment from Leatt is the ADV range, which deserves its own point entirely.

3. The new ADV range is Leatt’s fastest-ever growing product line

Leatt introduced a new range of products for the Adventure Riding market in late 2023, hiting dealers in early 2024. In typical Leatt fashion, the company brought a new “safety focussed” approach to the adventure riding market with the same level of comfort required for long range rides. The intial range was limited, including just gloves, boots and apparel. The initial launch was a huge success, and by 3Q24, ADV already accounted for ~8-10% of Leatt’s revenue.

In late 2024, the company expanded the range further to include helmets, protective gear and boots. By mid-2025, the ADV line had already increased to ~15-20% of group revenues, approaching ~$10m in annualized revenue after just 18months. In early 2026, the company announced a large expansion to its existing line including new product tiers and technical upgrades to some cateogries. With product momentum continuing, I would expect that growth should continue strongly into 2026.

Even today, the line-up is quite immature with a small number of SKUs compared to competitors, suggesting that there’s more growth in the existing line-up to come. It is also not available worldwide yet - I contacted a few local Adventure stores in Australia and they mentioned that the gear is not yet available but coming soon.

I also polled some folks on the ADV Rider reddit. Feedback on the products was unanimously positive albiet admittedly anecdotal.

I also spoke with a former employee who had this to say:

ADV is a massive market and they have started the right way. Jackets, pants and helmets are stacked with value and designed in a different way setting them apart from most competitors. A big challenge and opportunity is going to be marketing. Moto and MTB have racing, ADV is more about community. So it means connecting to your customers in different ways than they would be used to.

They have all the basics in place, including personalities. What I have always felt is that ADV requires more unconventional marketing, like finding a place in riding clubs, events and chat boards. Maybe even hosting events or riding workshops as a marketing tool. Leatt finally has a new marketing director and is busy changing up how marketing is done, so these unconventional methods might just be around the corner.

Leatt’s ADV products appear to be resonating well with customers. If the momentum continues, ADV could be a very meaningful contributor to growth over the coming 2-5 years having gone from 0% to 20% of revenue in the short span of 2 years.

ADV’s meaningful revenue contributions over the last 2 years also suggests that traditional MOTO and MTB end-markets are still quite a bit more depressed than they might appear, given the still weak consumer in Europe and the US. True MTB and MOTO demand is still probably pretty close to trough levels, albeit the major destock is past.

4. The company should experience meaningful operating leverage as revenues grow

As noted, Leatt has continued to invest in SG&A through the downturn which has depressed their earnings. However, the cost-base is now in place to scale the business up to $70-80m in revenue. The CEO outlined this on the 1Q25 earnings call:

“We can get to $70 million to $80 million of sales without having significant increases in our OpEx. So we’re setting ourselves up, obviously, for future growth.

In terms of our sales team, we’ve actually been working really hard in the U.S. to make sure that we are a lot more efficient in terms of sales structuring there. We bought Rob Ramlose on as a VP of Sales on the MOTO side, working with the brick-and-mortar dealers and the brick-and-mortar reps. And I think that’s going to add a lot of efficiency in terms of our sales.

So I mean, to answer your question, the kind of OpEx that we’re sitting on now, $20 million to $21 million to $22 million, I think we can still leverage that significantly before needing to go to any kind of much higher level. So we are building a base now really for future growth.”

Opex is currently run-rating at ~$21-22m/year and revenue for FY25 will land in the ~$62m-ish range. If the company can return to $80m in revenue at 45% gross margins on a fixed OPEX base of $22m, the company should be able to generate ~$14m of operating profit or ~$11m of net profit.

This is ~$1.75/share of EPS power on a current stock price of $9 with $2/share of cash on the balance sheet. With an active repuchase program, I have no problem crediting the company for cash on its balance sheet. This puts the stock at an ex-cash P/E of ~4x, or ~3x EV/EBIT. Even if it re-rates to 10x P/E, it’s at least a $20 stock from here.

This doesn’t seem like such a stretch. $80m in revenue is only ~30% higher than the FY25 level and is still well below the $100m peak that the company achieved during Covid. In FY25, they will register ~40% revenue growth over FY24, albeit off quite a low level. It’s not hard to imagine how a little recovery in the consumer and continuing momentum of the ADV line could get the company to $80m in revenue in FY26 or FY27.

For the dreamers: If we really get a cyclical recovery and the ADV lines strong growth continues, it’s not unreasonable to think that Leatt could re-trace and surpass the $100m revenue milestone. If we assume a ~$25m OPEX base at this revenue line, Leatt could generate ~$2.50 of EPS. For a growing, niche consumer business like this, a 15x P/E multiple is probably not outside of the realm of possibility, which would put the stock at ~$40 - or a 4x from here. We’d probably see a Nasdaq listing as well. You really don’t have to believe that much to see a >4x in LEAT.

If I’m wrong and growth stalls out, it’s hard to see too much downside. The stock is currently trading at ~12x FY25 P/E (trough levels) with >20% of its market cap in cash and 40% revenue growth this year. I’d imagine that at worst the stock is dead money, unless we have a severe recession which pushes the company back into the red, at which point it could probably head back to the $6/7 dollar range that it hit back in 2024.

Summary

The stock just isn’t liquid enough to warrant spending time/resources on boots-on-the-ground research. If I could get in touch with management I would perhaps feel comfortable having a bit of a bigger position. However the position will remain small and I look forward to following the company’s progress in the coming quarters.

Brief Overview of Investment Thesis

Guardian Pharmacy Services is a Long-Term Care Pharmacy. There’s an LTC Pharmacy 101 section in the appendix (bottom of page) if you don’t know what that is. The key points of the thesis are as follows:

1. LTC Pharmacy is an attractive industry to be in (which I will describe at length below). It’s sticky, defensive, capital light and has high barriers to entry.

2. LTC Pharmancy benefits from a robust multi-decade aging-population demographic tailwind.

3. Guradian has a secular market share gain tailwind from (1) small mom + pops who are increasingly struggling as ALF industry consolidates, and (2) Omnicare who is slowly dying.

4. Fred Burke (CEO) is a killer. He’s a multi-time entreprenuer (with exits to DnB and Cardinal Health) and has constructed a unique business model and culture.

5. Guardian has a tried-and-tested growth model that combines inorganic + organic growth. As the leader in LTC ALF with just ~13-15% market share, there’s a significant runway to continue growing. Guardian has opened or bought 50+ facilities with remarkable success and has nailed down the playbook. I expect that they could eventually reach 100-200 locations.

6. Both organic and inorganic ROIC are phenomenal. The company will generate >$120m of EBITDA this year on a net PP&E of just $55m. Their M&A targets are typically break-even or losing money, but amount to ~3-4x multiples after group profitability is achieved (within 3 years).

The growth algorithm is roughly as follows:

• HSD% organic growth

  • The industry grows at ~4-5% (CAGR of the ~80+ demographic population through to 2040).

  • Guardian takes share from Mom + Pops and Omnicare at a ~LSD% rate. They’re currently growing at low-teens but that won’t sustain forever.

• ~MSD-HSD% inorganic growth

  • Ideally, the company would do more M&A but is limited by number of available deals. ~4-7% growth/year would be doable and still only consume ~half of their FCF.

• Margin improvement through scale and improving operations

  • Currently, ~20% of the company’s pharmacies are recently purchased or open and only at break-even margins, dragging the company’s margins by ~90bp (on 8.6% EBITDA margin).

  • Over the long-term, margins should converge closer to 10%. However, as the company continues to acquire and open new facilities, the short-term PnL drag will continue. But true earnings power will grow.

• Hence, the company should grow EBITDA at mid-to-high teens over time and EPS at 20-25%+ (leverage over D&A and excess cash to buybacks).

There’s also an extra, significant option on growth. The company’s largest competitor - Omnicare (owned by CVS) - declared bankruptcy in late 2025 after years of CVS trying to flog it off. Omnicare dominates in SNF but also has a sizeable ALF business. Any break-up or liquidation of the company could have significant benefits for Guardian through (1) ability to acquire facilities cheaply, or (2) oragnic growth from closed stores.

Stalking Horse bids are due at the end of March so I will update on X as it goes along. Omnicare as a company is so busted based on all my conversations with customers and competitors that I’d imagine the only way to fix it would be very significant restructuring, which I’d imagine would have to include closing/selling parts of the business. I expect that the conclusion of the Omnicare bankruptcy could accelerate GRDN’s organic and/or inorganic growth. GRDN is interested in buying some facilities but only those that serve ALF (most of Omnicare’s business is actually serving SNF where they have 40% share).

At ~25x FY26e P/E (at $31) (with a meaningful PnL drag from acquired facilities), I’d expect that investors should experience a TSR roughly equal to EPS growth of 20-25% over a 5-10 year period. I don’t include model/forecast/numbers in this post.

What don’t people get about GRDN?

There are 2 things in my mind.

Firstly, until recently, the IRA changes to drug pricing under Part D have been a big overhang on the stock. Throughout 2025, Fred has been assuring the market that they’ll offset the headwind and in January he guided to that, before increasing guidance just 1.5months later on the 4Q call last week.

Secondly, people seem to underappreciate how difficult it is to get into and succeed in this business. Guardian’s success is largely owing to its attracting and retaining of high-quality operators as well as scale. I’ll get into some of this below. I’d say that a business like this which has an unfamiliar business model needs to show track-record until investors are willing to pay up. I think we’re halfway through this recognition process. Management also needs to build credibility, which they have been with consistent beat/raises and resolution of the IRA issue.

Thirdly, the stock is small and until recently was only trading ~$2m/day when we first started buying it. There’s barely any analyst coverage.

Lastly, some people were worried about an overhang from secondary. I had strong reason to believe (based on conversations with other investors) that any secondary would be well absorbed. It seems that some folks holding out for a secondary missed out and are now chasing the stock higher.

Brief overview and history of GRDN

Guardian Pharmacy was founded in 2004 by Fred Burke, Kendall Forbes, David Morris and Bill Bindley. The first 3 had previously founded a wholesale distribution business (focused on nuclear drugs), which was sold to Bindley Western, which was sold to Cardinal Health. Prior to that, Fred Burke had founded Sales Technologies, inc. – a salesforce automation tool that was sold to Dun & Bradstreet in 1989. This is the team’s second succesful pharmacy act and the playbook hasn’t changed.

Guardian started with a single pharmacy in Phoenix that was run by a guy called John Saliba. John Saliba started in the assisted-living industry in the 80s/90s and later started his own LTC pharmacy in 1999 after having great frustration with how drugs were handled withing ALFs and seeing an opportunity. John was having success but he was struggling to make money and got connected with Fred and the team (I think through Bill Bindley). Guardian bought 80% of John Saliba’s business back in 2004 and partnered with John to open more pharmacies, giving birth to Guardian. From here, they began building an optimizing a blueprint for the LTC facility.

From day 1, they focused intently on being tech-forward and building scale in an industry largely reliant on faxes and manual processes. Bill saw an opportunity to combine Fred and team’s operating expertise in Pharmacy with his expertise in drug wholesaling. Together, they could build a scaled platform with drug purchasing expertise.

At this time, Omnicare was a dominant force in the LTC Pharmacy industry, though they were mostly focused on SNFs. There are a few reasons for that. Firstly, the ALF industry was largely in its infancy in the 1990s and it looked very different today. Back then, it was mostly just converted apartments for old people with food/cleaning services and the operators were mostly real estate people. The scale and sophistication were much smaller. Secondly, Part D didn’t exist until 2006 and utilization of prescription drugs was much lower back then. As a result, re-imbursement was much less certain. It didn’t appear to be quite an attractive industry for Omnicare, who was dominating (and still does) in SNFs.

Guardian was somewhat in the right place at the right time back in 2004. They rode the ALF boom in the early 2000s/2010s and were perfectly situated to benefit from the explosion in drug utilization within ALFs as Part D was introduced, as well as the rapid growth of the ALF industry. As ALFs became more sophisticated, LTC pharmacy services became more critical.

Guardian expanded nationally both organically and via acquisition. Early The first few facilities were organic, riding the ALF/Part D boom in the early 00s, until the company had its process honed. Then it began to start purchasing financially struggling LTC facilities with existing customer bases in new markets under the assumption that they could implement their own processes (which included tech/automation – it was novel at the time) and improve profitability. This worked, and the company has also built out a muscle for M&A. ~2/3rds of their facilities have been acquired to date.

It’s worth mentioning GRDN’s unique ownership model, which is key to its success and was also implemented at Fred’s former pharmacy venture. In organic or M&A, GRDN seeks to partner with a local president who’s truly in charge of their facility. Historically, the owner would own 20% of the business and GRDN would own 80%, providing scale, back-office operations and other expertise. The pharmacy industry is unqiue in that many of the small pharmacy owners don’t really care about the finance/business side and just want to focus on running their pharmacy.

When the company went public, GRDN collapsed the 20-80 structure but it still implements 80-20 by using its public stock and RSUs tied to the PNL of the local pharmacy. It doesn’t just buy any facility - the company is highly-selective in who it chooses to partner with. I’ll discuss the success of this model more in detail later.

Why is LTC Pharmacy a good business to be in?

Let’s discuss some of the key points.

1. Sticky, economically insensitive customer base

LTC Pharmacy is, at its core, a services business. The service that Guardian provides to assisted living facilities (ALFs) is mission critical. Patients need to get their drugs on time and without error to avoid hospitalization. Facilities care deeply about avoiding hospitalizations – empty beds = lost dollars, and hospitalizations can mean fines/reputational damage. The administration of drugs at an ALF (the “med pass”) is a logistically complex process. It includes not just the actual passing of drugs but also the records keeping, claim adjudication, etc. As such, close co-ordination with the pharmacy provider is required.

Guardian’s service is deeply embedded into a facility’s workflow which creates switching costs for the facility. Significant process, software and relationship integrations exist. Facility staff are trained on the systems and processes which is labor/time consuming and often leads to high errors until processes are nailed down. Given labor is already a big issue for ALFs, this is a process that they’d rather avoid unless strictly necessary. Customers that we heard from discussed that they “would never switch pharmacies, unless there’s an absolute need.”

The most interesting and underappreciated characteristic of the business is that although the facilities have decision-making responsibility over which pharmacy to adopt, they largely don’t pay for the service (other than a small fee for training) – Pharmacies make their money through drug sales, which are paid by the residents’ Part D plan provider. As such, it’s very difficult to win a competitive account through price. Once Guardian is entrenched in an account, so long as their service levels and relationship management remain at a high-level, they’re highly likely to maintain the account. High switching costs + no incentives to switch = sticky customers and high barriers to entry.

From a top-line perspective, the four-wall demand profile of the business is one of stability and economic insensitivity. Patients need drugs in good times and bad, and the Rx/patient has been stable-to-growing over a long period of time. Guardian’s listed peers exhibited strong resiliency during the GFC as an illustrative example.

2. First-mover incumbency benefits add to high-barriers to entry

The ALF industry is still fragmented, but it is consolidating. These chains prefer to work with a single pharmacy provider for simplicity (one software connection, one throat to choke). This preference for a single pharmacy provider creates first-mover advantages.

For example, Guardian’s customer in Georgia – Phoenix Senior Living – uses Guardian for the 70% of their national footprint (everywhere Guardian has a presence). Phoenix recently opened a new community, which defaulted to Guardian. If Phoenix were to acquire a facility (happens all the time), they’d switch the provider to Guardian. In effect, a big portion of the industry’s growth (organic and M&A) comes from existing developers, favoring incumbents such as Guardian.

Guardian has national accounts relationships with the likes of Brookdale and others where they’'re basically the preferred supplier. When Guardian goes into a new area where Brookdale has a presence, they often win that business. Similarly, whenever Brookdale comes to a new area, that business often defaults ot Guardian. All this is to say that being early and having scale gives Guardian a significant advantage which would be extremely difficult for new entrants to overcome. As Guardian gets bigger, it becomes even harder to displace.

In summary, it’s a sticky business that favors encumbent relationships with high switching costs and high barriers to entry. The customers have very little incentive to switch and high incentive not to switch. It’s extremely hard to do a full greenfield expansion in a new area. In fact, I even heard of a facility from a competitor in Texas that went greenfield and won basically no business after 2 years. It was eventually turned into a training center.

3. The market just isn’t that competitive

LTC Pharmacy is a very regional business due to logistical requirements and time sensitivity. As such, there is no “national” LTC Pharmacy market, but rather a collection of hundreds of small MSAs. Whilst Guardian has ~15% national market share in the ALF market, they are only participating in 38 states. 66% of their pharmacies are in just 11 states, where they have >25% market share – a majority of which are the market share leader in their CMS-defined MSA. 12 Pharmacies (~25% of their pharmacies) have a >40% market share including a couple of pharmacies that have an effective monopoly. So, whilst the market is highly fragmented on a national basis, it tends to be more consolidated in regional areas. Their national market share is just 14% in ALFs and ~5% in total long-term care.

Competition generally consists of 3 groups: (1) large national chains, (2) mid-sized national chains and (3) independents.

1. Large Chains

There are two large national chains – Omnicare and Pharmerica – that historically dominated the skilled nursing facility (SNF) market. Both have been in business for decades and were previously listed – Omnicare was bought by CVS in 2015 and Pharmerica by KKR in 2017. Omnicare is ~double the size of Pharmerica. They tried to buy Pharmerica in 2013 but the FTC blocked it as the combined company would have had >60% market share.

It's important to note that there’s a general distinction between serving SNFs and ALFs. Guardian mostly focusses on ALFs. Pharmerica and Omnicare were traditionally 100% SNF but they try to do ALF as well. When Guardian talks about its market share of ~15%, it’s referencing the ALF market only. Guardian focusses on this niche – Omnicare and Pharmerica are present in ALF, but majority of their business is SNF.

Omnicare – once the 500lb gorilla with ~40% market share overall LTC market – has been a total failure for CVS. Purchased in 2015 for ~$12bn at 17x EV/EBITDA multiple, CVS has taken ~$8bn of write-offs since. The acquisition never delivered the synergies between retail and LTC pharmacy as hoped. In 2018, CVS gave a presentation on Omnicare’s disappointing performance and described that it was dealing with service-led retention problems. They installed a new management team, but they couldn’t turn it around. In 2019, the CVS CEO called the acquisition “disappointing.” In 2022, Omnicare was declared nonstrategic by CVS and put up for sale. Later, in 2023, CVS noted that Omnicare was unlikely to be sold in the near-term amidst multiple rumors of failed deals. In late 2025, it was put into chapter 11 bankruptcy.

The company has been undergoing constant layoffs and closures. I have spoken to a handful of customers and formers who have nothing but negative things to say. Reviews on reddit/Glassdoor paint a picture of a struggling company that is constantly losing customers due to poor service. The culprit seems to be mismanagement, neglect and incessant cost-cutting at the hands of CVS in an industry where service is paramount. “Do less with more” is the CVS motto, formers say. Omnicare was also largely built through acquisitions but never properly integrated them, which ended up being a big problem later down the road.

Pharmerica recently came publicly again via BrightSpring. From a high level, Pharmerica seems to be doing okay and isn’t in the same position as Omnicare. However, under KKR the company pursued an aggressive acquisition strategy outside of the scope of LTC pharmacy which includes other specialties such as IV Pharmacies. The LTC Pharmacy piece is now just ~20% of their pharmacy sales (a majority of which is SNF), so it isn’t quite so core to what they’re doing. That said, the segment continues to grow and is profitable – it isn’t obviously having problems like Omnicare. Pharmerica seems a formidable competitor in the market.

Guardian management believes that the two are not very focused on ALFs and that they don’t do a good job at it. From my indsutry work, this generally seems to be true. There are a lot of differences between the two markets – most importantly, SNFs are more sophisticated; they have a medical director in house and registered nurses (RNs) administering medication. ALFs don’t have a medical director or RNs, but instead a MedTech (effectively normal people that did a course to administer drugs) giving medication.

The significance is that ALFs require a lot more time, attention and service because their medicine administration personnel are less sophisticated. It also requires dealing with Part D plans, claims adjudication, PBM negotiations, RCM, etc. that SNFs don’t have to deal with because they mostly get paid through Part A or Medicaid. So, the ALF market does require a concerted effort to serve well. In fact, I spoke to some of their smaller competitors who mostly compete in SNF and they mentioned that they hate ALF work because it’s complex, labor intensive and often small in scale compared to SNF. So I do think that Guardian’s view that ALF is complex to serve is supported by evidence.

I expect that Pharmerica will remain a solid competitor in the future. Omnicare remains a wildcard – it may present further opportunities, or a successful turnaround could create a headwind. The former seems more likely in the near-term as they continue to go through the Chapter 11 process. Turning around a company of the size, complexity and regional dispersion of Omnicare would be a gargantuan task, and would likely require further shrinking.

2. Mid-Sized Chains

Guardian sits in with the mid-sized chains. Guardian has ~50 facilities, making it the biggest mid-sized chain. There are a handful (~4-5) national chains with ~10-25 facilities. Public information on these companies is light, but they all look pretty similar to Guardian – they’re privately owned and have been growing nationally whilst also acquiring independents. Similarly, they – along with Guardian – appear to be growing with the market and at the expense of independents. Whilst they do go head to head in specific states, some are stronger than others in particular regions. It’s also worth noting that a majority of these chanes are predominately focussed on SNF, confirmed by my own conversations with a few of them.

Competition with the mid-sized chains is something to monitor. Typically, Guardian grows organically through incremental facilities (i.e. strong position in a growing market, open a second facility to serve existing accounts and open up some more geography). Organic facilities in new markets (which are rarer) typically occur as a result of national accounts that request them to come and open up. For example, a certain customer’s facility expansion in the Pac NW area pulled Guardian into the market. This is likely the same case for when mid-sized chains move into Guardian’s markets. For example, there’s one player who is strong in Texas where Guardian doesn’t have much of a presence.

I’d expect that the landscape will remain stable as mid-sized chains continue to grow organically with the market and at the expense of mom + pops whilst also continuing to buy smaller players. It’s plausible that one of these mid-sized chains could be a transformative acquisition target for Guardian one day. I monitor the expansion of these chains for any step change in behaviour, but it’s not a cause for concern and the market is enormous for all.

3. Independents

Beyond that, there’s the local mom + pops that own 1-2 facilities and typically only operate in 1 state. Some of these are well-run, some less so. Research from SCPC (an industry advocacy body) shows that the average EBITDA margin across the industry is just 2% (compares to Guardian at 7.5%) suggesting that many of these facilities are close to break-even or even loss-making. Most acquisitions that Guardian has made in the past have been ~break-even, illustrating this reality. Independents can survive through great service and strong relationships but it’s very hard to make any serious money in this industry without scale. I expect that the independents will continue to slowly bleed share and get acquired.

Competition Case Study - the Georgia Market

To get a more tangible sense for competition, I’ve done case studies in a couple of areas. I pulled the LTC Pharmacies from Network filings in different states for Cigna and Blue Cross.

From Cigna’s Pharmacy Network in Georgia, we can put together a picture of the competitive landscape. Cigna has ~150 LTC pharmacies under network. I went through all of them individually. A majority (~70%) of the listed pharmacies are just retail pharmacies (many of which are attached to gift shops, retail, etc.). A further 8% are retail pharmacies that actively market their LTC services, some of which own a level of automation machines – likely servicing small and rural assisted living facilities. ~11% are inpatient pharmacies – none of which actively market LTC Pharmacy. Only ~12% are dedicated, closed-door LTC Pharmacies.

As one can see, there are only 16 dedicated LTC facilities in the entire state. We can take a closer look at the individual facilities to get a better picture for the competition.

A few observations. Firstly, ~half of the market is controlled by “chains” – i.e pharmacies with more than one facility. The remainder are mom + pop, single facilities. Secondly, whilst there have been some more recent expansions for mid-sized chains such as Guardian, Gayco and Pharmscript, there have been no independent entries since the GFC.

It’s also worth looking at the customer base in Georgia. There are ~1,650 assisted living facilities according to the Georgia Institute of Technology. However, of those, ~1,300 are “Personal Care Homes (PCH)” which are effectively converted homes/apartments where a small number of seniors live together. PCHs typically have <25 residents, in many cases they have much less. Many might just be 5-10. The other 300 facilities are true active living communities, which they define as having >75 residents.

Clearly, the needs/requirements of these facilities are vastly different. A senior apartment with 6 residents can be served by a local retail pharmacy offering delivery services. Some may not even use an LTC pharmacy at all, instead relying on the residents’ family to go to CVS and get drugs on their behalf. The volumes and complexity are low enough that a local retail pharmacist could plausibly handle the requests. A facility with 100-200 residents would not only overwhelm a local pharmacy from a volume perspective, but also requires much more sophistication in operations/logistics/electronic record-keeping/claim adjudication, etc. At a certain point, the manual labor would become very inefficient relative to robotic dispensing.

Breaking the market down like this helps us to understand the competitive landscape better. Whilst smaller retail pharmacies are active in the market, they really aren’t competing head-to-head with the dedicated LTC pharmacies. LTC pharmacies like Guardian are serving the large ALFs (generally, they are large chains like Brookdale) who require the scale and sophistication. Smaller retail pharmacies with LTC operations are serving the local small PCHs which have low volumes and sophistication and are of little interest to Guardian.

In summary, here we see that there have been 2 expansions for chains in the State over the last 10 years but no truly new entrants. It helps support the idea that barriers to entry are high for new players and favor encumbents.

Tangible scale benefits - Get better as you get bigger

I love companies that get stronger and harder to compete with as they get bigger and I believe that Guardian fits the bill. The industry benefits mostly from scale in (1) purchasing, and (2) negotiating power with PBMs in contracting. It’s clearly visible in the margins that Guardian earns compared to the companies that it buys. The Heartland acquisition in 2024 came with a ~13.5% gross margin compared to Guardian at 20%. It’s misleading to look at Guardian’s flattish margins and assume that they aren’t getting leverage. Currently ~20% of Guardian’s total pharmacies are at break-even because they’re recently opened or acquired (takes time to ramp). The underlying business, stripping out acquisition impacts, is growing margins.

One of the “secret sauces” of Guardian is in its purchasing. Bill Bindley, the chairman of the board, built and sold one of America’s largest drug distributors. He has extensive expertise and relationships in sourcing. Whilst the company refuses to go into detail into what they’er doing/how they’re able to produce better margins (other than scale), there’s clearly expertise within the organization for purchasing whilst most competitors will just rely on a GPO.

Stable, defensive business

Guardian’s track-record of stability in margins/growth illustrate the relatively low competitive intensity in the industry. Gross margins are more impacted by PBM re-imbursement than anything else, but have still remained quite stable in the face of headwinds over the last decade. Note: the company has been more active in M&A recently which has a dilutive impact on margins in the first 2-3 years, explaining the dip in GPM.

Other key operating KPIs – including drugs dispensed per patient and revenue/gross profit per patient –have also shown stability over time as one would expect.

Guardian wasn’t listed during the GFC but its biggest customer and 2 biggest competitors were. They both exhibited resilient operating metrics in the midst of a global meltdown. (Note: 2010 saw gross profit headwinds from a Part D re-imbursement headwind that will be discussed separately).

Brookdale’s occupancy fell by ~2-3% - to be expected given most residents in ALF are self-pay. It’s not a meaningful decline in context what the housing market was doing at the time.

Phenomenal ROIC

The return on capital in the business is supportive of our thesis of high barriers to entry. There has been just ~$130m of cumulative capital investment into the business + $10m of NWC since inception, which will generated >$100m of EBITDA last year. On a unit basis, each facility costs ~$2m to build. The average EBITDA/facility is ~$1.8m (including overhead costs). It is an extremely capital light business. I have visited a facility and can see why – it’s effectively an office with a bunch of drugs on shelves, a few automated dispensing machines and a few trucks. The company has generally spent capex at the level of depreciation over the last 5 years and grown healthily. Capex runs at just ~1.5% of revenue.

This is very clearly not a business where one could just raise money, build a facility and start winning business. If that’d be the case, nobody would do anything else given how attractive the ROIC is.

What makes Guardian Unique

There are a few considerations that I think are worthy of acknowledgement. First is that Guardian’s management team is accomplished and tenured – from C-Suite to facility-level. Second is Guardian’s unique decentralized model.

A quick note on management. Fred Burke is a 3rd-time entrepreneur. He had two successful exits – one to Dunn and Bradstreet and one to Cardinal Health. The Bindley family liked Fred so much that they tapped him on the should to come run/build Guardian when they originally had the first idea. David Morris (CFO) and Kendall Forbes (COO) joined Fred early on in his 2nd venture and the three co-founded Guardian together. Fred’s second business was a nuclear pharmacy which worked on many of the same principles upon which Guardian has been built. The learnings were taken from that and applied to create Guardian.

My impression of Fred and David from having spent a lot of time with them is that they are of high integrity. Their track-record both with Guardian and prior ventures largely speaks for itself. Their “people first” approach to acquisitions is clear evidence that they prioritize building a long-term sustainable business over quick-hits.

Perhaps most importantly, the company has a decentralized model where facilities are effectively independently run with support from HQ on centralized functions (HR, accounting, purchasing, etc.). Guardian seeks to “partner” with individual presidents who are effectively put in charge of running their own business. It’s almost a franchise model. Originally, Guardian had an NCI model where they’d own 80% and leave the other 20% to the president, and the earnings would be distributed annually directly to their president. During the IPO, they collapsed the NCI structure and issued equity to the NCI owners. Incentives are now done with an LTIP and RSUs that are linked to individual pharmacy PnL.

This decentralization has led to remarkable job satisfaction and retention rate of key personnel. I went through each of their 56 facilities individually. In 52 of the 56 facilities, the founding president (from acquisition or organic) was still in charge. In 3 cases, the long-term Pharmacy Director was appointed after the president retired and in one case an external president was hired.

This decentralization allows Guardian to maintain the service levels of a “small business” at the branch level whilst still benefitting from scale. It also instills an entrepreneurial focus from the local presidents as they have direct PnL responsibility and equity in the business, effectively seeing themselves as “owners” – many even have Owner/Founder in their LinkedIn title. The presidents are truly in charge of their businesses. For example, I overheard a conversation between Guardian Atlanta’s president and Fred Burke about potentially putting in some more automation equipment due to recent volume growth. Fred’s response was “okay, sounds great” – the presidents are truly in charge.

Attrition is the enemy of good service. As one of Guardian’s customers put it: “LTC Pharmacy is a people business.” The same customer noted: “I know when I have a problem, I can call Matt and he’s going to pick up and sort it out. That’s very important to me.” Service businesses are people businesses – your only asset is ensuring that people do their job consistently and correctly. I truly believe that Guardian’s ability to attract and retain good people is a competitive advantage.

The C-suite has built up a playbook for execution in the LTC pharmacy space and built a decentralized model that works. With 56 facilities, they have a lot of successful reps. (The last column in table below is years with the business)

The growth opportunity

The growth outlook for LTC Pharmacy is attractive on multiple levels. Firstly, the population keeps aging as the Baby Boomers will enter the later stages of their life over the coming few decades. The 65+ population will grow at ~2% through 2040. The 85+ population will grow from 6.5million in 2022 to 13.7million in 2040 – a CAGR of ~4.5%. Given that the average admission age of an assisted living facility is 84 years old (>50% of patients are 85 or older), the assisted living customer base should grow at ~4-5% for the next few decades – a robust, long-term, structural driver.

At the company level, mid-sized chains should be well positioned to grow even faster than the national average for a few reasons. Firstly, they have higher exposures to faster growing states. Secondly, gradual share gain from independents should continue. This will be driven by consolidation in the ALF market and the closure of smaller homes in favor of larger communities with more sophisticated pharmacy needs. Thirdly, Omnicare is not growing and in fact it appears to be shrinking (closing locations). The absence of growth alone from Omnicare – given its size – should create outsized growth opportunities for smaller chains like Guardian where a 1% market share gain is very meaningful to their business. Guardian believes its organic growth rate should be HSD % over the long term.

The company grows largely in three ways. First is increasing beds at existing locations. As existing customers build more facilities, increase occupancy or expand existing facilities, same-store volumes should grow. Second is new facilities. This typically happens in 1 of 2 ways. First is that they expand laterally from an existing facility in a growing market where they have a strong existing position. Second is a greenfield into a new market where they have meaningful existing national account coverage that they can win. The latter is rarer as the barriers to entry are high as we know – Fred admitted to me that it will be a minority of their growth. The former has historically been common. Of the company’s 53 pharmacies, ~21 have been organic. Both ways of growing are extremely capital efficient, as we’ve already discussed.

The second way is through M&A. Guardian has highlighted a target pool of ~100 facilities nationwide that would be attractive targets. These targets need to be (1) in an attractive market and (2) have a strong leader who’s willing to stay on with the business. Guardian is not interested in purchasing facilities with the intention of installing its own management team.

The acquisition playbook has been honed over 30+ acquisitions. Typically, year 1 is a lot of blocking/tackling – integrating the company into Guardian’s tech/HR/finance systems as well as training for the leaders on Guardian best practices (effectively knowledge sharing). In year 2-3, Guardian HQ provides a multitude of help measures to the facility including centralized purchasing, margin management system/tools (group learnings on how to run leaner), sales + marketing efforts (including introductions to national accounts), efficiency initiatives (including automation equipment investment where necessary), PBM negotiations and collections/RCM initiatives. In year 4, margins are targeted at group level.

The company has had success expanding revenue in targets post-acquisition, too. Below shows a case study of 5 anonymized acquisitions completed since 2015 (prior to Covid). In just 4 years, the pharmacies collectively grew revenues by ~40%. One can find many articles from local news sources, etc. of acquired pharmacies that have seen great success shortly after acquisition also. Many expanded to multiple facilities.

The economics of acquisition are extremely attractive. Typically, the facilities that they’re purchasing are at ~break-even, and the deals are done on a multiple of patients. In 2024, the company purchased 6 facilities for ~$20m. Those facilities are doing ~$85m in revenues. In the MD&A, we can estimate the profitability of the pharmacies by looking at the acquisition contributions on revenue, COGS and SG&A. Collectively, the pharmacies have a ~13.5% gross margin and SG&A rate of ~13.5% with ~no profitability. This compares to Guardian group with ~20.5% gross margins, SG&A rate of ~12.5% and an EBITDA margin of ~7.5%.

If Guardian were to get that $85m in revenues to group margins of 7.5%, it would represent a ~3x EV/EBITDA multiple in year 4. After accounting for growth, the multiple could plausibly be below 2x. The returns are understandably very attractive.

Guardian believes that it has ~5% of the LTC Pharmacy market nationwide, and ~14-15% of the ALF-specific LTC market (given that it doesn’t service many SNFs). Within ALF, Guardian believes that it is the largest player with the remainder of the market made up of mid-sized chains and independents. It’s hard to say how much market share they could ultimately take, but there seems to be no limits to their ability to grow. Just last quarter, the company purchased three independent pharmacies. Given that Pharmerica and Omnicare managed to collectively own ~60% of the SNF LTC Pharmacy market, I don’t see a strong reason that Guardian couldn’t get to 20% or more market share over the coming 5 years. Going from ~55 facilities today to ~100 facilities over 5 years seems entirely plausible.

Over the long term, 150 facilities doesn’t seem so unreasonable, though they might need to do some transformational M&A (i.e. purchasing another mid-sized chain or a chunk of Omnicare facilities) to get there, and its unclear if something like that would even work given the unique model that Guardian has. Either way, it seems likely that the company will be able to maintain its HSD% organic growth targets for a very long time with M&A supplementing growth.

APPENDIX

LTC Pharmacy 101

The drug administration regimen within a long-term care facility differs greatly from how any typical person would get their drugs. Residents, who tend to be old and often unwell, typically can’t leave the facility to go to the local retail pharmacy to fill a script. As such, drugs are centrally procured by the facilities and administered to residents daily. This process is called a “med pass” - a scheduled process by which facilities administer drugs to patients, typically occurring 3 to 4 times per day.

Unlike hospitals and other select inpatient facilities, LTC facilities (such as an assisted living home) are not large enough to have their own internal pharmacy like hospitals (except in very rare cases), so they rely on external pharmacies to deliver drugs. The average LTC resident is taking 12-14 drugs per day. Retail pharmacies are not set-up to deliver drugs in a bulk manner to an LTC facility. This is where LTC pharmacies come in.

LTC Pharmacies differ greatly from retail pharmacies. They are effectively fulfilment centers – there is no storefront. Here (below) is a good video showing the operations of an LTC pharmacy and Guardian’s looks pretty similar (albeit more automated - I have visited one). LTC Pharmacies are purpose-built for high volume throughput including automated script filling and custom packaging. The key differentiating service of an LTC pharmacy is that drugs are provided in specialty packaging. That is, the facility receives a blister pack for each resident that includes all of their medications with labelling (i.e. take at 10am). This packaging makes drug administration (a key process in an LTC facility) very simple, as they facility just takes the pre-packaged blister packs and hands them to residents.

LTC Pharmacies are effectively logistics businesses – they do daily deliveries of drugs and typically service a ~200mile radius – it’s quite a regional business. They are also on call 24/7, as emergency situations do occur often given the acuity of the population in LTC facilities.

Also of importance, LTC facilities face strict and increasing regulatory/administrative burdens on drug administration. LTC pharmacies provide services to LTC facilities including handling claim adjudication, coordinating formulary interchanges with physicians, mock audits to assist in compliance with state/federal regulations, EMAR integration, monthly pharmacist reviews of resident drug plan (CMS requirement) and other things. This is particularly pertinent in assisted living facilities where most staff are not registered nurses.

All-in, it’s a pretty simple business once it gets going. The flow goes something like this:

1. Resident has a physician visit who prescribes a drug, they get a script.

2. Resident hands the script to the staff member at the facility.

3. The staff member uploads it into the electronic medical records system (EMAR) which interfaces with Guardian’s software.

4. That script gets sent to Guardian for fulfilment (typically via Fax or eFax).

5. Guardian receives the script in its workflow platform, a technician reviews the script for errors, duplicates, insurance coverage, potential clashes with the resident’s existing drugs, etc.

6. Once approved by the technician, the order is sent to a pharmacist for review. The pharmacist then okays it in the workflow software, and it is sent to the automated machines.

7. The machines dispense medication and fill blister packs. The packs are then checked by an automated machine vision platform and okay’d by a pharmacist. The blister packs are then rolled up and loaded onto the truck.

8. The truck does bulk deliveries daily to facilities. The facility receives the drugs, and in the next med pass, drugs are given the residents. From the resident’s and facilities’ perspective, the process is seamless – they just take a picture of the script, put it into the software and wait for the drugs to be delivered.

9. Once the script is administered, the pharmacy technicians interface with the insurance company to adjudicate the claim and get paid. Facilities and patients don’t have to deal with claims at all.

How does Guardian make Money?

A majority (~75%) of Guardian’s revenue is earned through Medicare Part D, with the remainder coming from commercial insurance and Med Part A. It’s important to understand Part D as it’s the most impactful on the business. Below is a step-by-step guide to how money flows through the healthcare system to become revenue for Guardian. It includes an overview of some key concepts of Part D plans.

Note – the facility pays a small fee for the delivery of meds from Guardian and also in some cases charges residents a fee for drug services. A resident can technically opt out if their acuity is low and they’d prefer to have a family member visit CVS on their behalf and deliver drugs on an ad-hoc basis - ~90%+ of residents opt-in to the drug administration service. However, this fee is a tiny fraction of Guardian’s revenue (<1%). Most of their money is made from the sale of drugs, collected from insurance companies. As such, it’s important to understand whilst facilities chose the pharmacies, they aren’t actually the paying customer – residents pay for the services via their insurance plans. This will be more important later.

Medicare Part D is the federal program that subsidizes prescription drug coverage for Medicare beneficiaries. There are different sections of Medicare that cover different things. Traditional Medicare (Part A) covers hospitalizations, Part B covers doctors visits (includes X-rays, surgeries, etc.) and Part D covers prescription drugs. Prior to Part D’s introduction in the early 2006, Medicare did not cover most outpatient prescription drugs. Part D significantly changed the access and usage of prescription drugs by seniors (many studies show meaningful increase in drug consumption amongst seniors). When seniors reach the Medicare age, most enroll in coverage. Part A is typically automatic and premium-free. Part B is optional and typically comes with a premium, though the uptake is nearly 100%. Part D is also optional and comes with a premium, though uptake is a little lower at ~80% as younger/healthier people may not see the need for a prescription drug plan.

Part D is unique versus A and B because it is entirely privately administered. That is – private companies offer plans to customers which collect premiums and pay claims. In contrast, Part A and B are directly administered by CMS who receives premiums and pays providers directly.

The flow of money for Guardian:

1. Customer purchases a Part D plan: Customers enter a Part D drug plan with a monthly premium. This will cover their payments for prescription drugs. There are hundreds of plans that differ in premiums, benefits, co-pays, etc.

2. Customer gets prescribed a drug: When an enrollee is prescribed a drug, they go to a network pharmacy to get it filled. When they’re at an assisted living facility, the drugs are administered by the facility through an LTC pharmacy (which is in network). The pharmacy verifies the coverage and adjudicates the claim with the PBM. The Pharmacy provides the medication to the patient there and bills to PBM on behalf of the plan.

3. PBM pays claim: Each Part D plan appoints a PBM to administer claims on their behalf. The claim is typically processed electronically in real-time, though sometimes claims require prior-auth or manual intervention for some reason. Once the claim is accepted, money is paid from the PBM to the pharmacy. The PBM will then put in its own claim with the plan provider and get paid from them.

4. Medicare payments to plan: Typically, premiums only cover ~25% of overall benefits, with the remainder covered by CMS (it’s a subsidized program). This is done through a monthly capitated subsidy – a fixed amount per member which is risk-adjusted. This is based on the plans annual “bid”.

   1. There are a lot of things to consider: Risk codes, low-income subsidies, catastrophic reinsurance, rick-corridors, etc. that aren’t so important for Guardian.

How does contracting work?

For a pharmacy to administer drugs to a patient under a certain plan (and for it to be covered), they need to be “in-network.” That effectively means that the insurer has a contract with the pharmacy and agrees to re-imburse them based on contracted rates for their patients, amongst other things. Plans have to adhere to strict network adequacy rules from CMS – i.e. they must offer sufficient access for patients. Understandably, Plans generally only allow access to in-network pharmacies to control costs. All major pharmacies are generally contracted with all major plans with some exceptions.

Here are the key components to the process:

1. Plans, by law, must offer standard network contracts to any pharmacy willing to meet their terms and cannot exclude a pharmacy if they’re willing to contract at the set terms.

2. Plans typically have a standardized LTC pharmacy contract that specifies requirements and reimbursement terms for LTC pharmacies. The contract is itemized – i.e. it includes prices for individual drugs based on a specified ingredient code. The standardized terms of the PBM/Plan are regulated by CMS (more on this later) and plans cannot discriminate, meaning that they have to offer these terms to everyone and accept a contract. Rates, however, are not regulated. Plans can negotiate from these base terms with large pharmacies. Typically, the base terms would be unfavorable, and negotiations would be incrementally favorable for larger pharmacies.

3. Similar to GPOs, pharmacies have PSAOs (pharmacy services administrative organizations) that negotiate with Plans on a collective basis. This helps smaller pharmacies to get more favorable contract terms. Guardian does not use PSAOs anymore - they’ve built out their direct muscle for negotiating with PBMs. Hence, scale is important in the pharmacy business (it’s true of drug purchasing too, which we’ll get to).

4. There are two key components that are set out in the contract: ingredient cost (i.e. cost of the drug) and the dispensing fee (i.e. administrative cost for the pharmacy to dispense). These terms are not set by CMS but privately negotiated between the pharmacy and the Plan. CMS does not regulate the rates at all. In theory, there’s nothing stopping the PBMs from giving unfavorable rates to pharmacies. The free market should regulate this. However, the consolidation of power within the PBM industry has shifted the balance of power such that many believe the free market doesn’t work anymore. The PBM conundrum and PBM reform is a whole separate issue and outside of the scope of this note.

5. Typically, branded drug costs are based on WAC minus certain % and generics are set at a maximum allowable cost (MAC). Dispensing fees are also set – these are opaque and often don’t cover the actual cost of dispensing. These are the two key components. An example is this: drug is dispensed to a patient, the drug code is submitted in a claim to the Plan, the plan re-imburses the pre-defined contracted rate for the drug + a pre-defined dispensing fee. That is the Pharmacy’s revenue.

6. These contracts are typically made annually which coincide with the Part D bids.

From a high-level, it’s actually pretty simple. Below, we pull it all together:

1. Patient gets a drug from Guardian.

2. Guardian buys the drug at $X price from a wholesaler.

3. Guardian fills the script and dispenses it to the customer at the facility.

4. Guardian bills the customers’ Part D plan.

5. The insurance company (via PBM) pays Guardian $Y per drug + a dispensing fee of $Z.

6. Guardian’s revenue is $Y + $Z, it’s costs are $X + costs to dispense the drugs (labor, logistics, etc.) and it’s profit is the difference.

KVH’s business has transformed in the past few years and it is on the cusp of a huge inflection in earnings and profits. It has also undergone significant restructuring to lower overheads, positioning the company for meaningful operating leverage as growth re-accelerates. The market doesn’t seem to be paying attention to the momentum in the business, which is unsurprising given that the company has zero sellside analysts.

I estimate that KVH is trading at ~6x FY27 FCF with ~50% of its market cap in cash (therefore <3x FCF ex-cash). I see 3-5x upside as my base case.

How did we get here

KVH started life as an antenna manufacturer. In 2007, the started to buy up satellite capacity and sell their own data plans called “Airtime.” Airtime subscriptions were a huge success for the company and generated meaningful, consistent growth since their inception. Whilst consolidated revenues over time have been flattish and choppy, Airtime has continued to grow consistently for nearly 20 years, with the exception of a few digestion years during the Oil and Gas downturn ~2014-2016.

Over time, outside of period of network transitions, gross margins were relatively consistent at ~35-40%. The legacy GEO business required some level of fixed costs on capacity and satellite hubs which could cause some fluctuations in margins, such as the ~FY18/19 period. However, the underlying margins were robust and quite consistent up until FY24.

Obviously, things started to go bad in FY24 and FY25 for both Airtime Revenue and Gross Margins. So let’s look under the hood.

First, let’s look at revenues. Airtime revenues took a hit when LEO sat first came to market and disrupted GEO sat (if you don’t know what these things are, do a little googling - I won’t explain it here). The biggest hit was the loss of a large customer (US Coast Guard at ~$12m annual revenue) who went direct with Starlink. FY24 was Airtime’s first down year since inception, with stabilization in FY25.

The decline was driven by a mix of ARPU and Vessels, mostly driven by the US Coast Guard contract and Lesiure Marine vessels which were quick to switch to Starlink. There was a lag between Starlink launching and KVHI’s Starlink service going live, which also created some churn.

FY26e below is the implied midpoint of guidance for Airtime revenues in FY26. You will not that it implies significant growth, which I will adress shortly.

On a quarterly basis, we can see more clearly that GEO revenues fell whilst LEO ramped up. However, from 3Q25, total Airtime revenue returned to growth as LEO growth is now significant enough to offset GEO declines. On an adjusted basis (taking out the Coast Guard Contract, which was a one-of-its-kind), the Airtime revenue decline was far less stark, with LEO growth roughly offsetting GEO declines through 2024 and 2025. However, revenues and vessel counts recently have inflected to record levels. This shows that the overall TAM for Maritime has expanded meaningfully with the introduction of LEO. The company has reached a point where LEO gains are well-offsetting GEO declines.

For context, during FY25 the company shipped ~6,000 Starlink Terminals. Their vessel count at the beginning of FY25 was just 7,400. Starlink is growing rapidly and to a size far beyond where GEO reached.

Looking more closely, we see that the company’s vessel count has accelerated recently. 4Q25 is adjusted for a legacy 2,000 vessel GEO contract that the company let go due to contributing $0 of profit. The vessel count peaked at ~7,100 in FY23 pre-Starlink and is now >9,000 - a company record.

When we look under the hood on margins, things aren’t as bad as they seem either. LEO margins have settled in at ~high-30s/low-40s over the past few quarters, but GEO margins have taken a meaningful hit. The reason is the copmany has fixed capacity contracts that don’t flex down with revenues. These progressively roll off over time with some respite in FY26. My understanding is that there is a cliff in FY27, where GEO margins should recover meaningfully. Contrary to popular belief, GEO is actually dragging down margins today, not LEO.

In 4Q25, the company printed record Vessel and Airtime Revenue growth and LEO margins are tracking ~above where GEO margins were historically. The company is guiding to ~35-40% consolidated airtime margins over time - consistent with what they achieved in GEO for nearly 2 decades. They’re currently at ~35% consolidated, and should move up slightly over time as LEO mix improves and GEO cost-base contracts roll off.

Before I get into FY26 guidance, I want to address a few points.

GEO to LEO - is it a good thing?

I want to address a couple of fundamental concerns about the business briefly and why I think they’re misplaced.

1. LEO is a commodity and will be a race to the bottom

Some believe that reselling GEO had higher barriers to entry because of the hardware component. KVH would manufacture the VSAT antenna (which was more expensive and burdensome to install) and then get the airtime business. Today, the hardware and service are both purchased from Starlink and in theory anyone can purchase from and resell Starlink.

I think that the fear is overblown. Firstly, there were and are white label VSAT anenna manfaucturers. In fact, that is how KVH started. So anyone could have also just purchased antennas and capacity from SES/Viasat/Intelsat and built a product. There was nothing special about GEO that allowed KVH to make a margin which kept competitors out. GEO was similarly competitive, with 100s of Maritime connectivity companies globally.

What KVH really makes a margin on is two things: Breakage and Service. Fundamentally, these two things haven’t changed. The company has long-term customers relatioships for whom they provide service, support and advice. And the breakage element is still true - KVH can buy capacity from Starlink far cheaper than any given vessel owner and make a margin on that. As a top ~3-5 player, KVH has vast advantages on the latter point over the smaller rats and mice players out there.

The dynamics that would lead to a race to the bottom were also true in GEO. It’s not clear why that would change. I do have some sympathy for the argument that switching costs are probably lower if all the antennas are homogenous. However, there doesn’t seem to be obvious reasons that customers would switch or an obvious reason that competitors would race eachother to the bottom in LEO but not GEO. If the service works and pricing is in-line, it’s not clear that it’s worth the hassle. KVH should have some of the best pricing given that it is a larger scale player.

We already have proof from LEO margins relative to GEO margins that the economics of breakage/service really haven’t changed.

2. Starlink will steal the economics

Starlink has an important channel conflict to manage because they won’t be a monopoly forever. OneWeb has come to the market and is having quite a bit of early success particularly with its Inmarsat relationship (NexusWave). KVH and other installers have also begun to resell OneWeb and it is particularly popular in Europe where Elon Musk isn’t exactly popular.

If Starlink decides to get greedy and steal economics from the channel, KVH and other players can just shift customers towards OneWeb and eventually Kuiper/Telesat who will launch in 26/27.

There was significant concern that KVH’s new Starlink deal would be at worse economics, however the recent 18month bulk purchase has confirmed that margins will be ~consistent, which is a proof point that Starlink (at least for now) isn’t getting greedy.

3. Starlink Direct will make the channel irrelevant

When KVH lost the US Coast Guard contract, there was significant fear that Starlink would cut out the channel and go direct. Starlink Direct also signed Maersk and that fear was excacerbated. However, according the Valour Consulancy, Starlink Direct has a ~6% market share. Much of that is very large customers which are worth their while, and big enough to manage/negoatiate directly with Starlink. The US Coast Guard fits into that picture. They also have a lot of leisure marine customers.

Beyond that, Starlink Direct hasn’t made much of a dent. In fact, KVH has mentioned that they’re seeing 100s of Starlink Direct customers defecting from Starlink Direct towards them every quarter due to the abysmal support that Starlink provides.

Starlink Direct will be a threat for very, very large customers. However, most of KVH’s customer base does not fit that ilk.

LEO TAM expansion is underappreciated

Under the GEO regime, connectivity was expensive and frankly just not that good. LEO is much more affordable and way better. Under the new LEO regime, there’s no real reason that most if not all vessels will be connected to the internet.

Valour Consultancy estimates that there were ~42,000 GEO VSAT connected vessels globally in 2023. They also estimate ~180,000 vessels that were L-Band connected (mostly for safety comms). If we use the L-Band connected vessels as a proxy for total addressable market, we can estimate that only ~20-25% of the market was previously receiving GEO VSAT. Hence, LEO could unlock and ~4x expansion in the market as well as a long tail of leisure and commercial fishing vessels that previously couldn’t justify the cost. In fact, Valour Consultancy estimates that LEO connect vessels will grow from ~60,000 in 2025 to 138,000 by 2034.

We’re seeing the proof in the pudding in industry numbers and KVH’s numbers. Valour Consultancy most recently estimated that connected vessels grew ~15% QoQ in 2Q25 alone (the most recent free data that they released). KVH has seen a similar trend, wil the vessel count inflecting meaningfully in 2025. KVH has been shipping ~1,500 Starlink antennas per quarter. Their entire connected vessel based in FY23 before LEO was ~6,000, for context.

From 3Q25, we crossed a threshold whereby one could no longer doubt that the overally TAM expansion for LEO was going offset the GEO declines. FY26 is going to be the big inflection year.

Unpacking the FY26 Guidance and coming profits inflection

As discussed earlier, Airtime is set to grow meaningfully in FY26. The company recently guided to revenue of ~$130-145m, or ~25% growth at the midpoint. Assuming ~consistent hardware revenues, we get ~27% Airtime Revenue growth. The revenue inflection that begun in 2H25 is set to continue in FY26 and will represent the first year of meaningful inflection. There’s no sign that the market will slow down, and visiblity over FY26 is relatively high given the amount of terminals shipped in FY25 which activate with a lag (~6,000 shipped v.s. just ~2,000 growth in FY25).

More importantly, the company’s cost structure has undergone large, fundamental changes in recent years. KVH has by-and-large wound down its manfuacturing operations which has resulted in the removal of a lot of overhead. They’ve outsourced a lot of operatings to philipines and also cut overhead related to managing the complexities of the GEO business, including R&D which is less important when the company isn’t making its own hardware. Without going into detail, LEO is far more streamlined and simple to manage from an operational standpoint.

In fact, the company went through an interesting restructure ~2022/23 period where it became a pureplay connectivity player. That transition was overshadowed by the sudden impact of Starlink.

Run-rate adjsuted SG&A was ~$6.5m in 4Q25. Given the capital light nature of growth in LEO, the company really shouldn’t need to invest in overhead as the business continues to grow. This should lead to a big inflection in operating profits as the topline grows.

As far as Airtime margins go, Gross Margins should also benefit from a mix-shift towards LEO in FY26. In FY27, when the fixed-cost GEO contracts start to roll off, gross margins should take another step up towards the mid-to-high 30s.

In FY26, I believe that the high-end of the company’s guide of $16m EBITDA is achievable. In FY27, with 15% revenue growth (deceleration from FY26 but still robust), I estimate that the company can generate >$30m in EBITDA. With minimal capital requirements (I estimate capex ~$6/7m per year) and significant Tax NOLs, the company’s EBITDA should convert substantially to FCF of ~$25m.

The balance sheet has $70m of cash on it, or ~$3.50/share, which is a result of selling some businesses and buildings (manufacturing + offices) recently. At the current market cap of ~$138m, that’s an EV of ~$68m for a potential FY27 EBITDA of $30m, or ~2.3x EV/EBITDA and 2.7x P/FCF (ex-cash). The company has recently started to repurchase stock and has a current authorization of $15m. I expect that they’ll continue to retire shares.

What should a growing, capital-light, reseller of internet to boats like this trade at? If we are right about the fundamentals of the business (that Airtime will look substantially similar from a revenue stability/margin profile going forward as it has in the past with steady growth and margins), I’d argue that 6-8x EBITDA would be a fair multiple for a cashflowing company that should grow HSD% for some time to come. I see 3-4x upside on my FY27 forecasts.

Barbarians at the gate?

There’s been some interesting board-level activity in KVH over the past few years. Here’s a brief timeline of events:

1. In March 2022, Black Diamond Capital filed 12.2% ownership via 13G.

2. Black Diamong continued to buy shares in July and August, aggresively purchasing shares after the announcement of the internal navigation sale, which had been flagged by management as for sale.

3. On 9th of August 2022, KVH sold its inertial naviagtion business to EMCORE for $55m in cash, leaving the company with a significant windfall. They needed time to assess what to do with the cash.

4. On August 18th, the company adopted a temporary poison pill, capping ownership at 15% until the end of the 2023 AGM. The timing would suggest that they knew someone was trying to get control, though they framed it as a tool to fend off liquidators. This effectively froze Black Diamond from accumulating more stock.

5. The next day, Black Diamond changed its filing to 13D, and showed a 7.29% position.

6. On Feb 3, 2023, the company achieved a coopertaion agreement with Stephen Deckoff, the CIO of Black Diamond. He was not allowed to go up to 25% ownership without breaching the posion pill and was made a board observer. He was then officially appointed to the board at the AGM.

7. Bradley Radoff and Michael Torok (two small/microcap activists who often work together in tandem) both showed up as substantial throughout 2024. It’s unclear if they have any connection with Stephen Deckoff.

Before we go on we have to discuss Stephen Deckoff and Black Diamond. Deckoff draws his lineage from Drexel Bunrham and is a special sit/distressed guy. He was heavily involved with Speedcast (the 2nd largest Maritime Connectivity player) which went bankrupt during Covid, so it is not a random coincidence that he has stumbled into this industry. He was understood to have owned ~30% of their debt.

After the company went bankrupt, Deckoff got into a bidding war with Centerbrige for the company. In Jan 2021, they gave up and cut a deal with Centerbridge to sell debt interests to them. The company ultimately emerged from Bankruptcy shortly after with no debt and a $500m injection from Centerbridge. It is interesting that Deckoff wanted to own the Maritime assets and he’d have been intimately familiar with the indsutry and KVH during that process.

What’s more interesting is that the company emerged from bankruptcy under the leadership of a guy called Joe Spytek. In fact, he was made Speedcast CEO early in 2020 just as the company was entering bankruptcy, so he would have had an intimate relationship with Deckoff. Joe was the founder of ITC global - an offshore energy focussed satellite communications copany - which got sold to Panasonic and then was recently sold to Marilink. In mid-2025, Joe Spytek was nominated to the board of directors of KVH.

In July 2022, David Kagan was also nominated to the board of directors, who’d proviusly worked in exective rolls at ITC and Speedcast before moving to Globalstar. This was around the time that Deckoff was buying shares but before he was on the board.

Today the board has 5 members including Chris Bruun (CEO), David Tolley (satellite executive that was appointed at the same time as Kagan), Kagan, Spytek and Deckoff. 3 of the 5 directors on the board were involved with Speedcast, which is 3 of 4 independent directors. The board has entirely turned over since 2022.

What’s the end game here? Deckoff has continued to accumulate shares in the company, as have Torok and Radoff, albeit in small amounts relative to their positions. Together, the 3 have ~35.5% of the company, meaining that they could probably get a sale of the business consumated.

What is clear to me is one thing - this is a scale game. The bigger the scale, the better purchasing power one has with Starlink and the higher their margins will be. There is signficiant synergy in putting KVH together with Speedcast or Marilink. KVH also lacks in certain areas of value-added services such as IT versus Speedcast and Marilink.

I see either of the following outcomes as possible:

1. They want to sell the business to Speedcast.

2. They want to reverse merge Speedcast into KVH.

3. They want to build KVH into a Speedcast competitor by embarking on their own roll-up strategy.

All 3 situations seem positive to me. The smart money sees value in KVH as a company, either way.

I recently bought a position in Civeo. Life as a public company has been nothing short of misery for Civeo shareholders. However, I believe that the company is on the cusp of a potentially meaningful inflection in earnings and free cashflow. Recent engagement from a well known activist has pushed management towards significant capital returns. I also believe that their Australian assets are of under-appreciated quality worthy of more attention.

Civeo is in a cyclical business and so there are a range of outcomes and the risk is higher than things I would typically own. However, if things go well, I believe the stock could be a 2-3x over a 1-2 year period.

Introduction and Brief Investment Thesis

Civeo is a workforce accommodations company that provides permanent villages, temporary lodging and integrated services (catering, housekeeping, etc.) to workers in remote resource extraction regions. They have two businesses: Canada and Australia.

The Australian business predominately serves the Bowen Basin – the world’s premier high-quality Coking Coal region. The Canadian business predominately serves the Oil Sands regions in Alberta. In addition, they have ~2,500 rooms of mobile camp assets in Canada, similar to those owned by Target Hospitality (TH-US). These are currently sitting idle.

Life as a public company for Civeo has been nothing short of a disaster. The company was spun off from Oil States International in May 2014 at the peak of the Oil Sands and Met Coal booms. The company was severely over-earning in both regions with occupations at 100% and day-rates well above historic norms. The spin-off was supposed to be structured as a REIT, but the IRS denied it REIT status. The stock fell 50% in one-day, and then proceeded to decline ~95% from its peak. Profits fell by ~80%.

From 2016-2023, the underlying business was mostly doing fine, except for some volatility around Covid. The Australian business exhibited strong growth whilst the Canadian business had some volatility with oil but still generated plenty of cash. From FY16 to FY24, the company generated ~$700m of unlevered Free Cashflow, or ~$75m/year on average. A majority of that went towards paying interest and de-levering the balance sheet from 4.5x to <1x with a measly ~$70m ear-marked for investor returns. Historic underlying cashflows from Civeo have been very robust.

Since 2023, as one will see from the chart below, Civeo has become a tale of two businesses. Australia has continued to grow and perform extremely well whilst the wheels have fallen off the Canadian business.

Below I will explain the situation in a bit more detail, starting with the woes in the Canadian business.

Canada’s Woes

There are two parts to the Canadian business and both have been a thorn in Civeo’s side. First is the lodging assets in the Oil Sands region. The long-short of the situation is that Canadian oil sands operators have been on a multi-year journey to reduce their cost structures. A big part of that has been reducing labor intensity on site - particularly FIFO workforce. These really started to take place in 2023 and hit the company hard in 2024/25. The company’s EBITDA collapsed over this period from $80m to just ~$20m in FY24 and FY25.

In early 2025, the Canadian business tipped into losses and Civeo made the decision to restructure the business. They cut overhead meaningfully and closed multiple lodges with poor occupancy. They’ve effectively right-sized the business for the new reality of Oil Sands. Whilst it’s possible that the region could show strong occupancy growth, the major oil sands companies have favored share buybacks over new investment, and Civeo doesn’t expect a meaningful uptick.

The good news is that this should be a trough level. Occupancy seems to have stabilized in 2H25. In fact, the company is expecting that FY26 may even be a slight uptick from FY25. Customers have cut labor so meaningfully already that Civeo struggles to see how they could possibly go much lower - of course anything could happen. The Oil Sands companies have also done a great job lowering their break-evens to low $30s and even into the high-$20s, meaning that they’re much more resilient to $55-60 WTI than they used to be, and one wouldn’t expect meaningful business disruption at that level outside of a material negative oil shock - current occupancy levels are ~consistent with $55 oil in late 2025.

I estimate that the business should be able to generate ~$22-25m in EBITDA in FY26 and build from that level if things improve. Canada is not the key earnings driver, but stabilization in this business is key as it was previously loss-making.

The other part of the Canadian business is the mobile camp assets. These are deployed to remote projects which are large but temporary in nature. For example, new pipelines. Civeo has ~2,500 readily deployable camps sitting idle in Canada. These were previously deployed into two pipeline projects: TMX and Coastal Gas. Both projects finished ~22/23 but the camps weren’t redeployed - there just weren’t any large infrastrcuture projects to redploy to.

The outlook here seems to be warming up. There are currently a large handful of large pipeline, LNG and infrastructure-related projects in Canada that are awaiting FID with an enormous amount of addressable rooms. The company is hopeful that these projects will achieve FID this year, after which a decision on the rooms will be made shortly after. Civeo has one of the largest idle fleets of mobile accomodations in North America and should be well-positioned to serve at least a portion of this demand.

I’m no expert here but I’ve done a little digging on these and it seems likely that we’ll get some FIDs in 2026. Coastal Gaslink Phase 2 is expected by its owners to take FID in 2026. Civeo supported Phase 1 of this project and is likely well positioned to support phase 2. PRGT is sexpected to achieve FID in late 2026 and has been selected by the federal government for fast-tracking. FID is also expected this year for the BC Hydro transmission line where early works have already taken place.

In addition, there’s also the Data Center piece that I’ve written about in the past. Target Hospitality recently won a 1,500 bed project and has discussed a 15,000 bed pipeline. Civeo is seeing a similar thing noting that the sales folks are as busy as they’ve ever been. Given the sizes of these projects, Civeo only really needs to win 2 or 3 projects to soak up its capacity.

As I understand it, at the peak in FY21/22, the Mobile Camp assets generated ~$25-30m in Gross Profit (effectively, EBITDA) on an average deployed room count of ~1,600. They currently have 2,500 rooms sitting idle and another 1,000 rooms that are attached to lodges in Canada that could be mobilized with a small amount of capex.

Hence, with ~2,500 to 3,500 rooms to deploy, the assets could generate an incremental $30-60m of EBITDA for Civeo. The company has significant NOLs in the US and Canada and deminimis capital would be required for the first 2,500 (small amount for the next 1,000), meaning that EBITDA should largely drop straight down to FCF. For a company that will do ~$35-40m of FCF this year, this would provide an enormous windfall.

The company is not in control of project FIDs, however they’ve mentioned that revenue generation from the mobile assets could come as early as late 2026 and full run-rate in FY27. I would expect that we will hear announcements in the coming 1-2 quarters about projects here. I don’t believe that the stock is currently pricing in any success in re-contracting these assets.

Put together, with just the 2,500 rooms deployed, I estimate that the Canadian business could generate ~$70m of adj. EBITDA in FY27.

The Aussie Cash Cow

The Australian Assets includes two businesses: Owned Villages and Integrated Services. Owned Villages consists of rooms that are owned and operated by Civeo. Integrated Services consists of services provided for assets that are not owned by Civeo (i.e. managing captive assets on behalf of a mining company). The services business is growing well, but ~80% of Australia’s EBITDA today is still provided by the owned villages. The owned assets are predominately located in the Bowen Basin of Queensland. The Bowen is predominately a Met Coal region.

The Bowen Basin is located in central Queensland. It hosts ~48 active coal mines and produces ~half of the world’s seaborne supply of Met Coal. It’s an enormous region, covering ~60,000 square kilometres (larger than Tasmania) and its mines are scattered across vast, sparsely populated landscapes ~150-300kms inshore. The closest major regional city is Mackay – a coastal hub about 200kms East.

Bowen is blessed by remarkable geological characteristics. The basin’s Permian-age coal seams were deposited in conditions that produced thick, laterally extensive measures with excellent coking characteristics. It is known for its high-quality coal – low ash, low sulphur and high vitrinite. Bowen Basin coking coals command premium pricing on global markets, with brands like Peak Downs, Goonyella, and Saraji setting benchmark prices.

Given the quality of assets, the region has historically been relatively stable outside of the 2010-13 boom/bust period. Production and exports have declined recently largely driven by a decline in demand from China, partly offset by an increase in demand from India. Met Coal is predominately used in steel production, so China’s slowdown/global weakness in construction/automotive production has impacted seaborne demand in recent years. Exports are forecast to rebound in FY26 and FY27.

It’s worth noting that the fundamentals of the Bowen Basin are quite a bit different to Abathasca oil sands. Oil Sands produces a small fraction of the world’s oil and was generally considered to be high cost. The Bowen Basin produces 50% of the world’s seaborne Met Coal exports and is home to the highest quality Met Coal in the world. Whilst it is still susceptible to impacts from commodity prices, the Bowen Basin is more critical to Met Coal than Abathasca Oil Sands are to Oil. It should be more stable as a result.

That’s visible when observing the company’s sound operating history throughout multiple commodity price cycles.

Over time, employment in the region has been relatively stable-to-growing, and the number of avaialable beds has remained relatively flattish. The fundamentals are strong for workforce accommodation in the basin.

~50% of beds are owned by mining companies themselves with ~50% by private companies like Civeo. I estimate that in the Isaac LGA (main LGA within the Basin), Civeo has ~80% of non-captive beds. In other words, they have a very strong competitive position.

From a high-level, Civeo’s Isaac villages are well-situated in the major cities of Isaac, nestled between large operating mines and dozens of proven deposits. Copabella, Moranbah and Nebo are well-situated in Northern Isaac, each surrounded by a handful of active mines. The Middlemount assets are similarly well positioned to service a cluster of open-cut mines in Southern Isaac. The Dysart village acts as an overflow for assets between Moranbah and Middlemount, as well as serves Lake Vermont and Saraji.

According to management and confirmed by 3^rd party reports, Civeo’s Northern Isaac accommodations (~55% of its total legacy room count) are at capacity and have been for some time. From a late 2024 economic needs assessment (report).

“Based on conversations between Pacific National and Civeo, the Civeo Nebo Village, Coppabella and Moranbah all are currently at capacity with no additional availability for Pacific National employees in the foreseeable future”

And from the company’s earnings call: “3 of our Bowen Basin villages continue to be effectively operating at full capacity, and we’re seeing strong occupancy across the remainder of our owned village portfolio.”

All of this is to say that the Basin’s long-term fundamentals are sound and that the supply picture is favorable. However, it is not immune from commodity weakness. In FY25, the Met Coal price weakened and occupancy took a hit in 2H25 for Civeo, particularly in 4Q25.

The company has guided to lower occupancy in FY26. However, Met Coal prices have recently recovered from their FY25 lows. On the 4Q25, Civeo seemed confident that levels would stabilize at these levels and potentially improve if Met Coal prices were to hang in at this level. Of course this could change at any given moment.

Though fluctuations will impact EBITDA for the Aussie business over time, it seems to be a fundamentally sound set of assets capable of generating significant cashflows through the cycle and on more structurally sound footing that the Oil Sands assets given the position of the underlying Basin relative to global commodity demand/supply.

Unpacking FY26 Guidance and upside

The company guided to $85-90m in EBITDA in FY26 which would amount to ~$30-40m of FCF depending on capex.

I think that there could be upside in FY26 guidance if things continue as they are. The key assumptions in the company’s guidance from their call were:

• Stabilization in Canada with flat to slightly up occupancy.

• Weakness in Australia, with consolidated occupancy roughly flattish YoY (would amount to ~MSD% organic decline in occupancy YoY).

• No contribution from the mobile camp assets.

With these assumptions, I get towards the high end of the EBITDA range. I have modelled flat occupancy in Canada and taken a slightly more bleak view on the Australian business with organic occupancy -HSD% YoY. This gets me to ~$87m of EBITDA on constant currency rates.

However, the company’s average AUD/USD exchange rate in FY25 was $0.645. Though the company didn’t explicitly comment on the exchange rate in their guide, my modelling suggests that they’ve baked in no impact from FX. The AUD/USD currently sits at ~$0.71, or ~8% higher than their average for last year. This should provide a tailwind from 1Q26 but especially in 2Q26. If the exchange rate holds at this level, I estimate a ~$7-8m tailwind to EBITDA guidance in FY26.

Altogether, I’d expect EBITDA of ~$95m and FCF of ~$43m in FY26, putting the stock at ~7x FCF. The company has already repurchased 500k shares in 1Q26 (~$13m, or ~4% of S/O) and has another authorization. Going forward, they’ve comitted to at least 75% of cash returned to shareholders. I would expected most of that cash to be used for buybacks with some modest develeraging.

In FY27, with a little fortune, things could really inflect higher. If occupancy were to recover to FY25 levels in Australia (which should be consistent with met coal prices at current levels of ~$220-240) and the CIS business continues to march towards its $500m AUD revenue goal, I estimate that the Australian business could do ~$115m of EBITDA in FY27. As discussed, if the 2,500 mobile camp rooms get deployed, then the Canadian business could do ~$70m in EBITDA. With ~$26m in overheads, that would get Civeo to a consolidated EBITDA of ~$160m and ~$100m of FCF.

Share Count should be ~10m by the end of the year, which would be a market cap of ~$280m, putting the stock at <3x forward FCF. Even at a modest 6-8x FCF multiple, there’s significant upside to the stock at current levels.

Loads can go wrong for Civeo but you don’t have to believe much for this scenario to come true. The stock is on a single-digit FCF entry multiple with a commitment to repurchase stock. There’s risk but I like the R/R at $28.

Ticker: PUMP-US
Stock Price: $11.26
PF S/O (post-raise): 121.5m
Market Cap: $1.37bn
ADV: ~$30-40m
PF Net Cash (post-raise): $105m
Enterprise Value: $1,260m
FCF Yield: 15%

Overview

ProPetro is an oilfield services company. Their legacy business is tied to completions largely in the Permian; specifically, pressure pumping. It’s a nasty business - capital intensive, competitive and highly cyclical. The stock’s historic multiple of 4-5x EBITDA reflects that reality. It does, however, throw off a lot of cash.

In late 2024, ProPetro entered the on-site power generation business through the formation of a new subsidiary: ProPwr. ProPwr provides onsite generation through small turbines (20-50MW packages) and reciprocating engines (~5-15MW packages) for the Oil and Gas and Data Center industries. In effect, they bring power to your site and run a small (or sometimes large) mobile power plant on medium or long-term contracts.

The economics and growth potential in ProPwr are distinct and far superior to the legacy completions business. I forsee significant growth in ProPwr. I believe that ProPwr alone could be worth >$30/share within 3-5 years. The current market cap is incorporating almost zero value for the ProPwr business today despite its phenomenal success over the last 12months. My 2-year base case is ~$35.

I have an enormous portion of my net worth invested in this stock. Details below.

The Behind-the-meter (BTM) power opportunity

It’s not news to anyone that the grid is struggling to connect required large loads on requested timelines. This hasn’t stopped the data center industry from building as quickly as possible. The key problem is that data centers are either (1) following construction timelines that run ahead of the grid, or (2) built for loads well in excess of what the Utility can realistically deliver in the short/medium-term. Many data centers today are being started without line of site on the utility or a plan to connect to the grid at all. The significance of this is that data centers need to find power to get their data centers energized before the grid is available.

Elon put it most elegantly himself earlier this year:

“The limiting factor for AI deployment is fundamentally electrical power. We’re seeing the rate of AI chip production increase exponentially, but the rate of electricity being brought online is negligible. It’s clear that we’re very soon, maybe even later this year, we’ll be producing more chips than we can turn on.”

— Elon Musk, World Economic Forum, January 2026

“Between now and then, the constraint for server-side compute, concentrated compute, will be electricity. My guess is that people start getting to the point where they can’t turn the chips on for large clusters towards the end of this year. The chips are going to be piling up and won’t be able to be turned on. So how are you going to turn the chips on? Magical power sources? Magical electricity fairies?"

— Elon Musk, Dwarkesh Podcast, February 2026

The grid continues to move at a glacial pace, and whilst most investors assume that the grid will “catch-up”, by all accounts it appears to be falling further and further behind. This has caused an urgency in the industry to bypass the grid entirely and build onsite power plants.

Realistic options for onsite power are largely limited to two: Nuclear or Gas (Solar and Wind are ruled out due to intermittency). Nuclear is not really part of the conversation for the next decade and is also not fast, which just leaves gas. However, anyone that’s read a GEV earnings call knows that Heavy Duty Gas Turbines are heavily backlogged and EPC labor is heavily bottlenecked.

So what choice do we have? Fortunately, we do have a solution. Utilites already have a need to deploy fast emergency generation when large generators go offline due to weather damage. Their solution hss historically been to deploy a fleet of packaged mobile generators (mixture of large reciprocating gas engines and small single-cycle turbines) to plug the gap. These are basically “power plants on wheels” that can deploy fast and reliably.

Below is a picture of what these look like. It’s literally a generator on wheels which includes a mobile substation (balance of plant). Sometimes they are skid-mounted when the deployments are longer-term in nature.

There’s strong consensus amongst industry participants that a mixture of packaged reciprocating gas engines and small gas turbines will be the BTM solution of choice.

The most famous case for this solution is Elon’s Colossus cluster, which relies upon 750MW of small Cat turbines (going to 1.5GW by early 2027 and likely beyond). Solaris (SEI-US) is the vendor for this project and really was the first to prove out the concept at scale.

Why the recent fuss about BTM?

I go to a lot of Data Center Industry conferences. In mid-2025, I started noticing a significant pick-up in discussions for BTM power from industry folk. This was followed by a handful of public deal announcements from VoltaGrid, Cat, Wartsilla and ProPetro as well as private deals without public announcements that I picked up from field work.

In October, things seemed to really picked up steam. I spoke with multiple private engine OEMs at conferences that spoke of significantly extended backlogs for within the last 4-8weeks. One of them sold ~2 years of production to a single customer. Another spoke of multiple gigawatts of late-stage, firm deman. Around the same time, we also saw a capacity announcement from Cat (the industry’s largest supplier) on the back of “unprecedented demand” in early November. One private BTM provider told me at a conference that he was getting multiple calls asking for 300-400MW of capacity per day.

It’s unclear exactly what has catalyzed the mad rush for capacity. My hypothesis from extensive conversations and following the space for some time is that it was driven by a collection of the following:

1. Existing grid capacity seems to have been exhausted, and projects are now necessarily being started without a substation or even transmission line with the hope that it may come eventually, necessitating BTM.

2. Many large-scale projects were started in 2024 or 2025 which should be nearing energization in 26/27. As utilities are slipping on timelines, BTM crunch-time is coming.

3. The sheer scale of power demand has pushed customers to look at all possible options for power generation. Small Turbines and recips are the most proven and fastest amongst available options.

4. Recent legislature in ERCOT (SB6) has given Oncor the ability to “force curtailment” of load. This has necessitated BYOP. It now seems that most, if not all, projects in Texas will be built with onsite generation. Texas alone is an enormous data center market especially for large training sites.

5. PJM has been a little bit messier. They initially proposed mandatory curtailment in August 2025. However, stakeholders pushed back on it and they removed the mandatory concept in October. In Jan of this year, they released a new proposal that achieves a similar outcome: voluntary BYOP can achieve expedited connection and avoid curtailment. If one doesn’t BYOP, they are subject to curtailment before emergency demand response.

6. Utilities have generally continued to be unreliable outside of Virginia. Many DC developers are adopting the attitude of “you can’t trust the utility.” Timelines are dissapointing and being met with more scepticism.

7. XAI’s Grok 4.1 model was a huge success and enabled by having the largest cluster of GPUs in the world. It was a sort of real-world verification that scaling laws hold and that Coherent clusters have distinct advantages over distributed compute. It was also built in break-neck speed. A key enabler of the speed/scale of the Colossus clusters was Solaris and its BTM turbines. They have enabled Elon to scale >1GW in a matter of months.

8. The Chip shortage seems to have caught up but power remains a constraint (Satya comments, Elon comments, etc).

9. Oracle’s recently announced mega-deal with OpenAI will be powered by 2.3GW of Jenbacher reciprocating engines by a company called VoltaGrid. This has probably caused people to pay attention and perhaps a “rush” to secure capacity in recognition that megadeals are coming which could clog up the supply chain just like they did with large turbines. There’s already proof from my work that this is happening.

Who’s playing? Oil Services companies have a right to play and right to win

In much the same way that Bitcoin miners were in the right-place-right-time for large-load sites, oil services companies are in the driving seat for BYOP.

The reason is that the grid in the Permian is almost non-existent. In fact, the Permian has been facing a similar power problem as the data center industry over the last few years, albeit at much smaller scale. Historically, equipment (such as a frac pump) in the permian was powered by diesel engines. As the key E&Ps have become more concious of emissions, they’ve been moving away from diesel and towards electric powered equipment. However, the unreliability of the grid has caused E&Ps to look towards small natgas microgrids to power that electrical equipment.

Some E&Ps or pressure pumpers (such as ProPetro) have experimented with doing this themselves. However, this trend has given birth to a new business model of BTM power in the Permian. This is the use case for which VoltaGrid was created.

In addition, companies like ProPetro have a lot of experience dealing with large scale diesel and gas engines that have powered their frac fleets for many years. The heavy mechanical skills are largely transferrable to BTM and their relationship with CAT (from whom they’ve purchased over $1bn of equipment over time) is critical in securing a reliable supply of generators. In fact, one might argue that running a stationary microgrid on a clean Data Center site is easier than mobile engines in the desert out in West Texas.

The key players today in the market all started in the oil patch (Solaris, VoltaGrid, Liberty and all the private players) because they already have the product and business model to service data centers.

The big debate: A grid stop-gap or sustainable business model?

Anyone who has followed Solaris knows that the bear-case on BTM been some form of the following: Sure, Solaris is pretty profitable today, but when the grid catches up to demand, Solaris will have no business model! It’s just a stop-gap to 2030.

This was the prevailing view even ~6months ago and the investment markets (even those of us who traffic heavily in power stocks) were skeptical of BYOP, seeing it simply as a bridge-to-grid.

I have come to a very different view over the last 6 months.

Firstly, it seems that the supply/demand gap for power is worsening and will continue to worsen as the scale of sites keeps growing. Many analysts and consultants have done estimates of power supply/demand and whilst it’s hard to model with accuracy, Morgan Stanley shows that the deficit through to 2030 continues to grow and will reach 40-50GWs by the end of the decade. It’s not hard to reach a view where this is seen as conservative. NERC’s recent reliability assement shows that PJM and MISO are facing significant power shortfalls in the short-term - i.e. this is not a 2035 problem, it’s a 2027/28 problem.

This shortfall is BYOP’s market. Overall demand is significant and will continue to grow for the foreseeable future with new sites needing onsite power even if the older ones transition to the grid.

However, where I have really changed my thinking is on this point: the assumption that data centers will rent equipment for a few years and then hand back the generators seems to be a huge a misconception. In a mad rush to scale up, why would a power-constrained data center hand back valuable power when the grid arrives instead of adding the grid to existing onsite gen? In fact, I have heard multiple anecdotes of exactly the latter happening. A VoltaGrid employee that I met at a conference noted that they’ve never had any equipment returned once it’s gone to site. He showed me a picture of a deployment in San Antonio and asked, with a brief pause and a wry smile, “do you think they’re moving those any time soon?” The CEO of a large DC operator joked with VoltaGrid after they deployed mobile turbines that “they’ll never get these back.”

Solaris’ experience with XAI has been similar. XAI originally expected to rent 80MW of turbines for 6 months whilst they waited for the utility. It then went to 400MW for 2 years and now 1.5GW for 7 years. The utility arrived with 400MW, but their demand just kept increasing. Instead of returning the power, they’ve used the grid to boost their overall power. XAI’s commitment to funding a JV with Solaris also shows their long-term belief in onsite power. It’s one of the very few examples that we can actually point to, but it’s a positive signal.

A discussion I had with a procurement person at one of the hyperscalers explained to me that their strategy is to use onsite gen to power all of their cooling loads where possible (less critical and complex loads, typically making up for 30-40% a DC’s power usage) to free up more grid power available for scaling up data halls. This also validates my view about onsite power being more permanent in nature.

It’s also true that most existing sites will continue to scale in size beyond what they initailly get from the utility. For example, someone at another Hyperscaler explained to me that maybe you asked the utility for 500MW by 2030 and bridged the gap with onsite power, but by the time that comes, you need 1GW which the utility can’t deliver until 2035, so the turbines will stay. For the foreseeable future, it’s likely that individual sites will continue to add campuses as fast as they can. It’s much easier to add a campus to an existing site than it is to build a new site and so existing sites will look to scale up where possible.

We also have recent proof from XAI that scaling existing campuses is important for model performance. Coherent training clusters are important drivers of model performance and Semi Analysis has written a lot about this.

In addition, rack densities are going to grow by a factor of 10-20x over the next 3-5 years which should lead to exponential increases in power requirements for existing sites that wish to host the newest generation of servers. Take, for example, a data hall that today only has 100kw rack densities increasing to 300kw and eventually 1MW. Without more power, this data hall will sit only 1/3 or 1/10^th full as each rack is now pulling 3-10x more power. This will happen over the medium-term and put consistent upwards pressure on the power needs of existing sites. A site that finally got 300MW from the utility (or built on an existing site with power) can’t simply get another 300MW just because they got connected – they have to go to the back of the queue. BYOP should play a role in helping existing campuses continue to scale.

It’s also worth noting that this trend isn’t just about power quantity, it’s also about power quality. Large training loads pose significant challenges for grid operators as the have enromous surges and drops in power draw that occur in a matter of miliseconds and cannot be forecast by the grid operator. When we’re talking about 1GW+ clusters, these loads could make up a significant portion of the local grid operators service area, and a swing like that could interupt the frequency of the grid so meaningfully that it would bring the entire system down.

One large Data Center Operator that I spoke to who works on the Abilene project noted that one of the Stargate projects would have been ~40% of the load in that particulary electric service area. How can a grid planner possible cope with a 40% load swinging around like crazy?

This poses an issue for those wishing to power their data centers entirely from large CCGTs, also. CCGTs can only run ~80% of the time and take ~30 minutes to go from zero to full capacity. How can a CCGT cope with the load variations of training data centers? They can’t. In fact, META famously dealt with the load variations by having its chips solve for 1=0 (or some other impossible calculation) so that they would always keep working (a gigantic waste of resources) and not interrupt the grid. Microsoft does a similar thing.

Onsite generation, using batteries or intertia, can cope with these swings much better than a large scale CCGT. Most industry folk believe that Large Data Centers of the future will be connected via microgrids where small turbines can play an important role alongside more economical CCGTs or SMRs. A micro-grid like this could also build-in enough redundancy to support the grid and make back-up diesel gensets redundant. (Most people don’t appreciate that Diesel gensets are insanely expensive and almost never get used).

Onsite Gen also has another distinct adavnatage - it’s fast and scalable. Once a fleet of engines is deployed, the Data Center can scale up in a modular fashion by adding more engines and upgrading the balance of plant. It’s much easier than getting more power from the utility which would require going to the back of the queue and also timely and costly infrastructure upgrades.

Lastly, onsite gen might prove to be even cheaper than grid power in some areas and on some timesframes. The significant pushback against utilites passing transmission costs for Data Centers onto customers means that transmission upgrades are going to be borne solely by the data center operator. Inflation in equipment and labor means that building transmission today is way more expensive than in the past. Today’s power curve is extremely misrepresentative of where grid prices are going, and when one compares the gargantuan costs to upgrade transmission to serve these new loads, onsite gen is likely to be much closer to parity with the grid on a 5-10year view.

For example, if you can sign a 10-year PPA with a BTM provider at a 15% premium to the grid today, it might still be NPV positive considering where grid power prices will be towards the latter end of that contract.

I am aware of atleast one deployment in Texas where onsite gas gen was cheaper than the grid (advantaged access to cheap gas + unique situation where the utility tarrif was super high, but it’s an illustrative datapoint of where we could be headed).

My view is that there will be multiple forms of onsite power. First, there will be true bridge power - short-term contracts where the power needs are truly of an emergency nature and returned shortly after. But there will also be more permanent deployments of onsite power which are effectively acting as scalable onsite power plants. The latter will come with long-term, take-or-pay contracts. And demand for the latter seems to be on the cusp of entering a boom era.

Settling the “Stop-Gap” debate: Long-Term Contracts are here

6 months ago, my hypothesis was that we’d see contract tenors extend. Today, you don’t really need to take my word for it. Long-term contracts are here.

Late last year I was at an industry conference and VoltaGrid gave a presentation. I nearly spat my coffee out when I saw the bottom right textbox of the following presentation slide:

15 years… 15 years?! Yes. 15 years. That’s about the same length as a data center contract.

It suddenly dawned on me that there’s a huge disconnect between what the investment markets perceive to be “stop-gap solutions” and what industry folks are viewing as long-term deployments. At the time, the public markets had not yet any proof that these long-term deployments were going to happen as VoltaGrid is a private company and Solaris’ longest term contract with XAI was ~7-8 years (investors remained sceptical about the repeatability of such a deal).

Since my initial work last year, we’ve seen further proof of where the market is going:

• In November, Williams announced a 10-year Data Center contract with 2x5 year options.

  • In February, Williams announced that both of their data center contracts saw the tenor extended to 12.5 years.

“And we've showed the team has done a great job of extending the contract terms on 2 of the projects, but we're seeing additional momentum for that going even further over time.”

- Chad J. Zamarin, CEO of Williams, Feb 2026

• In February 2026, Solaris announced a 10-year contract with an unnamed Hyperscaler with a 5-year option to extend.

• In January/February 2026, Liberty announced a 1GW framework agreement with Vantage Data Centers and an additional 330MW deal with an unnamed customer in Texas. Liberty noted that these are “long-duration” but didn’t explicitly clarify the term.

I strongly believe (based on conversations with private players) that current deals in the market are being negotiated with long-term tenors such as these above, and I believe that we will see a flurry of long-term contracts singed this year which validate my view that BTM power is not a stop-gap but rather a critical long-term power solution. This supports my view on multiples for companies in the space (which I will address later).

Let’s talk about ProPetro

ProPwr was officially launched in December 2024 when ProPetro hired Travis Simmering and Dave Bosco from Dynamis Power. Dynamis is a leading microgrid company operating in the Permian and has >500MW under contract (mostly in the Permian), so these guys are intimately familiar with running and scaling a BTM power business.

ProPwr originally targeted the oil and gas market. In early 2025, it became clear that the data center market would be a big opportunity for them. In May, ProPwr signed their first deal for 80MW in the Permian and in late October they signed their first Hyperscale data center deal for 60MW, which shocked the market. Whilst Permian power will continue to grow, the emerging data center opportunity is significant if they can execute.

On the most recent call (Feb 18), Sam acknowledged that Data Center will soon overtake Oil and Gas in importance:

“Moreover, we are seeing a growing number of inquiries from potential data center and industrial clients. Over time, we anticipate these opportunities occupy a higher share of our overall capacity, driven by both their larger load needs and longer-term strategic commitment. - Sam Sledge, Pro Petro CEO

The ProPwr business currently consists of ~240MW of contracted power. 60MW of the total is with a Hyperscale data center customer whilst the rest is in the Permian. The oil and gas business is out on 10-year tenors whilst the DC business is on an undisclosed “long-term” tenor.

The company’s plan is for 750MW of contracted power by the end of 2027 and 1GW by the end of 2030. Given that they finished 2025 with 230MW under contract, the 1GW by 2030 target is probably conservative. My base case is that they will end 2030 with at least 1.5GW of contracted power. Even at their current run-rate of ~230MW/year, they will surpass their 2030 target by ~400MW. In fact, the company has recently increased its 2027 target from 475-525MW to 550MW and the 2028 target from 675-725 to 750MW. I expect that the below chart may also be updated shortly.

It’s important to put in context how relatively little power 1GW is in the context of a data center. Recent deals from SEI, WMB and LBRT have ranged from ~350-500MW for a single data center. Sam also acknowledged on the call that a chunky data center deal would push their targets way up. I believe that the recent capital raise was to prime the balance sheet for a larger deal.

“That said, to the upside of that, some of these non-oil and gas data center, industrial type projects can be much bigger and chunkier in nature. So one of those can potentially change that time line and that mix very significantly if we're able to capitalize on one of those opportunities soon.”- Sam Sledge, Pro Petro CEO

ProPetro has taken a little bit of a different approach than LBRT and SEI. LBRT and SEI’s approach has been to order a very significant amount of equipment in order to have capacity for the “Mega Deals” - GW+ scale stuff. ProPetro has been a little bit slower and methodical in its apporach, instead preferring to pick off smaller contracts and get equipment deployed into the field faster. I do think that recent comments from the company + the recent capital raise indicate a willingness to push for larger contracts.

For the “why not just own WMB or SEI” askers, I say that I do own both of those stocks. However, I see a much more attractive R/R for ProPetro and it is my top pick amongst the small handful of companies pursuing this opportunity. This is why:

1. Team and Ability to Execute

Travis and Dave built Dynamis largely from scratch to >500MW under contract. They were both in the engine OEM business at Baker Hughes and MTU, respectively, before getting into the mobile gen business. Granted, they know exactly how to build a mobile power business and they’ve assembled a strong team including a handful of ex-employees and ex-competitors. Feedback on the team from competitors (including from Solaris and other private competitors) has been positive.

Travis hit the ground running in his first 12 months. He has taken the business from an idea on paper to 230MWs within 1 year, including a contract with a Hyperscale data center – an elusive target even for some of his competitors that have been in the business for multiple years.

ProPwr is earlier in its life than Solaris and though they’ve secured their first contract, they still need to prove to customers that they can execute. Their original contract is a great starting point but it is admittedly small.

With this 60MW contract, they’ve got a chance to perform and cement their position in the data center market just like Solaris did with XAI. There’s more power at this existing site and at other sites for this customer, and Travis is confident that they’ll win plenty more business if they can execute on this contract. For context, Solaris started with an 80MW contract from XAI in May 2024 which will be >1GW by the end of 2025. Solaris’ team was similarly acquired from a private company with Permian experience.

The hyperscale contract will go live from 1Q26 after an extensive engineering period. I think they have a real good shot at executing but only time can tell. If they can do a good job for their first customer, I see no reason that ProPwr couldn’t scale at a similar pace to Solaris and smash through its 2030 1GW target.

I also like management. I’ve travelled to meet the ProPwr teams and the C-Suite in Midland and Houston. Sam and Caleb are straight-shooters and err on the conservative side. They understand that this is a significant opportunity for the company but have erred on the side of conservatism with managing street expectations (relative to peers such as LBRT who’ve promised huge deals).

Travis and Dave are also great. They’re power experts and they know what they’re doing. They’re also straight-shooters and don’t get caught up in the hype. Exectution is priority #1, 2 and 3. Travis put to me that “demand is the least of his concerns” as it pertains to hitting their 1GW target - execution is everything.

2. Highly asymmetric valuation

Below, I’ve provided a very simple upside analysis for the stock depending on how much capacity we can deploy in the future. I want to stress three points.

1. The 240MW scenario (what we’ve currently contracted) gets us to a stock price above the current level. That is to say: the recent move in the stock price has not even reflected what we’ve contracted let alone upside from here.

2. There’s very significant upside in the stock if ProPetro extends its Data Center business. 3-4x is my base case. I think a 10x over 5-8 years is totally within the realm of possibility if ProPwr exectues flawlessly.

3. Downside is heavily protected by (1) a net cash balance sheet and (2) 15% starting FCF yield to EV based on just the contracts that we’ve signed today and trough permian conditions. Completions FCF was $100m in 4Q25 - it was an extroadinary quarter (low capex, good market conditions, some WC unwind), but FCF could be very siginificantly higher than $150m/year if we get $70+ WTI. Things are already improving in the Permian and FY26 should be better than FY25.

Before I do the SOTP, I want to show the FCF Yield calculation. I don’t use the company’s FCF for completions because it excludes the lease costs. Instead, I calculate my own as EBITDA less Capex (cpy. has significant NOLs so I don’t subtract Tax, but you can if you want).

Below is a quick and dirty SOTP on my 3-year price target. Here are some key points to consider:

1. Multiple of ProPwr: I use 9x EBITDA for the multiple. Some will probably pushback here but this is my explanation: I think there are three suitable peer groups - Data Centers, IPPs and Contract Gas Compression. Data Center companies trade at >20x, IPPs at ~12-13x and contract compression at 9.5x (AROC/KGS). I believe that BTM power should trade higher than contract compression due to (1) better structural growth prospects with less cyclicality, (2) longer-term contracts (contract compression has ~3-4year contracts v.s. power which could stretch 10 years plus). So somewhere between 9.5x and 12-13x seems fair to me. 9x seems right if we see contract terms stretch to 7-10years. JPM is currently using 8x for LBRT’s business. KGS just bought a Permian Power player (~20% of its fleet with a data center) for 7.5x EBITDA, as a floor.

2. Multiple of Base Business: The stock has historically traded ~4-5x EBITDA. The street values LBRT’s pressure pumping business at 5.25x EBITDA in its SOTP. I use 5x given it’s a very similar business and EBITDA is currently recovering from trough levels. It doesn’t move the ultimate dial all too much anyways.

3. Add-back of cash generation: I add back $150m/year of FCF from the completions business and a cumulative $300m of cashflow that the power business would generate if it hits planned timelines over 2026-28. Given that we are subtracting the debt for equipment purchases we ought to credit cash generated also.

   1. A key point of the thesis is that the completions business trades at a single-digit FCF multiple because the market hates the business, but if you can take that cash and deploy it to a high-multiple business then suddenly that cashflow is worth a lot.

4. 2-year target: LBRT anaylsts are valuing the stock based on its end-of-2028 3GW deployment target (see below) and applying a ~8x multiple to it. LBRT stock price largely reflects this today. They have only contracted ~750MW so far (zero deployed). To be clear, 3GW is a target - not a contracted number. I value PUMP on a 2-year-view (Jan 2028) which would be looking at end-of-2030 1.5GW deployment for PUMP. So PUMP will deploy 1.5GW by 2030 (can be easily viewed as conservative given that at the current first-year run-rate they’ll hit 1.4GW), discounted back to a Jan 2028 PT just as the LBRT valuation is an end-of-2028 target discounted back to Jan 2026.

   

ProPetro is perhaps the most asymetric opportunity I’ve found in public markets with serious trading liquidity.

Catalyst

Another Data Center contract. Either an extension of the existing one or a new one of large size. I expect one in 2026.

The stock could be range bound until we get the contract. It may also run up into it like LBRT did. Either way, over 2 years, I am very confident that we will sign some more Data Centers. If we don’t, I don’t think I’ll lose money.

Appendix

JPM’s LBRT SOTP valuation:

Stock Price: $119 | M/C: $2.75b | 2024e P/E: ~10x | P/B: ~1.6x | Target IRR: ~20% |

LGI Homes is a founder-led entry-level homebuilder in the US. It has a truly best-in-class track-record of growth and returns that is rivalled by few. The stock has produced a TSR of ~25% since IPO in 2013.

The three key points to the thesis are:

1. There is significant pent-up demand for entry-level housing caused by a shortage of cheap home construction over the past decade+. Demographics should continue to be a tailwind for entry-level homes as millenials move increasingly into the first-homebuyer category over the next 5 years. The company sells out of ~100 communities today. It should be at 150 by the end of 2025, and has a medium-term target of 240. It had ~10 in 2013.

2. LGI has a unique model in both sales and construction which has led to industry-best return-on-assets over a long period. Despite efforts, most other public builders have failed to rival LGI in its area of expertise. The company is still founder-led and upper management is long tenured in the business. I have no reason to believe that the company’s best-in-class performance shouldn’t persist.

3. LGIH has de-rated considerably on a P/B basis relative to pre-covid (2017-2019) mutliples, whilst most of the large-cap builders are ~equal. This is due to (1) mortgage rates have impacted the entry-level builders disproportiantely, (2) the company is digesting some short-term margin pressures, (3) the company is working through some short-term land inventory issues, dragging on asset efficiency and growth. I walk through all points in the note below and don’t believe any are structural. LGI should produce better growth at similar or better ROA than peers. There is room for multiple expansion, also.

I’ve modelled a couple of different scenarios below. Whilst LGI is confident that margins and absorptions will recover (and they have indeed started to do so), I’ve included a more sombre scenario in my base case to account for uncertainty around mortgage rates. With community count at 150 by Dec 2025, this gets us to EPS of ~$23 and BVPS of ~$134. At a constant P/B of 1.6x and under conservative margin assumptions, the company can increase its BVPS by ~80% by 2027, and hence increase stock price by 80% over that period (~17.5% IRR). Under more positive assumptions, including a faster margin recovery to pre-covid levels and a re-rating to ~2x P/B, the stock could perform much better, ending 2027 with a BVPS of ~$146 and a stock price of $290, producing closer to a ~25% IRR. I suspect that high-teens/low-20s IRR is achievable.

Given the company’s medium-to-long term targets of achieving 240 active communities, LGI could produce such returns well beyond 2027.

The key risks and cavaets for the case are:

1. It’s all about rates, rates, rates. We can make structural arguments for supply/demand until the cows come home, but ultimately LGI is in the business of selling monthly payments. If mortgage rates continue to climb to 7.5, 8%+, LGI’s sales are going to start slowing again. LGI sells cheap, entry-level homes, which are the most rate sensitive, so they will probably slow more than peers. I’ve accounted for this by modelling lower margins in the next few years (i.e. company can use incentives to aid affordability if it needs to) but ultimately the path of rates in uncertain. If mortgage rates keep going up, the stock will not perform well and there isn’t much anyone can do about that.

   1. The positive note is that we are starting from a severe S/D misbalance (pent-up demand) and mortgage credit conditions have been sensible since the GFC, so whilst higher rates will impact demand (see: 2022), it isn’t going to be GFC 2.0. The problem then was artifically high demand + overbuilding. Today we have pent-up demand + underbuilding. When rates went from 3 to 7 last year, it was disruptive, but the company still made plenty of money - it isn’t a case of heads-you-win and tails-you’re-dead; it’s a case of heads you do really well and tails your margins will come under pressure. Also, if mortgage rates actually pull back to 6 or even 5%, there is so much pent-up demand that LGI is like a coiled spring. It had the most amount of leads (inbound interest) in 2Q23 in its history despite having somewhat lacklustre sales absorptions, evidencing the point.

2. Lets not sugarcoat it - Homebuilding is a shitty business, period. There are some builders, be it through niche market exposures, sales models or scale, that have generated satisfactory returns, and LGI is one of them. But generally, the business model sucks - it is capital intensive, the product is a commodity and you need to use a lot of debt to get an attractive ROE. So, when shit hits the fan, you’re stuck with a huge capital position, funded by debt, which can go down in value and wipe out your equity. Shit doesn’t hit the fan to the point of armageddon very often, to be clear, but you have tail-risk owning a builder (like you do with a bank) that you just don’t have owning an unlevered industrial company, for example.

   1. That said, the stocks have always been dirt cheap and have massively outperformed the S&P as a group (nearly 2x since 2010), suggesting that valuations have more-than compensated for the risks. The returns alrogithm for LGIH is simple: If you can produce an ROE of 15-20% and hold your P/B ratio constant, you should be able to produce stock returns of 15-20%. If better times are around the corner, and LGI can return to an ROE of 25% and get a bit of multiple expansion, then we’ll get paid really well owning the stock. LGI will never be my biggest position (unless the fed goes crazy and starts cutting aggresively), but it is an attractive proposition in the context of a long-term portfolio and history suggests that, over time, you get paid handsomely to take the risk so long as you can avoid the big GFC-style blowups (which I clearly do not think we are facing).

Some have asked: why LGI? My answer is:

1. I like LGI’s entry-level exposures given supply shortages and supportive demographics. LGI should be able to outgrow its large-cap competitors who are more focussed on trade-up and/or luxury.

2. I like LGI’s management team and track-record of strong return-on-assets, which I go into in more detail in the note. I also like the differentiated model. Higher ROE at similar P/B multiples should produce better stock-returns.

3. I like that LGI has de-rated and underperformed peers recently for reasons that I view as mostly short-term in nature. I think the market is especially underappreciating that the company’s bloated land position will turn from a big negative (low asset efficiency) to a big positive (big growth with lower capital outlay) over the next 2 years as communities under development finally come active.

I think one will do just fine buying DHI or NVR given that they also skew towards cheaper homes. LGI will probably be more volatile (upside and downside) due to its exposure to entry-level. But it should be able to grow faster (mostly due to smaller size) and has more room for re-rating, so I think the potential upside is better. It is something each individual needs to consider for themselves.

What am I doing with the stock

As usual this is not financial advice. I have a small starter position. I am hesitant to take it up too much right now because the builders have had such a big run and 2H23 estimates have come up across the board, but since 2Q earnings calls we’ve seen mortgage rates re-accelerate and the HMI (builder sentiment index which is a pretty important data point) has started to tick down for the first time in ~9 months. So, I suspect that the builders will have little in the ways of positive surprises for 3Q but potential for negative surprise. If mortgage rates start to come back considerably, I’d probably change my thinking. I’ll be watching closely for both weakness in the stocks and mortgage rates before taking the position up.

That said, if you’re a “I buy and don’t check the price for 10 years” type of operator, the current valuation is pretty attractive on a medium-term view, in my opinion. That is, of course, not advice.

Contents

1. Brief History of LGI Homes and Overview of the LGI Model.

2. The Entry-Level opportunity.

3. Short-Term Housing Market Trends and what to watch (hint: rates).

4. LGI’s recent hiccup, and why it isn’t a structural problem.

5. Financial Analysis, Forecasts, Valuation, Risks.

Brief History of LGI Homes and overview of the LGI Business Model

LGI Homes, in its original form, was founded in 1995 by Eric Lipar’s father, Thomas Lipar. LGI was originally in the land development business, specializing in acreage home sites located 45-60 miles outside of urban centres. After many years of doing so profitably, Eric had the idea of moving downstream into homebuilding.

"We were selling lots and home sites to other builders and we didn't think they did a great job with sales and the customer experience," said Lipar, especially in the entry-level market.

In ~2003, they started building homes, an initiative led by Eric. The idea of focusing on the entry-level really came from Tom Lipar, who after many years in the homebuilding industry, including through the late 90s housing downturn, had decided that entry-level housing was the most resilient.

Early on, having worked as a used car salesman and later a timeshares salesman, Thomas Lipar instilled the importance of direct-to-consumer selling in his son. He told Eric that he needed to spend the first 5 years of his career as a comissioned salesperson. As Eric led the company’s efforts in homebuilding, he took the idea of direct-to-consumer selling and developed a rock-solid system that has really been at the core of LGI’s success. There’s a great interview with Eric Lipar here, where he talks a little about the founding of LGI and the culture of the company. This should be required listening for anyone interested in the company.

LGI’s unique model largely comes down to two differentiating factors. First is the DTC, heavily incentivised sales model. It’s part culture and part systems, as Eric explains, that enables the company to turn its inventory faster than the competition and ultimately produce a better return on assets.

LGI sells homes to renters in their ~late 30s, early 40s and the name of the game is converting renters to homeowners. LGI often joke that they’re a sales organization that sells homeownership, and they also happens to build homes. LGI’s criteria for a new community is pretty simple: there needs to be atleast 50,000 renters in a 30 mile radius. Once they move in, LGI turns on the marketing machine, bombarding sorrounding areas with provocative fliers and online advertisements generating inbound leads to their closing team. Based on that 50k number, management knows from over 20 years of geographically disbursed history that they can convert enough renters to get attractive absorptions and sell-out a community.

The DTC model (that is, generating your own leads and closing your own homes, instead of relying on realtors/buyers agents) is a true differentiator in entry-level for the simple reason that the company is selling homeownership, not homes. First-time buyers are typically “fence-sitters” when it comes to homeownership - it’s something they’ve maybe thought about, but don’t know much about. Or maybe they have pre-conceptions that it is unattainable. Or maybe they just have no idea where to start. They receive a flier in their mail, which makes them consider the possibility of buying, and then they call LGI. Having a DTC strategy allows you to actively convert the fence sitters into customers in a way that just wouldn’t be possible if you were sitting in the office relying on realtors to bring customers through the community. The same is true for move-up/luxury homes, which tend to be more considered purchases. LGI tried move-up, and it didn’t work so well.

To illustrate the point, >50% of the company’s closings come from its direct salesforce (the rest from realtors) which compares to most builders which rely on realtors for probably 80-90%+ of sales. It is a truly different way of selling homes - something that most builders are just not set-up to do physically or culturally.

The system-selling culture at LGI runs deep. Before starting as a sales person, an employee will do an intensive 100-day off-campus training where they learn about the company, its culture and how to sell under its system. This includes direct training talks from Eric Lipar himself. Across the company, there are 10 large manuals (yes, physical manuals) that explain different roles in minute detail. Some of the details in the sales manual are as specific of where to stand in the kitchen when giving tours and the order at which features should be shown. Eric is absolutely meticulous about having his salespeople follow the manual to a tee. If you enter a sales office in Texas, the Carolinas or California, you’re going to hear an almost identical pitch.

When I met with Eric in 2021, he explained their sales process as a “franchise opportunity.” He guarantees his salespeople (pre-selected for general sales acumen) that if they follow the system, they will succeed. It isn’t an easy place to work - it is long hours and management are strict about the system and demanding on targets. It’s sink or swim, but LGI’s salespeople are the best compensated in the nation. He joked that salespeople often offer suggestions about things that they might change to the manual (which has barely changed since 2003), and Eric responds by saying “when other builders are achieving better absorptions than LGI, then we’ll change to script.” He has a point.

There’s a good article in ProBuilder here (from when LGI won builder of the year in 2016) which does a good job explaining for those interested in going a bit deeper. The article also gives an example of one of Lennar’s poor-performing communities, Deerbrook Estates, that LGI purchased and turned around back in 2013. The absorption rates that LGI were able to achieve after rolling out their system-selling relative to Lennar is pretty staggering and illustrated the difference of their model versus the older, stodgier builders.

As it turns out, building and selling entry-level homes profitably is pretty hard to do. Many of the large public builders shied away from the entry level after the GFC as young buyers got hit hard by the crisis. When the entry level started coming back in 2014/15, most builders rushed to open entry-level communities, but most just couldn’t make it work (see above example on Lennar). DR Horton’s Express brand is the only company that has really posed significant competition at scale. NVR has also done a decent job, but still has a much lower absoprtion than LGI. There is competition, to be sure, but the entry-level has proven an elusive part of the market to operate in profitably. Ultimately, the company says its real competition is rentals, which is true.

I could write pages about the company’s model and why it is really hard to replicate for the bureaucratic volume builders, but really, the best proof is in the pudding. Below is a table of absorption rates in 2022 for LGI Homes and a handful of the largest public builders. I couldn’t get absorption rates for all of the builders as they don’t all disclose community count, but you get the point. You can also see ASPs. LGI sells the cheapest homes in the nation and is simply the best at selling them.

The second factor in LGI’s differentiated model is the construction method and land strategy. LGI builds 100% spec homes - no customization, no frills, no straying from the system. Most builders will build some spec, some semi-customized and some fully customized and try make margin with add-ons. Being only spec and low frills allows for (1) the cheapest materials costs, and (2) the most efficient construction methods (free-flow - no disruption and no variability). LGI builds the cheapest homes in the nation. Also, not only is LGI most often developing its own lots (allowing it to also pick up a lot development margin on-top of homebuilder margin), but it focuses on lots that are typically less desirable. John Burns has often joked that if A lots are in urban areas and B lots are in surrounding suburbs, LGI are buying F lots. Naturally, lower competition for these lots leads to higher returns.

These factors together have allowed LGI to generate a return-on-assets which has been the envy of public builder peers. Below is a chart of ROA for LGIH (light blue bars) vs publically listed peers (mostly the top-10 builders). The only company that has consistently produced a better ROA is NVR, who has a unique asset-light model utilizing optioned lots. You’ll see that ROA for LGIH has dipped in recent quarters below its peers - there are a few reasons for this which are specific to LGI which I will discuss later in the note. I expect that in the future the company will revert to industry-leading returns.

Whilst, yes, on average, LGIH has produced a better ROA than peers, I also want to acknowledge that operating at the entry level has certain risks. Firstly, first-home buyers are very sensitive to interest rates and prices such that higher-value builders (i.e. Toll Brothers) are not. The biggest block in the sales process for LGI is having the customer qualify, and when rates go up, it isn’t just that customers don’t want to pay - it is that they cannot physically qualify.

The other point is that land is far less liquid - getting rid of A and B lots in a downturn, in the case of armageddon, is easier than liquidating F lots. Developing the land themselves, as opposed to purchasing finished lots from developers or using optioned land-banking agreements (see: NVR), also exposes them to more land risk, even though it ultimately leads to higher margins.

Clearly, I think that the S/D scenario, higher growth and higher returns compensate for the risks you take. Higher rates has, and will likely continue to have, a disproportionate impact on LGI relative to peers such as Toll and Pulte, for example. This is part of the reason that the stock, and its ROA, have underperformed recently. It is just something to be aware of.

To summarize, LGI has a unique culture and model which allow it to build and sell homes more efficiently than peers, in-turn leading to a higher absoprtion rate and return on assets. I view this as a structural advantage, and it should persist going forward once recent trends normalise.

The entry-level opportunity - LGI is well positioned to grow for many years

Addressing the US home shortage is a task that goes well beyond the scope of this memo. There has been much written on the topic. For those not really accross it, I’ve collected some reseach that lays out the barebones of it and then you can go and check the data yourself. Though some try to argue against US housing (mostly twitter doomers), there are very few people that have actually seen data and disagree that there is a shortage of homes in the US.

• Freddie Mac original study (2018)

• Freddie Mac follow-up (2021)

• Harvard Annual State of the Nation’s Housing (2023)

• Congressional Research Service (2022)

• NAR/Rosen Consulting (2021)

I will spend a little bit of time discussing the entry-level market specifically as it is directly relevant for LGI. The point I want to make here is that, due (mostly) to rising costs, the construction of entry-level homes has fallen considerably over the last few decades (but especially post-GFC), creating a pent-up demand for cheap homes. LGI has ridden the tail of this with its low-cost, high-absorption model over the last decade to grow almost 10-fold since 2013. There is still a significant shortage of entry level homes, and LGI is well positioned to continue to grow in the future.

As we can see, starter homes have decline considerably over time.

There are a bunch of reasons for this both on supply and demand side, though it remains largely a supply story. Entry-level homes have been getting more and more expensive to build. NAHB often refers to the issues as the “4 Ls” - Land, Labor, Laws and Lending. Though the latter has generally loosened in recent times (was more of an issue immediately post-GFC), the first three continue to add costs to the homebuilding process.

On Laws, the NAHB has conducted multiple thorough studies (the most recent here in 2021) of the costs going into a new home. It found that regulatory costs account for almost 25% of the price of a home - a staggering figure. I won’t re-hash it, but the regulatory burden has generally increased over time. This burden is even more accute for entry level homes. It has shown little sign of easing recently, despite increasing pressure from industry bodies such as NAHB for governments to re-consider housing policy. It’s an area that is, generally, likely to get more and more attention in coming years as the affordability problem worsens.

Land has also been a big problem. There was a good (yet controversial) WSJ article late last year that examined rising land prices and their impact on the nation’s ability to build cheap homes. The real issue is land-use restrictions and a lack of public investment in roads, rail and other infrastructure. According to the PRI, growth controls, urban growth boundaries, conservation easements and other public policies that forbid or limit development on large tracts of land are the primary culprit. So are governments that refuse to invest in infrastructure for open areas adjacent to booming metros. Lightbox points out the difficulty in solving such zoning issues due to competiting interests, and why it has persisted so long. The Reason Foundation also chalks the problem up to land-use regulations. All have resulted in soaring land prices, well in excess of incomes.

The US also has an extreme labor shortage in construction. Post-GFC, a large amount of construction labor exited the industry due to the low-level of construction that ensued in the coming years. It was slow to come back, and havs only recently reached pre-GFC levels. Unlike manufacturing, which has automated, productivity in residential construction is deplorable. In fact, it is down from where it was pre-GFC in single-family residential construction. It’s been hard to get more labour into the industry, especially with tighter immigration policies and general shortages across the country. It will likely persist as an issue until the builder embrace more offsite automation (which is not happening at a particularly fast rate).

So what?

The so-what of all this is that there is a huge supply/demand imbalance for entry-level homes, leading to pent-up demand. The biggest barrier is being able to build homes cheap enough. Homeownership rates in the typical first homebuying age group have fallen considerably since the GFC and are still below even historic recession levels, despite having picked up somewhat in recent years. Prices for smaller homes have soared, and construction + inventories for cheaper homes have fallen dramatically.

There is also a demographic tailwind as millenials move into the fisrt homebuyer age over the coming few years. NAHB estimates that the average first homebuyer age is ~36 as of 2022, and LGIH’s target customer is generally 38-42. The ‘bump’ in the green bars below shows millenials aged 26-32. Over the coming ~5 years, these millenials will move increasingly into the first homebuying age bracket. This will only add to the S/D misbalance.

LGI has been part of the solution for this issue over the last decade. The differentiated model has allowed it to build and profit on cheap entry-level homes where other builders previously haven’t been able to (note LGIH’s ASP relative to competitors). It has continued to grow from essentially nothing pre-GFC to be one of the top-10 builders in 2021, all in the face of the above headwinds, proving out the potentcy of the company’s differentiation in spite of tough conditions.

To illustrate the growth, at IPO (2013), LGI had an active community count of ~10 (that is, ~10 communities that they are selling homes out of, and they sell on average ~6 homes per community, per month). Today, the company is at ~100, and it peaked at ~115 in 2020. By the end of 2025, it plans to get to ~150, and it has a medium-term target of 240. There is still a significant amount of growth for LGI addressing the pent-up demand of first-time buyers, and as we’ve seen, the need for first-time homes has only grown greater over the last decade.

However, all of that is not to say that the affordability issues won’t make things tough for LGI going forward. The big question-mark for entry-level is, and will always be, affordability. Right now, affordability has never been worse. I’ll address this and short-term trends next.

Short-Term Housing Market Trends - it’s all about Rates, Rates, Rates

There is no shortage of data and news resources out there for getting across the housing market. Two of my favourites are Housing Wire’s weekly housing tracker (who also has a daily podcast which is great for staying up to date on Housing) and Altos Research on Youtube.

Altos’s Youtube is great. They are a data company and do weekly and monthly webinars based on what they are seeing in the field. Below is the most recent update. It will give you a good sense of what’s going on in the US Housing Market with data.

Again, analysing the entire housing market is beyond the scope of this note, but i’ll summarize the last ~12 months in a few key points.

1. Higher mortgage rates have created significant friction in the housing market. With most of the nation refinanced to a ~3% rate, refi’ing again as a 7%+ rate makes moving homes quite expense. The result has been a huge decline in existing home sales and active listings. Inventory has fallen dramatically.

2. This has created a shift towards demand for new homes. Active listings are so low that there is a real shortage of existing homes on the market. Active listings are ~50% lower than 2019 levels. New home sales have started improving as a result. Homebuilders’ sentiment increased through much of this year and earnings expectations for 2H23 have also risen.

3. Home prices have reversed their decent and have began to climb again. Homebuilders are still using incentives to bring buyers in, weighing on margins, though recent commentary has been that incentives are slowly coming off and new home prices are sequentially moving back up.

4. In the last month or so, mortgage rates have continued to climb higher. Some of the higher frequency data, such as mortgage purchase applications, has slowed and the NAHB index has dipped (NAHB HMI is essentially a sentiment survey of builders and is a really good barometer of field activity). This is sign that the recent leg higher in rates is slowing the market again.

Where to from here?

Right now, it’s all about affordability. The price of building homes is unlikely to fall much other than some modest inflation in materials for reasons discussed above, so its really all about rates. Many people just can’t afford the payment right now. The housing affordability index has never been worse, as of June, and rates and home prices have only gone up in July and August, so it is likely to worsen.

The 30-year mortgage rate is the king of all datapoints to watch for housing right now. Ultimately, given affordability is the key bottleneck (and structural S/D remains supportive), the direction of rates will dictate the entire market. Starts, sales, etc. will all be a function of homebuiler and homebuyer sentiment which will ultimately be a function of rates.

The good news is that LGI has proven it can live in a world of higher rates. After the market slowed in mid-late 2022, the company started using incentives (including mortgage buydowns) and increased marketing spend to improve sales. It was effective, and LGI got back to an absorption rate in 2Q of ~6.1 when rates were ~6.5%. Yes, margins took a hit, but that was also partly a result of higher building costs that are flushing through the PnL and should ease in coming quarters. In addition, the company noted that demand was strong enough to start increasing prices - ASPs in the backlog were ~3-4% higher than closings. I will go into this in more detail below, but want to make the point that whilst 8% mortgage rates are going to hurt, we also don’t need to bet on rates going back to 3%.

I will also point out that the housing market has done a surprisingly good job at digesting higher rates in the past - it just takes time. Historically, the rate of change in rates has been more important than the rate itself. Incomes are now growing at a faster rate than home prices, and ultimately people want to own a house (evidence by record in-bound interest levels). Lets also not forget that rents are now catching up with home prices, which will also aid the buy v.s. rent decision. Maybe people have to wait a bit longer to buy, whether it be for the next pay raise, or change their spending habits for 6-12 months to get there, but eventually the economy will balance as it always has and the American Dream of owning a home isn’t going anywhere.

LGI’s recent hiccup - ROA will recover, path of margins will be somewhat dependent on rates

LGI is facing two problems which have weighed on its ROA - bloated balance sheet and lower margins.

On the balance sheet, LGI suffered from a particular set of circumstances through covid. When the market heated up, absorprtions went through the roof. In addition, the company’s order book filled up so much that it began selling houses far earlier in the construction cycle than it usually would. This led to the company running through inventory much faster than it had originally planned which was compounded by the fact that it paused starts/land acquisitions for a few months during covid. It was effectively a “pull-forward” of demand at the same time that supply went down (paused starts).

As the housing market went nuts, finished lots effectively evaporated as every Tom, Dick and Harry was trying to get land to fill the insatiable covid demand for new homes. Builders, both private and public, were buying land anywhere and everywhere because for a brief moment, it didn’t matter where or what you were building - it was selling and you could make a good margin on it. LGI, who is a strictly returns-focussed company, largely stepped away from finished lot deals at this time with the expectation that the market would ultimately cool down and that the short-termers would get burned. This meant that the company was not able to replenish the finished lots that it was pulling forward, and ran out of homes to sell.

LGI did continued to re-fill their land pipeline, but did so mostly with raw, undeveloped land, as that was all that they could get at attractive returns. This isn’t a problem as the company is used to developing its own lots, however, raw land takes 2-3 years to develop, so it wasn’t available to refill the pulled-forward lots in the immediate future. This, compounded by the fact that supply and permitting constraints through covid pushed the timeline out even further for land development, meant that LGI had an air-pocket in its community count. Below is the quarterly trend of active communities through covid. You can see it took a big dip in 2021/22.

Communities peaked in 2020 at ~117 and proceeded to decline through to early 2022 on abnormally high absorptions. This was not only a drag on sales, but also on return on assets. The company’s inventory position bloated out considerably, but was comprised mostly of non-revenue-generating raw land. As you can see below, the company’s inventory years increase considerably as you had increasing inventories and lower sales.

We can also see that the company’s lot count was heavily weighted towards unfinished. Finished lots declined from 2020 onwards, even though total lots more doubled through to 2022.

This is entirely a short-term timing blip, and the trend is finally coming to an end, with communities being released in 2H23 and ramping into 2024 and 2025. The company expects to be at 115-125 communites by the end of this year, and expects to grow community count by 20-30% in 2024. As the investment into land has already largely occured to support these communities, volumes and sales should increase considerably in the coming two years wihtout much investment needed in the land bank. As you can see above, the company’s inventory years was at 15 at the end of 2022, relative to an average of ~9-10 throughout the past. With a roughly ~50% increase in community count coming in the next 2.5 years, there won’t need to be a considerable amount of investment in inventory to grow into that number. This is a big part of the reason why the company’s ROA has been poor in recent periods, and should subside in the coming periods.

The other thing weighing on the company’s ROA is margins. Prior to covid, gross margins had typically been in the 24-26% range. This increased to ~28% through covid, peaking intra-quarter at 32%. This was a result of rapid price increases and also lower spending on sales and marketing (the company effectively stopped marketing because homes were just selling themselves).

When the market slowed, these factors swung back the other way. Home prices moderated an in-fact started falling with rapidly rising rates, and sales/marketing spend not only increased back to normal levels but also beyond as the company began training for the coming growth community count + needed to hit marketing harder than usual given the poor demand environment. The increase in SG&A and decrease in GP Margins at the same time caused operating margins to fall considerably.

The SG&A issue is mostly temporary. As community count expands in the coming 1-2 years, LGI will absorb the extra investment in sales staff that have been hired and trained recently, and I expect that will return to pre-covid levels in the coming 12-24 months.

The Gross Margin point is a bit of a tougher one. GP fell off a cliff in 4Q22 from ~28.5% in 3Q22 to 20.7% in 4Q22. This was mostly due to the company using incentives (effectively, a fancy way for saying they were cutting prices). In early 1Q, the company noted that it found an absorption rate that it was happy with, and began to slowly raise prices again. From the 1Q23 call:

Throughout 2023, gross margins have already begun to move upwards ahead of mgmt’s original expectations. Incentives are slowly being lifted and ASPs are moving up, and construction costs have eased somewhat. In 2Q23, the company reported a gross margin of 22% and called for another 100bps of improvement in 3Q23 to 23%. Its FY23 guide implies 22.5% at the midpoint, calling for a 4Q23 gross margin of 24% (another 100bps sequentially). That exit rate of 24% would imply a healthy gross margin in FY24.

There are levers that the company can pull beyond price incentives to aid affordability. One of them is building smaller homes, which they’ve been doing. And that has been working. Between 2Q21 and 2Q22, houses <1,500 SQFT accounted for 19% of sales compared to 27% in 2Q23. These homes are cheaper for the consumer but the company makes a similar gross margin (lower ASP and gross dollars, but same margin). It has evidently been working.

Ultimately, the big question mark, again, will be rates. LGI will use sale incentives to target an absroption rate of ~6, and increase prices beyond that. The good news is that LGI has proven the demand is there when they can get the affordability piece right. In fact, in 2Q23, the company generated the most amount of leads in its history, proving that the pent-up demand is there. This is why I think LGI is a coiled spring. The level of pent-up demand/shortages are real and staggering. If we do get a normalisation and/or even easing in mortgage rates, LGI is going to absolutely shoot the lights out, but it should continue to do well even if rates hang in around where they are.

Financial Analysis, Forecasts, Valuation and Scenario Analysis (risks)

First, a quick high-level on the company’s financial history. The growth is truly something. ROA is lousy compared to most industries, but it has consistently run at HSD/LDD through time (good for a homebuilder), and with leverage, ROE has consistently exceeded 20% since 2015. If you are able to re-invest all earnings (they do), then ROE of 20%+ is pretty darn attractive.

ROE is an important figure for homebuilders as investors typically look at them on a P/B basis. So, if you can hold your P/B ratio constant, the stock return will be ~ROE. On that basis, you can also see why a 20%+ ROE is attractive.

Here are my key modelling assumptions:

• Community Count: This is how many communities are actively selling homes. The company has guided to this and it’s a pretty high-visibility number, seeing as these communities have been under development for 3+ years. I have 120 at the end of this year, increasing 25% YoY and then up again to 150 in 2025. This has been more or less guided to. The company wants to get to 240 in the medium term, so I’ve grown communities at +15 in 2026 and +10 in 2027, which is fair relative to history of ~15/20 per year.

• Absorption rate: The right way to think about LGI is that they have a bunch of regional communities surrounded by a certain amount of renters, and they close a certain amount of renters per month in each individual community. Hence, I model homes sold (closings) using the number of communities open in a period multiplied by the absorption rate. I have used 6 going forward, which is at the lower-end of what the company has achieved historically (~6.5 is a pretty decent year historically). I think it is fair an a touch conservative. In better times the company should outperform 6. In 2023, they will do ~5.5 to give a sense for what a bad year looks like.

• ASP’s: Grow at 2% per year. That is conservative versus history (implies negative real house price appreciation). It is probably too conservative but I think most would expect home prices to be pretty benign in the next few years. Ultimately margins are way more important to profits than ASPs so I don’t view it as a key input.

• Gross Margin and SG&A: Below are my assumptions for GPMs and SG&A as a % of revenues, and you can see those versus history. There isn’t so much science in the GP margin - just a gradual recovery back to a historic level by 2025 for the sake of conservatism. Actually, given the company’s guidance for an exit rate of 24% in 4Q23, my FY2024 estimate of 23.5% is probably conservative.

  • SG&A, similarly, remains elevated before normalising to 2025. In 2025, the company’s operating margin is ~in-line with historic averages, if not slightly below. Ultimately, it’s going to depend on rates. If they keep going up, the company will need to use incentives and gross margins will drag. If they stabilise or fall, the company should be able to recover historic margins faster. If absorprtions are healthy, then SG&A will come in line faster.

And this is how profits shake out:

I’ve kept the debt-to-cap ratio at ~the middle of the company’s guidance and used any excess profits to repurchase stock. I might also add that the average communities at the end of 2027 is 170. Given the company’s target is 240, the cpy. can still probably continue to grow beyond 2027. This gets us to the following multiples, based on $115 stock price.

So, if we assume that the stock holds its P/B ratio, stock returns should be ~roughly equal to growth in BVPS (book-value per share). Below you can see BVPS growth, as well as ROE and ROA. The difference between ROE and BVPS is that there are some share buybacks in there as well, but they are roughly equal.

My forecasts are pretty conservative, as you can see from the fact that the company’s ROA going forward is much lower than the past. I’ve also been a bit conservative on the debt-to-cap ratio, which would juice the ROE if I were to bring it up to the low 40s (company guides to 35-45). Under this scenario, stock returns should still be mid-teens at a constant P/B. If we actually assume that the company’s ROA returns to 10% at a 40% debt-to-cap, returns will be low 20s (that is my more optimistic base case).

And, if we assume that the P/B ratio returns to prior levels of >2x (see below), we could see particularly strong returns in the mid-20s. Actually, in better times such as 2019, the stock traded with P/B multiples in the mid-2s, and even up to and above 3 when the market got really excited, so a re-rating to >2x is really not such a big ask. The current P/B multiple is actually consistent with prior downturns, such as late 2018 (it was a pretty bad time for housing). The market is discounting permanently lower returns and growth for LGI and I clearly disagree.

So, in a conservative case, we make mid-teens IRR and robust base case we make mid-20s. In a bullish scenario (god forbid rates come down - lets not forget that current forecasts are expecting cuts) where we get a re-rating, LGI can be a really, really good stock and do 30%+ returns if, for example, we re-rate back to 2.5x. Pretty good upside.

What’s the downside? Well it depends how bad the economy could get at any given point. The lowest level the stock ever got to was 1.15x book in 2022 when mortgage rates were going crazy. Given the situation in the housing market right now, it’s hard to imagine how the company could get itself into a position where it has to impair its book. Also, the LGI has, even in trough year 2023, an operating margin of ~10%, which is a huge amount of cushion for ASP that it can give away before it becomes loss-making. The world would really, really need to fall apart for LGI to lose money.

So, if we trough at 1.15x book, downside from here is ~30% (which doesn’t include armageddon). I view that as a particularly attractive R/R given that we’re playing for 20%+ IRR. If it works, LGI could produce 20%+ returns through to the end of the decade as it races towards its 240 community count target. LGI has been a serious multi-bagger, but it still has to potential to add another couple of bags through to 2030.

I am a long-term believer in DG and their track-record really is world-class. But they have some issues which I found to be bigger than I’d originally anticipated after doing some more work. I think it will take some time to right the ship. The company faced similar problems back in 2016, and the result was a significant and prolonged margin hit, one which the company never really recovered from (margins peaked in 2015, other than one-off Covid boom margins). One could reasonably argue that the current situation is tougher than 2016.

That said, I also want to make the point that whilst I am selling the stock, DG isn’t broken - it is still a perfectly fine business than can produce SQFT growth and has produced a sensational return on capital over a very long time. After the recent haircut, the stock is trading at ~16x forward EPS. Investors will probably do fine owning the stock today with MSD top-line growth profile and LDD EPS growth as they did back in 2016. However, to get really excited, you need to believe that margins are going to rebound pretty quickly to 9%+. I don’t feel confident about that right now.

Contents

1. Cyclical issues

   1. General Macro and non-consumable uncertainty

   2. SNAP roll-off and Tax-Rebates

   3. Shrink issues

2. Execution/idiosyncratic issues

   1. Freight, warehousing and logistical issues

   2. Ill-timed price investments

   3. Ill-timed labour investments

   4. Merchandising Errors and inflated inventories

   5. Distractions from PopShelf and other initiatives

   6. Management Turnover

3. Competitive concerns

   1. The Family Dollar situation

4. Is this a broken story? Let’s look at history…

Cyclical issues

1. General macro weakness and uncertainty in non-consumables

We don’t really have to overcomplicate this and it isn’t really that big of a big concern. Goods spending has slowed and that has hit DG’s core customer disproportionately. In addition, consumers have been trading down to private-label and/or cheaper alternatives. Non-consumable discretionary items hold as high as ~2x margin to certain consumables and that has come under pressure, so they are very important to the company’s margin profile. It is also facing an extra headwind as people over-indulged on certain categories such as home and seasonal goods during the pandemic.

Non-consumable sales/SQFT are elevated relative to the last cycle’s trough, but there has been signficant inflation in the economy, so it is hard to draw comparisons with certainty.

There are also questions around the strength of the company’s recent non-consumables initative (essentially a merchandising refresh). Historically, when the company has done a merchandising refresh, it has been followed by a few quarters of success, and then continued to recede. This is the fourth non-consumables refresh since the GFC - it has long been a thorn in the company’s side, and every time they have proclaimed to have finally figured it out. The jury is still out if the recent increase in non-consumable sales/square-foot was structural or a temporary hit.

Ultimately, spend is cyclical. It will be under pressure for some time and then come back and isn’t that big of a deal, other than to say that pressures might persist a little longer than past cycles for the above reasons.

2. SNAP roll-off and Tax-Rebates

This is a bit of a bigger deal. In March, the SNAP benefits being recieved by some 30m US residents enrolled in the program began to roll-off. During Covid, an emergency allotment of ~$95/month was implemented. It ended for a number of states last year, but up until March 2023 was still being implemented by 35 states.

Howard Jackson from HSA Consulting (source: Guidepoint call) gave the following analysis to help quantify the SNAP impacts for DG.

In 2022, the Covid emergency supplement was ~$37bn on total SNAP spend of ~$118bn. There was an increase to base levels of ~12.5% between 2022 and 2023 plus another 4m people put on the plan, but that emergency allotment is now gone, meaning that total snap spend for 2023 will be around $103-106bn people. Howard estimates that ~10% of Dollar General’s sales come from SNAP benefits, meaning that the impact of SNAP benefits alone could have dragged on DG’s comp by ~1-1.5%. This is a pretty big deal and it is important because SNAP is not cyclical - it isn’t coming back.

The other factor that is impacting spending is tax refunds. They were increased in 2021, and in 2023 returned to 2019 levels. This included a decline in Child Tax Credit from $3,600 to $2,000, the Earned Income Tax Credit of $1,500 to a maximum of $530 and Child and Dependent Care Credit from a maximum of $8,000 to $2,100. It is hard to measure the impact, but given DG’s customer demographic, these sorts of changes can have a material impact. In a similar sense, this is not cyclical and will not come back next year.

3. Shrink Issues

Shrink is not a new issue and is not confined to Dollar General. However, DG is coming off a period where shrink was very low, and 2023/24 margin pressures from shrink are not temporary but are likely reverting to a more ‘normal’ level. DG’s current labour situation, which is widely reported to be inadequate (understaffed, generally) further complicates the issue as lower staff hours are directly correlated to shrink. This issue appears to be here to stay and will continue to put pressure on margins.

Execution issues

1. Freight, warehousing and logistical issues

In mid-2022, the company announced that it would expand its warehousing capacity by adding three new facilities. By the end of the year, things weren’t going to plan and they took a $40m hit in Q3. Channel checks that I conducted on the business suggest that this was just due to poor execution - the company didn’t expand its capacity enough, for one reason or another, and got caught without capacity.

Issues outside of the company’s control had also contributed. Most merchandisers in 2022 suffered from a phenomenon where in 2021/early 2022 they couldn’t get product, so double/triple orders were placed, and then supply chains eased all at once leading to too much product in the channel. This compounded DG’s woes as they got caught with too much inventory and not enough space to store it.

It’s not a structural problem, but it is the first amongst many execution issues from the company’s new management team.

2. Ill-timed labour investments

Dollar General has been criticized significantly in the last couple of years for inadequate levels of labor caused by supposed labor hour cuts in 2022. Whilst this is anecdotal, the reports of labor shortages at Dollar General have been widespread, and have even attracted attention from the OSHA leading to hundreds of stores being fined for unsafe conditions, generally caused by a lack of labor to stack shelves. There have also been labor strikes recently.

Dollar General has been victim, as have most, to the Great Resignation and labor shortages that occured across the economy in the wake of the pandemic. DG is widely reported to be one of the lowest wage payers in the country - a conertstone of their ability to offer low-prices.

However, labor shortages have created significant issues in out-of-stocks and employee churn, which have forced the company to make ‘$100m of investments’ in labor - i.e. hire more people and put wages up. The so-what of this is that labor investments are happening at the same time as the economy is weakening, which will lead to further margin pressures. Given the state of the stores in recent times and increased OSHA fines, this might not be temporary, either - it might signify, in-fact, that the company was over-earning in recent years as a result of not having enough staff in stores, which had reached breaking point.

In 2016, the company took a similar investment in labor which, amongst other things, led to significant margin pressures - pressures that lasted for years. This could be, and in fact appears to be, a similar situation with real medium-term margin implications.

Some have also speculated that the recent $100m is not enough. With nearly 20,000 stores, $100m only represents ~$14/day per store in labor investment. That is roughly 1-2 labor hours per day. It just illustrates that any investment in labor will have significant impacts on the company’s bottom-line.

2. Ill-timed price investments

The biggest shock for the market from the April results was that DG had a comp of ~2% vs Walmart and Family Dollar at 6-7%, suggesting significant market share losses. The most common disgnosis for such discrepancy in comps from people in the industry is that the company’s pricing was not right. Walmart and Family Dollar made significant investments in price in 2022, but Dollar General didn’t.

The company told the market on the December 2022 earnings call that:

“where we're at in price, we don't see a need to invest.”

And then on the the Jun 2023 earnings call, after the comp was reported:

“First and foremost, we are taking action to provide even more affordable solutions and lower prices for our customers. We are doing this in a targeted fashion on the items that matter most, as we believe we can be even sharper within our established target range.”

The company blamed the SNAP roll-off, macro-weakness and weather for its lower comp, but what seems clear is that its pricing fell out-of-line with Walmart, which led to market share losses. This has had a considerably impact on margins, causing a downgrade which implies EBIT margins this year of ~8%.

It’s curious why the company didn’t invest in price and has left industry participants puzzled, as well. Perhaps they didn’t expect economic conditions to deteriorate quite like they have. In either case, it is another misstep from new management that the market has not taken to forgivingly.

3. Merchandising Errors and inflated inventories

The company keeps telling us that inventories are only inflated due to product cost inflation, but I just don’t buy it. Inventories/SQFT were up 14.7% YoY in the most recent quarter against a comp of just 2%. At a time when the company is investing in price, either they are about to see a very significant margin pressure from COGS, or they just have too much inventory.

Channel checks including expert calls from Guidepoint suggest that the company is sitting on a whole bunch of seasonal and home goods inventory. This is partly as a result of over-ordering and inability to distribute it with supply chain issues, and now they are stuck with it as discretionary spend has slowed. This inventory will likely need to be marked down and cleared at some point, which will drag on margins.

Much in the same way as they did with labor and price, the company has refused to acknowledge the issue on earnings calls, suggesting that it might be a skeleton in the closet for further downgrades. It is another misstep in execution and communication.

4. POPShelf

PopShelf is a new initiative from Dollar General. The pace at which they have ramped up the store is really something to see. The company debuted the box in 2021 and hoped to have 1,000 stores by 2025.

Unsurprisingly, that has been pared back citing economic conditions. Explained to me by an ex-employee as a ‘Todd Vasos Baby’ (outgoing CEO), it is unclear what the future holds for PopShelf. In any case, it’s possible and in fact likely that the company got false-positives from PopShelf as the launched the covid during a hobby-spending boom. Its recent pausing suggests that the box is not quite performing as hoped.

It isn’t ultimately so important for the company today in terms of footprint, but it has been slated as a big future growth driver as the traditional box will begin to reach saturation.

This adds to the list of execution issues from the company - ramping a new concept discretionary retailer into the peak of an economic boom.

5. Management Turnover

Much of the recent execution issues can probably be explained in part by management turnover. In late 2022, Todd Vasos (CEO) stepped down, and shortly after, his CFO resigned with no real explanation. Did they step down because they foresaw issues on the horizon? Well, now it is seeming more and more likely.

In addition, Family Dollar recently promoted Rick Drielling (ex-DG veteran) to CEO of Dollar Tree. Rick was largely responsible for building DG into what it is today, and is planning to roll-out the same playbook at Family Dollar as he employed at FD. Rick has also hired a handful of his ex-bench from DG, including John Flannigan and Larry Gatta, who were instrumental in DG’s success.

Recent management turnover and execution issues have created a crisis in confidence for investors, and the stock has de-rated as a result. It will take some time for the current management team to win back investors’ trust.

Competitive concerns - competition from Family Dollar

I’m not going to write about this at length, mostly because it doesn’t concern me so much. Generally, DG and FD’s footprint doesn’t really overlap that much. Also, FD and DG compete with Walmart and other local grocers, not really eachother. Generally, DG is used as a fill-in trip between larger shops at Walmart and tries to price within 2-3% - you get the convenience with a little bit of a higher price, but you don’t mind because you’re just going to buy a few items.

When prices fall out of step with Walmart, people notice, and they are no longer willing to pay the higher price for the extra convenience. DG’s customers are extremely price sensitive - mostly because they need to be. To say that the company’s comps have been weak because of competition from FD is for the most part misleading, especially considering that DG tends to be more rural whereas FD leans more urban. The comp discrepancy has to do with DG vs Walmart, not DG vs FD. Yes, they have lost share to Walmart, but it isn’t because FD has been re-invigorated.

Where I would be more concerned is on new builds. For years, FD has not been focussed on growing footprint. However, FD will likely start to build new stores, meaning that competition for real estate might become higher at the margin. Again, their different focus areas will limit this competition somewhat.

I would temper expectations for square footage growth in the future as a conservative measure to adjust for this, but I certaintly wouldn’t see a re-invigorated FD as some sort of existnetial threat to DG. Walmart did over $600bn in sales v.s. DG at less than $40bn. That is the real competitor and that is the real TAM for FD’s invigoration.

Is this a broken story? Let’s look at history…

There is no reason to think that the DG model is broken. They will continue to grow squarefootage and comps at a modest rate over the medium term. There is good reason to believe that DG is going to face lower profit margins going forward. And to underwrite an investment at today’s price, you need to believe that margins will get back to 9%+ sooner rather than later.

In 2016, the company faced a confluence of issues including general macro weakness, SNAP reductions as well as labor and price investments. It led to significant margin pressure which, really, never quite recovered.

Whilst, yes, the company has invested in a number of other initiatives to increase margins such as private-fleet, there have also been a number of detractors to margins such as higher consumables mix today v.s. 2016. In hindsight, it appeared that 2016 was a reset to a more sustainable base, as opposed to a one-off margin hit.

The current situation is similar. SNAP is permanent and economic conditions are not likely to improve immediately. The company is seeing cost pressures from labor, and at the same time it is having to cut prices to maintain market share. My fear (and suspicion) is that the recent adjustments from the company are akin to a 2016 ‘reset,’ or a departure from the over-earning years from 2020-22 which shan’t repeat.

Dollar General’s business quality comes into question in times like these. Whilst return on capital is phenomenal, the company ultimately is a high-fixed-cost retailer and sells to a customer who is brutally price-sensitive. The ability to protect margins in times like these is simply out of DG’s control. And it is hard to see how such external pressures abate in the near-term.

Conclusion

At 16x P/E, I am relatively unenthused by the stock unless they can expand margins, which just doesn’t seem likely in the current scenario. With ~MSD top-line growth, capital return + sales growth on a constant multiple should lead to a ~11% return, which just isn’t good enough to get me excited. For now, I watch.

I’ve taken my EPS numbers down from $18.90 to $18.10 to reflect the lower sales, which was partially offset by and increase in CWT and CCM margins. It also reflects the impact of removing CFT EBITDA/using the proceeds to repurchase stock, which is relatively neutral from a financial perspective on the year.

That said, I’ve taken up numbers for 2024 and out based on (1) lower sales this year caused by destock means that 2024 will likely experience growth with the absence of destock alone, (2) strong margin performance at CCM and CWT has given me the confidence to underwrite 30% EBITDA margins at CCM in the medium term (I was previously being overly cautious).

That puts the stock at ~15x 2023e P/E after the recent run, which should be trough. For 2024e I’ve got ~$20 and ~$22 for 2025e. That puts the stock at ~14x 2024e and 12.5x 2025e.

I talked about the two main catalysts for CSL re-rating in my original piece as (1) proof of sustainably higher margins and (2) further portfolio rationalisation/capital returns. Both are underway, with CCM margins holding strong despite large volume declines, and the CFT sales/buyback. In 2024/5, we should get the sale of CIT and further affirmation that 30% margins are here to stay, which I think should see the stock re-rate to a high-teens multiple (16-18x P/E if fair in my view).

So, it isn’t quite the gift that it was given that part of the re-rating has occured, but if you put a 17x P/E on 2025 numbers, we get to a stock price of $375 - which is still upside of ~35% from the current price. I’m holding and will probably add a bit more here on confirmation that the thesis is working.

Paying subs can email me guastywinds@gmail.com for my updated model.

Points from call

As usual, I’m not going to summarise the result as y’all are capable of reading it yourself. Press release and earnings call can be found here.

I will just talk about a few points that I thought were important from the result.

1. CFT sale and return of capital - this happened intra-quarter, but it’s encouraging to see that management’s immediate response to the sale of CFT was to return cash stratight to shareholders instead of sitting on the cash and waiting for an acquisition. One of the concerns is that mgmt. incentives are not based on ROIC but instead sales and EBITDA, which risks high-priced acquisitions. Not much else really needs to be said - the financial impact was actually minimal - but signalling is positive and it’s further confirmation for the last 3-5 years that management are responsible stewards of capital.

2. Cyclical - the company continues to face cyclical headwinds from de-stocking and weather, but the worst seems to be done. The cpy. expects sales down low-teens this year, but 3Q will be ‘3-2%’ better than 2Q and 4Q should be up YoY given that the destocking started in 4Q22. Next year, on flat demand and absence of de-stocking, the CFO believes that sales would be up 10% (theoretically). I have modelled for sales of +2.5%. I build in some conservatism around what might happen to end-market demand and maybe a bit of price slippage. But I think we are set-up for 2023 being a trough year. If demand holds in, not only could we get the benefit of higher sales growth, but we might also see a re-stock. There is significant potential for upside suprise in 2024 and/or 2025. We need to have a really bad recession for sales to fall at CCM in 2024.

3. Margin Performance - really, CSL is all a margin story. Margin performance in CCM in 2Q was really quite impressive. With sales down ~15% YoY (all volume), CCM margins only declined from an all-time peak 33.3% to a still-respectable 31.2%. That’s a very impressive result for an industrial company and shows the strength of the company’s variable cost-base. It is also worth noting that we are now 3 quarters through a pretty violent de-stock - volumes in last 3 quarters have been -10%, -34% and -15% - but pricing has held in. Whilst we aren’t totally out of the woods yet on pricing, I am encouraged by the increased discipline that we’re seeing in the industry. CWT margins were also very impressive, increasing YoY from 18.5% to 22.5%. This was achieved in an environment with volumes down 20% YoY. The company continues to improve margins there post-acquisition, which is tracking ahead of the deal model, and also is seeing the benefit of its variable cost base too. Whilst it isn’t a massive needle mover, my LT margin assumptions for CWT are probably too low in light of this. I’m just going to leave the below snippet too from the call which highlights management’s confidence on margins.

4. CIT recovering - the company made some comments around improvements that CIT is seeing in both margins and sales. The business has undergone restructuring in recent years, and the backlog is now even higher than Pre-Covid. It is still under-earning relative to where it has been in the past. This improvement hopefully sets up the company for a sale in 2024, but we’ll have to wait and see.

Here is a high level overview of my assumptions as it stands. I think there is upside to both margins and sales growth in 2024 and beyond.

I’ve lowered my EPS from $20.80 to $18.90. The stock is trading at $207 after-market, which puts it at ~11x updated EPS.

It’s going to be a bumpy couple of quarters as we are somewhat earlyish in a downgrade cycle. My position isn’t full, I’ve been buying down from $260. Mistake to move so early into a downgrade cycle probably but on a medium-term view I think it’s too attractive to ignore. Will buy a bit more here and be buying more aggressively if it gets under $200.

Overview of numbers

Here’s a high-level overview of the numbers v.s. what I had modelled. Not sure what the street had because I don’t currently have access to a BBG. Not going to just summarize the result/call as you guys are capable of flicking through the docs yourselves and I don’t think it adds any value. 8-k and Transcript here and here. But I will add my thoughts.

CCM sales miss was the big surprise. The company called out the destock and some weather-related impacts at ~80-20% weighted towards destock. This has been happening across the whole building products industry (both resi and commercial) and is well understood, though the magnitude was a little higher than originally anticipated. Basically what happened is that when contractors couldn’t get product they started leasing warehouses and stocking up and nobody (OEMs or distributors) knew exactly how much product was out there. Contractors would never do that in normal times; 90% of the time product is delivered direct to site, so this is not an issue that the business is used to dealing with. And so I think we can forgive management for getting the forecasts wrong. Carlisle feel that they have good visibility over that now, and that whilst it will continue to drag a bit on 2Q, it should be done by mid-year.

On end-demand, the company’s channel surveys suggest that sell-through volumes remain solid with contractors and that the downgrade to sales is on the back of inventory clear-out, not end-demand. That said, it’s an uncertain time. Backlogs for re-roof remain healthy, which was the same feedback that I received back in 1Q when I did checks. New construction is obviously a big question mark and I think that management have about as much visibility on that as any of us. Some conservatism is called for with the current state of the world.

The big positive was that margins held in at ~24% given a huge (~35% YoY) volume destock. That is encouraging for two reasons. One, it shows the strength of the variable-cost nature of the business, still able to generate significant profits in weak-volume environments. Two, it is positive to see that, despite a significant volume decline, pricing held in. It’s unlikely that recent price increases pushed through will stick. In fact, commentary suggested that pricing will probably slip a bit this year as raw materials come down (which was the feedback I had in 1Q from the channel). Mgmt. has guided for flat pricing YOY, and given that there was probably 1-2% of carryover pricing, that would suggest some light declines in the backend of the year as the company starts to feel some relief on costs.

Mgmt. are confident that next year, CCM will return to 30% EBITDA margins and that the new target is for 30%+. This still stands to be proven but they’re on the right track, despite a cyclical set-back. Reminder that I had 2024e margins at 28.5% in my base case numbers.

On CWT, sales were similarly impacted by destock/weather but margins came in better than expected, which was mostly just a realization of synergies from the Henry business. Not much else to call-out there and I’m not going to bother with the other businesses, they are both producing good results.

Updating my 2023e numbers

I’ve brought down my sales forecasts for CCM to reflect the new guidance of Down HSD. The company is also expecting margins to be down 100bps instead of slightly positive. I previously had down 40bps so I’ve taken that down a bit. Reminder than I have margins falling considerably in 2024e whilst management expects 2024 to expand to 30% - my base case had always built in a few 100 points of margin degradation to ~28.5% (from 31.6% in 2022 and 30.6% in 2022e).

This lowers my 2023e EPS from $20.80 to $19.10. Outer years remain relatively unchanged but come down slightly also. At the post-market price of $207, this still puts the stock at ~11x P/E.

Paid subs can ping me for an updated model.

Summary

It’s just a quarter and it is the weakest quarter of the year. The path forward is likely to be bumpy, as we already know. We could/probably will see further downgrades to numbers in 2023/24. Margins/pricing are holding in as hoped thus far, which I think is the most important thing to watch for the stock. So for now, structural story is on-track and I’m holding/buying on weakness.

Open House (3288-JP) is a niche homebuilder in Japan that has a special expertise in building/selling cheap detached homes in urban areas. It participates in many segments of the residential property market, though earns the lions share (~60-65%) of its profits from the single-family homes business. Despite its relatively modest profile in t…

Read more

You never quite know how these things stick. Carlisle was out with one in Feb and I haven’t done any checks on how that went. We will see in the next quarterly earnings how it stuck for Carlisle (expectation was that little would stick). Either way, it’s positive to see GAF holding the line on price.

I am not short OLO. I am, however, short-OLO-curious. I am trying to figure out if OLO is a true platform, or a piece of middleware which is being disintermediated. This is just the start of my work (hence Part 1) - I haven’t reached a conclusion yet.. but there’s definitely a set-up here for a good short (I lay out at the end).

These are my high-level thoughts so far:

1. OLO was an an early mover, and in the right-place at the right time when Covid came.

2. Growth is now slowing. Some is cyclical, some might be structural.

3. Competition is intensifying and the tech stack is evolving, which may or may not be good for OLO

4. The payments business will drive revenue/ARPU growth but it will come with a much lower gross profit margin. The stocks is pricing in a lot of payments success.

The rise of OLO

OLO was technically founded in 2005, but it hasn’t been around in its current form for that long. It was originally a SMS-based platform which was bootstrapped for a long-time in the days before cloud-software and smart phones were a thing. In 2014, it had just $100m in GMV. By 2017 it had $1bn and by 2019 >$5bn.

2020/21 was the company’s boom period, as can be seen below.

But even before Covid, OLO was growing well. The proliferation of e-commerce finally found its way to the hopstality sector, helped by online marketplaces and delivery service providers (DSPs). These 3rd parties started as a positive for the restaurants, but then became a threat with their high take-rates.

OLO rode the omni-channel wave by counter-positioning against the DSPs - somewhat similar to Shopify vs Amazon. It focussed on the enterprise market, selling to brands instead of individual locations to give brands a consistent online presence. It had some good success even before covid riding this trend with a good product and some good funding behind it.

The other phenomenon that aided OLO’s success is that the average restuarant tech stack is a white-hot mess. POS is a fragmented market. Majority of systems are still on-premise, and many large brands have franchise models, which may not have historically enforced strict standards around POS systems, meaning that franchise locations might run different systems across the portfolio.

OLO had a large suite of integrations (eco-system) which generally meant that it was better able to integrate into the disparate stack than others; a nice benefit of being a first-mover. Whilst many of the POS players didn’t have a native online ordering module back in 2020, this phenomenon od dispirate stack also made/makes it hard for the POS providers to sell online-ordering on a brand-wide basis (unless the brand has a brand-wide POS system).

Also, on-Prem POS providers have something of a reputation for not being the most tech-forward companies in the world. Being in the business of POS historically meant being in the hardware business, not the software business. This is only just now changing, very slowly.

OLO’s growth obviously accelerated through Covid - if you weren’t digital, you weren’t alive. As the leader, OLO saw a lot of this new business. It now has ~87,000 locations - nearly 30% of the enterprise restaurant market in the US. OLO was in the right place at the right time with the right product, and it did very, very well.

Despite this, trends have changed somewhat since early 2020.

Still growing, but slowing

Growth has understandably slowed in the business since covid. This is a function of (1) slowing trends post-Covid and (2) saturation of online-ordering. I also suspect that increased competition and belt-tightening are playing a role.

Locations Growth has slowed to ~10% YoY.

Ex-Wisely and Payments, which contributed ~$10m and ~$6m respectively in 2022, OLO’s core only grew at ~14%. ARPU (based on average locations, which is not a perfect measure) actually declined YoY when excluding payments and Wisely contributions.

In addition, the company’s guidance for 2023e ($215m in revs, on average cpy. has ~$5m of service revenue) would imply an increased of only 8-10% YoY for the core business depending on how you believe that Wisely will grow into 2023. Payments will contribute “in the mid- to high-teens millions for the year” according to management. That’s a significant deceleration.

The company is expecting to lose 12,500 subways next year (been disclosed), and add 6,000 locations excluding that. Subway was only on the Rails product, so had ARPU of ~1/3 or less of the typical customer. Hence, location count next year will drop by net 6,500, but revenue will increase by ~3% (if Subway was 1/3 of the ARPU of corporate average). So the company is also factoring in some decent increase in ARPU and maybe a bit higher growth in Wisely.

OLO’s profitability over time has also continued to decline, driven by a combination of higher sales & marketing costs + lower gross profit margins.

Acquiring customers has become a lot more expensive. In 2022, the company spent $33.5m on sales/marketing to add 8,000 locations. In 2021, it spent just ~$18m to add 14,700 locations. In 2020, it added 22,300 locations for $8.5m in spend. It is spending more and growing slower. Even Pre-Covid it was acquiring more efficiently - in 2019 it spent $6.3m to add 7,000 locations. In 2022 it spent ~5x more than 2019 to add just 14% more locations than it did in 2019.

And the non-payments ARPU growth, as we saw, has been lacklustre. The NDR tells a similar story. NDR has slowed signficantly post-covid. Some of this is the obvious post-Covid hangover of online-ordering.

The company discloses its revenues from Doordash (makes up a majority of its rails revenues), and that has been declining. In 1Q21 there was a one-off drop in Doordash revenues due to a renegotiation of the contract to much less favourable rates. Revenues from DD peaked in 4Q21 and have declined since. Part of this is explained by Subway coming off the platform. But a 107% NDR for a business which has effectively 0% gross churn is not so great.

The company has ~87,000 locations, and claims that there are ~300,000 enterprise locations. However, just the top 20 brands in the US account for ~90,000 locations. OLO only has 5 of those customers from what I can see - others are either with competitors or build their own tech stack (mostly the latter). When you consider that, the total market is much smaller, and OLO’s true addressable market share is arguably much higher.

Customer additions have become more and more incremental. In 4Q, the company appeared to add just two large customers - one, a virtual kitchen that added rails for 800 stores (rails has a low ARPU), and one Kroger, which added OLO functionality for their sushi and floral shops in 1,600 locations. In 3Q, they added a number of smaller restauarants with ~200 locations, and one with 900 locations, plus a few convenience stores.

This explains why OLO has been pushing into the “emerging enterprise” space (20-100 locations - much more competetive) and C-Stores. At next years rate of 6,000 locations, OLO will be growing at ~HSD on a locations basis. It’s nothing to balk at, but the company’s growth rate has certainly slowed. And when you consider their market share, you’d have to think that they are starting to bump up against limits in a market which should be saturated.

Consider the competition, too. The emerging enterprise/SMB space has been brutally competitive, with dozens of competitors. OLO is a clear leader in the large enterprise space, but it isn’t a monopoly. There are middleware competitors out there like ItsaCheckmate, Onosys, Otter, Deliverect, Cuboh, Mobi2Go, LunchBox, Novadine, etc. They tend to play more in the ‘emerging enterprise space’, which is ~10-100/200 locations. And then there are a bunch in SMB like Chowly and ChowNow. However, some of them have done larger deals and aren’t so small. ItsaCheckmate, for example, has >20,000 locations, including some inspire brands. Onosys has some large customers too, with >200 location brands like Jack’s and Pizza Inn.

As the largest deals run out, OLO is going to have to go fishing in the more competitive ponds, which we are already starting to see in both the brands that they are adding, and the sales expenses/new location.

Locations are slowing, and ARPU/NDR have been relatively lacklustre. Without a doubt, part of that has been a return to more in-store dining. The question is - can they re-accelerate ARPU/locations, or is there something more structural going on?

Competition is increasing and the tech-stack is evolving, rapidly

OLO’s original 3 key products were Online Ordering, Dispatch and Rails. Online ordering is a white-labelled website for ordering that syncs into the POS. Dispatch is white-labelled delivery product using the DSPs (see here). Rails is essentially a middleware layer that connects the DSPs into the POS (so that restaurants don’t have to have multiple tablets and reconclie data afterwards.)

Let’s start with its core product - Online Ordering. There are a couple of points to make.

Firstly, the online-ordering product is now a commodity. OLO’s lead in integrations has been bridged, and making an online-ordering website is not hard to do. Go check one of these things out - they’re very simple. I accept that it needs to be enterprise grade and integrated properly, sure, but there is nothing special about what they are doing. Anybody that sells software to restaurants (lots of them, by the way) has an online ordering module. Whilst most of the new cloud point-solutions are focussed on the SMB/Mid-Market, the products are more-or-less the same. Some checks I’ve done even suggest that OLO’s tech is actually worse than some competitors, given it is so old. There are 10s, if not 100s of options out there from both point solutions and POS providers.

The next point: there is a convergence occuring in the stack. Payments, POS software, back-of-house software, delivery and marketplace tech-integration, point-solutions such as online-ordering, reservations and loyalty as well as data-solutions such as CDPs and marketing (CRM) are all coming together. There has been a TON of M&A.

New players like Toast have been a big driver of this, offering a total bundled platform where the restaurant can do/see everything from one place, though the key reason is that restaurants are increasingly omni-channel. Once upon a time, you went to a counter and paid someone cash for an order, and that was it. And so the POS was just a cash register with a screen. That isn’t the case anymore, and it doesn’t make sense to silo data/software functions.

This has forced everyone in the industry to try become a platform. The old-school POS players like NCR Aloha now have a full line-up of products. Check-out Aloha essentials, for example. It took them a while - they only released the product in 2019 and only released a proper cloud product in 2022. But it’s here now. See below for example of what Aloha Essentials now gives you.

And Aloha Essentials has been growing rapidly. I only have data up to 2021 (last disclosed).

The payments players have pushed into the POS space with acquisitions too, such as Clover (owned by Fiserv through First Data deal), XPIENT (owned by Global Payments, now Heartland), BRINK (owned by PAR Tech) and FuturePOS (now SkyTab, owned by Shift4 Payments). They’ve all been hard at work beefing up their software stack.

Clover was late to online ordering, only adding it in 2020. They have progressively added products. Their line-up is now pretty comprehensive.

PAR has added loyalty/CRM (2021) and ordering (2022) to its line-up for the Brink POS platform - it is building a full platform. SkyTab also all has a comprehensive line-up - it released online ordering in 2020. This is what’s being called the unified commerce platform (PAR has a pitch for it here), and it’s a relatively new concept.

As the pandemic burst onto the scene, guests rapidly demanded new ordering modes and brands added lots of point solutions (like OLO) around their POS, continuing to build on top of an already shaky foundation. But it wasn’t about building a nice stakc, or even price - it was about survival.

Each new order channel introduced different data repositories, incompatible data silos and additional complexity. Most of the POS providers didn’t offer online ordering back then (as we can see above). For large restaurant brands running on outdated client-server technology (while also supporting a myriad of digital order channels), the situation is untenable. Even though some have moved to the cloud, many brands are running on first-gen cloud architectures that mirror the challenges of legacy client-server technologies.

Even integrating different solution providers brings only small relief for restaurants, because after the integration is done, restaurants are still left to deal with multiple technology vendors, each having its own support services, invoicing, communication, training, deployment and solution development patterns and roadmaps.

The omni-channel trend is here and it isn’t going away, with many brands for example trialing pick-up only concepts (and here) to match demands from customers around how they want to get food. As the lines blur between channels, it’s going to become increasingly important for brands to harness data in a way that is useful. The latter point is especially true as brands are fighting to build direct relationships with customers as DSPs are also vying for that relationship (and keeping their data).

Interestingly, the 2023 state of digital report from QU, which is a survey of ~30,000 restaurants, showed that focus on online order has fallen quite considerably between 2022 and 2023, which likely shows saturation. In-Fact, online ordering was a clear #1 priority in 2019-2022, and not in the top 3 for 2023.

In-fact, Cloud Pos was a clear #1. In 2019, the survey found that only ~30% of participants were Cloud v.s. on-prem. It’s crazy to think that, but it’s likely that well more than half of enterprise restraunts are still hosted on-prem.

I don’t want to draw conclusions from a single survey, but it makes sense. Everyone has online ordering now, and it would appear to be at or close to saturation point. And you can see that in OLO’s slowing numbers. Whilst OLO is still winning customers, it’s mostly winning customers from legacy solutions (mostly home-grown) as opposed to greenfields, and it’s happening at a much slower pace.

The current system of archaically stitching together point-solutions, to me, doesn’t seem sustainable, and the industry is focussing on transitioning to the cloud. It is also expensive and inefficient to have multiple of these point solutions. So were does OLO fit in all this? If my POS provider now offers online-ordering (at a significant discount to OLO) as a add-on module for cheaper than I am paying, and I am transitioning the POS to the cloud, do I still need to pay a high price for OLO?

Transitioning to the cloud is hard - I don’t want to understate that. And it is also why the disruptors have focussed on SMB. Not only for the obvious reasons but also that it requires a replacement of hardware, which is very expensive for thin-margin businesses. So it will take time. But it will happen, slowly but surely, and we can see that it’s starting. There are long-term benefits both in cost-savings and data-usage to cleaning up the stack.

OLO bulls will make two key arguments. Firstly, the legacy players’ tech still sucks. That has some merit. I can’t prove or disprove that and would need to do some checks. My sense is that Aloha it isn’t all that bad. When you look online, there are many that like Aloha. And it also has a strong presence in servicing/distribution (for larger multi-unit restaurants, integration/support is generally done by a 3rd party who needs to get trained with a specific player). Alot of what OLO offers is commodity and provided by others.

Also - yeah, if you’re running a 15 year old on-premise software programme on windows XP, it’s probably gonna suck. POS is also something that people love to hate - you either don’t think about it (because it’s working), or you hate it, because it made you embarrassed when it stopped working infront of a line of impatient customers.

The second point is that even though NCR Aloha might be able to offer online-ordering, the restaurant chain is using 5 or 6 different POS systems across the brand, so OLO is still needed to tie it all together. This has certainly helped the company grow. I actually don’t know how true it is that most tech-stacks are disparate, it’s something that OLO talks about a lot, but have never actually seen any data. It’s something I’d like to understand a bit better. In a few checks I have done in the past, it did seem common. But I just don’t really know if it’s 20% or 50% or 70% etc.

However, the latter becomes irrelevant if the industry goes through a large re-platforming at some point in the future - likely during a cloud-transition. It’s happening in the SMB space - just look at Toast, Lightspeed, TouchBistro, etc. In the past, when there were no online channels and data to be collected from credit cards, it didn’t really matter if everyone ran a different on-prem POS, as long as it could all feed data to a central accounting database. So you probably choose the POS based on whoever had more service-distributors in the area. I read anectotally on a message board somewhere that there are ‘Micros towns’ and ‘Alhoa towns.’ This has changed somewhat, and surely at some point in the future brands will see the neccesity to clean up their tech stacks.

I think that most will agree that trying to connect everything in/around the POS causes bloated, over-engineered POS systems. Constantly adding on expensive point-solutions doesn’t seem like the answer.

Beyond Online Ordering, it looks even worse for OLO. Rails is being disintermediated. Doordash, Uber Eats, etc. have been hard at work building integrations directly into POS. Subway built an integration and dropped OLO altogether across 12,500 locations. Why? Well, OLO’s revenue model is that they charge the DPSs. This is because DSPs have to pay significant upfront costs for tablets to send to their restaurants, so they are incentivised to bypass rails. This will take time as building these integrations is hard, but they’ll slowly get there. Also, they’re connected to a dozen other aggregators, too - OLO doesn’t do anything special, and some aggregators have even more integrations than OLO. Infact, OLO recently bought one (Omniovore) because it had a bunch of integrations that OLO didn’t have.

Dispatch is also being disintermediated. All of the delivery drivers have white-label delivery products and have been announcing deals either directly with brands, or with POS systems to integrate into new products. None of this is proprietary to OLO and both of these products, whilst originally innovative, are now a commodity. That’s not to say that people will switch off OLO - but it probably means that OLO doesn’t have pricing power.

So, that brings us to OLO’s dilemma: Is OLO a platform of the future, or a point solution that fixed a short-term issue? Does OLO do anything really well that can’t be replicated by others? With the tech-stack converging to a single unified platform, what is OLO’s role? Do customers care enough about OLO to keep them around at premium pricing as they reasses their stacks?

Reforming retail wrote a really interesting piece on this recently. Below is a quote from an enterprise operator.

One thing is for sure - OLO will survive in one shape or form in the future for a long time. This is a sloooow moving industry. Many of these customers are too small and under-resourced to worry too much about this stuff - OLO works, just leave it.

But to buy OLO’s stock today, you need to believe that there is still quite a signficant amount of growth ahead for the company (numbers later..). I’m just not convinced that is the case.

I need to chat to some customers and do some checks. That’ll be the focus of Part 2.

Everybody loves payments! But what are they worth?

I’d love to sit here and hang shit on OLO for being a sheep and limping into payments, but they do make one good case for borderless payments. Bascially, the idea is that because OLO has huge share in the restaurant space, they can create a borderless payment environment between all of their brands and 85m customers that interact with the brands. That means if I made an account and left my card on file when I bought Maccas, it will be there when I go to buy Five Guys next week. I’m not really a payments expert but I believe this is also a key premise behind Shop Pay (I stand to be corrected).

I don’t really know how unique that is to OLO Pay or if it can be replicated with other payment processors. Like I said, I’m not a payments expert. But I know that friction is important.

There are three reasons why I was originally skepitcal about OLO Pay. Firstly, unlike Toast, OLO serves large customers who would have the scale to achieve lower payment processing terms - there is no way that OLO will have Toast’s margins servicing customers like Dennys and Wendys. Secondly, unlike the POS which is the centre of the universe for a merchant, OLO is just an add-on. Merchants already have a payment processor - why would they change to OLO, other than if not for price (see above - maybe there is merit to that - but it’d still have to be price competitive). Thirdly, whilst the company has GMV of $23bn, much of that doesn’t neccesarily go through native payment processing. If you consider rails, for example - that order is technically executed on the OLO platform - it’s done on the Dash marketplace. It’s unclear how much of the GMV is transferrable to GPV, but it isn’t 100%.

But lets just say that, for arguments sake, OLO Pay is amazing and the large merchants are happy to pay the same price as Toast’s merchants, and OLO can make the similar margins (15% vs Toast at ~20%, but Toast has >4X GPV). Shop Pay is now at ~56% penetration, which has taken ~6 years. That is up from 43%, 49% and 54% in 2019, 2020 and 2021. It has been growing at ~5% pretty consistently.

ShopPay is probably an easier sell, because the merchants likely don’t have pre-existing payments relationships with really attractive rates (they are small). OLO really ought to be a great customer experience.

But lets just assume for a second that it is, and that the company can get to 50% penetration within 5 years and grow ARPU at 4% P.A. And then lets say that it achieves nearly the same economics as Toast on payments (240bps of interchange at 15% GP margin). And then lets say that OLO grows at 10,000 locations in 2024, and then 8,000 every year after that (remeber than next year it loses 12,500 subways, adds 6,000 others for net 6,500 loss).

As for operating expenses, it’s very hard to know how/if this company will scale. Since 2020 it has shown nothing but dis-economies of scale (likely due to competition forcing it to spend more on sales and R&D). In 2021, gross profit grew by 49% whilst operating expenses grew by 130%. In 2022, gross profits grew at 8% whilst operating expenses grew at 23%. Sales and operating expenses grew ~in-line with eachother, but gross margins came under pressure through a mix of (1) payments penetration and (2) more hiring.

So let’s say that the company, from here, grows its operating expenses at ~1/2 the rate of its software revenues (completely arbitrary but shows it can scale). It had ~$173m of operating expenses in 2022 including stock-comp but excluding some one-offs. Software revenue over this period (2022-27) increases by ~80%, so opex will grow by ~45%. This gives us the below outcome:

What I’ve outlined here is pretty much a best-case scenario for the next 5 years. OLO Pay smashes it out the park, cross-sell continues with a decent ARPU increase and locations re-accelerates from +6,000 in 2023 (+8,000 in 2022) to +10,000 in 2024 and +8,000 every year after that to grow at a really healthy rate (to the point that OLO would gobble up pretty much all of the addressable market left).

Then the business finally starts to get some scale on its cost base. It adds nearly $160m of gross profits in this scenario over 5 years, with incremental expenses of just $70m - pretty good incremental return on capital.

The outcome is pretty measely - $52m in operating income for a company with a capitalization of ~$1.2bn (Cash of ~$400mish). So call it an EV of ~$800m. So if things go really, really well, we get to a touch below 15x 2027e EBIT. That’s pretty punchy when MSFT is walking about with a ~20x EBIT multiple.

So you either gotta really believe that the company can rip a bunch of costs out and stay competitive (feels hard with the current rate of increasing competition), or that OLO Pay can penetrate well in excess of 50%. Either way, you’re paying for a lot.

Summary

This is why I hate people saying “you looked at software for a year and everything is down 80%, so must be finding so many bargains!”

No - these things still aren’t cheap. In fact most are still very expensive. This is why I am so attracted to shorting right now… there are so many stocks than can easily go a lot lower, or are likely to go sideways for a long time. OLO is still pretty much pricing in the world at the current price (god knows what it was pricing at $45), despite the potential headwinds against it.

I’m not ready to make a decision yet but there’s a set-up here for a decent short.

A good set-up here looks like:

1. Locations growth continues to slow.

2. OLO Pay penetration rates are slower than expected.

3. We start to see more customers leave OLO, or confirmation from the channel than people are looking hard at their stack and how they can clean it up/take out cost.

More to come in Part 2….

Market Cap: $11.4bn

Debt: $2.2bn

EV: $13.6bn

2023e EPS: $21.20

2023e P/E: 10.6x

2023e EBITDA: $1,559

2023e EV/EBITDA: 8.7x

Introduction

Carlisle Companies is one of the more interesting investment oppourtunities I’ve come across in a while. I think it is both cheap, high-quality, misunderstood and has clear catalysts for re-rating to a more sensible valuation. The 3-5 year risk/reward is very attractive.

I’m not going to give a description of what the business “does”, so maybe read the 10-k first. But it’s a pretty simple business.

Transperancy statement: I started buying this at $258 and I’ve been buying down with all this bankrun overhang. My average is $247. I have a pretty big chunk of my net worth in this stock now and will probably buy more. This isn’t financial advice - do your own research.

Wanna chat Carlisle? Email me: guastywinds@gmail.com. I might share my model if you ask nice enough.

High-level overview in 5 key points

1. Carlisle’s roofing business, which comprises the majority of its construction materials earnings, is high-quality and under-appreciated. It is a leader in an oligopolistic market structure, has defensive qualities due to high (~65-70%) exposure to non-discretionary re-roofing, is capital light and has a mostly variable cost-structure (allowing it to still generate significant cashflows through the GFC). Carlisle should grow due to (1) growing re-roof demand from buildings completed in the late 90s and early 00s that now need replacing (average age of roof = 20-25 years), (2) energy efficiency trends increasing content in the building envelope, such as spray-foam, polyiso and air-barriers. Management is confident that the business can grown ~MSD through to the end of the decade.

2. Carlisle is an old conglomerate that has undergone significant simplification over recent years. The company is still being valued like an old conglomerate, and the value of the underlying roofing business is not well appreciated. This is partly because the stock is not well followed - it has no major brokers, despite having a $11.5bn capitalization - and because the simplification is not yet complete (there still remains two non-related businesses which make up ~20% of sales, which are by all accounts for sale). Currently, Carlisle trades at ~10.5x consensus P/E for 2023e. I think it is worth closer to 16-20x - similar to something like Rockwool or Kingspan. In addition, two peer transactions were recently done (Firestone in 2021 and DuroLast in early 2023) at ~12x EBITDA - today, Carlisle (the clear industry leader) trades at 8.7x. I see that as a significant disclocation.

3. Beyond the conglomerate discount, two other cyclical concerns are weighing on the valuation. There are two points to address:

   1. Volumes: Whilst volumes have been strong in 2022/2021, 2020 was down ~8% and capacity/labour constraints put a dampener on the industry’s ability to respond. Hence, volumes from 2019-2022 only grew at ~4%, versus the company’s pre-covid expectations for MSD. Given that a signficant portion of revenues are re-roof, which is largely non-discretionary in nature, revenue is resilient over-time. My channel checks (including two largest contractors and distributors in the nation) suggest that backlogs remain healthy and there is still a large amount of re-roof activity which has been deferred and constrained through covid. Whilst a slowdown in volumes is inevitable at some point (I model 2024e -4%), I think that concerns for a significant drop in volumes are probably overstated. Sales in this business have only ever declined in one-year since 1991, and that was 2008, from which they promptly recovered within two years. The largely non-discretionary nature of the business gives it natural resiliency over the medium-term.

   2. Pricing: Pricing has surged throughout Covid and is likely up in the region of ~40% since 2019, and as a result, Carlisle has expanded its EBITDA margins from ~23% to ~31%. The market is concerned that the increased profitability in the segment is temporary, however under-the-hood, there have been structural changes. These include (1) ownership changes of major players, (2) management personnel changes at a key competitor, and (3) shifting mindset in the industry as a ‘volume-growth’ business to one that is more value-oriented. I also find that whilst prices have risen, they have predominately done so in the TPO product, which was (and still remains) the cheapest roofing membrane material (I will go into this in detail below). TPO membranes on my estimates comprise maybe ~5-10% of the total roof job, and hence I think pricing power is underappreciated. My view is that margins are more sustainable than the market is giving the company credit for. In addition, the company should benefit from materials deflation this year (mostly oil-linked resins/polymers) which will provide support to margins even if pricing slips a bit.

4. Since the management change in ~2015, the company has a good capital allocation track-record. It has begun paying a dividend, repurchasing stock and exiting business. The share-count has shrunk by over 20% since 2015. The cpy. has also done a few acquisition at post-synnergies multiples of 7-10.5x EBITDA. I expect that the company will continue to repurchase stock whilst its share price remains so cheap.

5. There are two catalysts to re-rating. First, the company needs to sell off its two other remaining businesses. This might take a bit of time, but it will inevitably come and cash will likely be used to repurchase stock or do accretive acquisitions. Second, the company needs to show that current margins are more sustainable. This might take a few quarters or even a year-or-two to prove out.

I think Carlisle has a clear path to do ~$29 of EPS in 2027e even with a partial retracement in margins and softness in volumes in 2024 (I walk through my assumptions at the end). I think a 16x P/E multiple is sensible with room for upside. Pick your multiple - it’s still cheap at 14x P/E. I’m just sure it is worth more than 10.5x P/E. My price target is hence $462 (current price, $223) and I think a 4-year IRR including dividends of ~20%+ is achievable, even when assuming negative volumes and margin compression in 2024.

My upside case gives IRR of 27% and if we assume margins contract back to 2019 levels and 14x P/E, we still get 11% IRR. I think it is a very attractive risk/reward.

Risks/Where would I be wrong

1. Obviously, Macro and all that. If you think the world is ending then maybe don’t buy the stock, or stocks at all. I think that roofing will slow in coming years and I’ve built that into my model but I don’t think the world is ending. The nice thing about the non-discretionary nature of re-roofs, which drives a big chunk of the business, is that revenues lost today are revenues gained tomorrow. Interest rates don’t generally effect how much it hails (I haven’t actually looked at the correlation). Don’t get me wrong - there is still newbuild and some resi - the business is cyclical. But it’s not as cyclical/discretionary as you might intuitively think.

2. Margins. If I’m wrong I think it’s probably because I got margins wrong. I think this is a rational industry which has become more so. Could one of these guys get really aggresive and start pushing product into the market really cheaply? It’s possible if we have a GFC-esque event. I chat about this in detail below. It’s the biggest LT risk, and whilst I have built in ~3.5 points of margin compression in 2024, if we sit here with 22% margins, 2027e EPS will be, understandably, a bit lower. That said, as I illustrate at the end, even with 22% terminal margins and a 14x P/E multiple, our IRR is ~11%.

3. Asset sales/general capital allocation. The company needs to sell these non-core assets to re-rate. If we’re stuck holding them, or forced to sell at a really cheap price, then it will take longer to re-rate. I have no insight into this. The company has sold multiple large assets in the past decade so I’m sure they’ll get it done. I’m similarly not holding my breath on both getting done in 2023. My assumptions on sales prices are (I think) sufficiently conservative. Mgmt. could also blow us up with a big expensive asset purchase, but their track-record is pretty solid.

Why is the market giving us this oppourtunity?

These, I think, are the key reasons.

1. General cyclical conerns.

2. Conglomerate discount and messy past financials (at group level, underlying biz is clean).

3. Not much coverage (no buldge-bracket banks, Jefferies is the biggest broker).

4. Under-appreciation for the structural changes that will likely lead to higher go-forward margins. Market thinks they are not sustainable.

5. Under-appreciation for the structural growth drivers of the business.

Contents and key points of my thesis

1. Brief history of Carlisle and the roofing industry.

2. Quick Roofing Industry Overview - fundemantals of commercial roofing.

3. Why Roofing is a nice business, and how the industry has structurally changed.

4. Structural Growth - Go Green or Go Home.

5. Numbers - Valuation and Forecasts.

Brief History of Carlisle

Carlisle is a 100+ year old industrial conglomerate, whose business was originally rubber manufacturing. In the post-war era, Carlisle followed its peers into ‘multi-industry’ and underwent significant diversification. It was George Ohrstrom who steered them this way (industrial history nerds will know him as the guy who built Dover and Roper). Come the turn of the century, Carlisle was involved in a plethora of commodity manufacturing businesses - all-in, a pretty unremarkable conglomerate. However, standing at ~20% of sales in 2000, was a hidden gem - Carlise Syntec.

In the early 1960s, some chemical engineers at Carlisle were experimenting with a material called Butyl for waterproofing farmers haystacks and from that they developed a new type of material called EPDM. It has both excellent waterproofing abilities and ozone resistance. One brilliant engineer decided that perhaps EPDM could make a great material for roofs, and thus in 1962 the synthetic roofing industry was born.

Prior to this, the way to build a commercial flat roof was asphalt ‘built-up-roofing (BUR)’, where they would lug a hot kettle of asphalt onto the roof and mop it on with layers of fiberglass as membranes. It was expensive, labour-intensive, and bloody hard work! Watch this video below.

Synthetic roofing started to gain a little bit of traction in the 60s due to its lower cost and labour intensity, but it was in the 1970s when the oil embargo caused issues with the price and availability of asphalt that synthetic roofing really took off. The industry scrambled for alternatives, and the synthetic roofing industry gained serious acceptance. Through the 70s and 80s it boomed, and it attracted the interest of many players - both large rubber manufacturers and new entrants.

By the mid 1980s, despite its strong growth, synthetic roofing was largely a profitless wasteland. Firestone, who had once come to dominate the tire business by using excessive scale to price competitors out, employed the same strategy in roofing and managed to gain a strong foothold in the industry. I even heard an anecdote that they used to give away a free flat-bed truck with every 20 roofs (or something like that).

However, it turned out that manufacturing this stuff well and cheaply was actually a pretty tough thing to do. Lots of companies had quality issues, and by necesity, the industry consolidated. By the early 1990s, the many of the smaller companies either went broke or were gobbled up by Firestone or Carlisle. The big four rubber manufacturers, due to their scale and expertise working with the matieral, came to dominate the industry. And then by the early 90s, it become two - Goodyear sold their business to Carlisle and General Rubber sold theirs to Firestone. Even today, EPDM is largely a two-player duopoly, with a small amount of share from Johns Manville. Anyone else who claims to be selling EPDM is just white-labelling one of these three manufacturers.

Even in the early 90s, it was by all accounts a really nice business. In 1991, Carlisle Syntec did $28m of EBITDA on a sales base of ~$200m and on a total asset base of $102m, a cool 27% return on assets. It only spent ~$3m in capital. What a business!

However, whilst EPDM would continue to boom through to the early 2000s, two new materials would come to rival its success. First was PVC - a plastic material that was popular in Europe, and the latter was TPO. PVC had some early success, but ran into quality problems as the original manufactures didn’t tailor the product enough for the US climate. PVC had some benefits over EPDM - namely, it was white, which meant that it reflected heat instead of absorbed (EPDM is black and hence absorbs heat - not good for south-western climates). However, PVC never really took off in the US in the same way that it did in Europe, partially due to (1) cost - its an expensive material and much was being imported, (2) early reputational issues and (3) the rise of TPO.

Carlisle first commercialised TPO in 1998, which was a science project at Versico (Goodyear) when they bought that business in 1993. TPO was a huge success - it was white and significantly cheaper than PVC with similar benefits. Whilst it wasn’t yet cheaper than EPDM, the industry knew that with scale, it could make it so. TPO had some early issues too, as any new material does, but over time TPO proved itself as a good material and it helped accelerate the synthetic market’s rise against asphalt.

The introduction of TPO, of course, attracted new entrants. Synthetic was taking share from asphalt really quickly through the 90s (Carlisle grew from $200m in 1991 to $400m in 1999). The asphalt guys, namely GAF and Johns Manville, saw TPO as an oppourunity to get into the market of synthetic roofing, and they took that. Firestone couldn’t get their TPO right organically, and bought a leading new entrant, but they caught up eventually.

For those unaware, GAF and JM are two of the largest roofing manufacturers across the US, across both residential and commercial. The channel (distribution and contract work) for resi and commercial has a lot of synergy, so it made sense for these guys to try play here and try save their commercial asphalt businesses from shrinking. No non-roofing players managed to get a foothold, and it’s really hard to do organically. I’ll go through that in more detail later. Others came in, but either got bought or failed. Today, TPO is mostly a 4 player market, with much of the share split between Firestone and Carlisle.

At the turn of the century, asphalt comprised still >50% of the roofing industry. TPO was just being introduced and it grew like wildfire, taking signficant amounts of share from asphalt through the early 00s. As TPO got scale throughout the early 00s, it managed to get priced much lower than PVC and even lower than the more mature EPDM, which helped it grow to become what is today by-far the most popular membrane.

Over time, the company also did a few smart acquistions in the polyiso space and grew it organically. Over time, that market consolidated signficiantly such that most of the polyiso insulation market is now owned by the commercial roofing manufacturers. Most roofs are sold as total systems with “system warranties” and insulation can’t be changed without changing the roof, so that level of vertical integration makes a lot of sense.

Below is an illustration of three-decades of success for Carlisle’s roofing business - most of which was organic (up until last few years - I give more in-depth breakdown later). Also pay attention to how the company performed through 2000 and during the GFC - profitability was really solid. Note: there was a shuffling in segments in 2020 - sales were down 7.5% organically, it looks much worse on the chart.

Despite huge successes in the roofing business, Carlisle as a whole was still largely comprised of other commodity manufacturing businesses come 2008. These were exposed to highly cyclical and competitive industries such as automotive (brakes and tires), heavy machinery and vehicles, trucks, industrial components, power transmission as well as legacy rubber manufacturing. The businesses all suffered pretty bad from the GFC, and the company was forced to re-structure both organically and inorganically. Despite incredible growth, in 2007, Roofing was still less than 50% of sales.

In 2009, the company kickstarted a simplification strategy. The first leg was simply getting out of businesses which had been permanently hampered by the GFC. The trucks/trailer business was losing money, for example, and few other businesses were left in such a state of overcapacity that they weren’t hitting their returns on capital.

The company would go on to sell a couple of businesses, though the second wave of simplification would really come in 2013 when Christian Koch took over as CEO. Koch accelerated the simplification strategy, and steered the business towards “high-value industries” - that being, CIT (interconnects - think similar to Ametek), CFT (Fluid control - think similar to Graco) and CCM (Construction Materials - mostly roofing/polyiso). He also began to repurchase stock and pay a dividend at this point.

This was a pretty succesful strategy and got the company’s portfolio to the position that it is in today. Below is a high-level overview of the path taken. The stock performed quite well over this period.

After Covid, the company changed it’s tone a further time to pivot entirely towards building materials. CFT and CIT, the other two assets, would be sold, and the company would become a pure-play specialty construction materials company focussed on the ‘building envelope’. Whilst a vast majority of the simplifaction is complete, Carlisle still has two businesses (CIT and CFT) which are comprising ~17% of sales.

Quick Roofing Industry Overview - Fundemantals of Commercial Roofing

I’m going to keep this brief-ish because there is a lot of information out there on commercial roofing and you guys can go read for yourself. Below should give you a good primer.

On Products

Here is a great resource on the different membranes which I took from some website. I couldn’t find the source so if this is your website, sorry!

Article Explaining The Different Membranes Well

138KB ∙ PDF file

Download

Download

Here is a really good lecture from a Firestone rep. It’s a bit old but pretty good:

Here is a good video on how TPO and EPDM are installed (gives you a sense for the products and the sense that this is a true ‘system’).

I’d say it’s good to watch the first video so you can follow along, but it isn’t abolsutely required. The installation videos are more just for the nerds out there that like to understand the products and how they work, but again, it’s not neccesity.

Market Structure

The three major synethetic materials are TPO, EPDM and PVC. Estimates I have seen suggest that TPO is ~50-60% of the market, EPDM is 15-20% and PVC is ~10-15%, then asphalt is somewhere in the 5-10% range. This isn’t exact and there are big geographical differences, and I get different numbers from different channel checks, but it’s a good proxy.

TPO is manufactured by Carlisle, Firestone, GAF and Johns Manville. TKO, a canadian manfuacturer, also have a small plant. Market shares are available but I estimate that Carlisle and Firestone probably have ~30-35% each, GAF has maybe 15-20%, JM and TKO probably have 5-10%.

EPDM is manufactured by Firestone, Carlise and JM. Estimates I have seen suggest that Firestone and Carlisle probably have 40-45% market share each, and the remainder sits with JM.

PVC is manufactured by everyone other than JM (I think). Carlisle only has a small exposure to PVC, and my understanding is that they only play in high-end product. There are two big private players - Durolast (who just got bought by Firestone) and IB Roof Systems. Sika also has a PVC business.

There are a bunch of players in Asphalt, but GAF and Johns Manville are big players there. Carlisle doesn’t touch asphalt.

So the four major players are: Carlisle, Firestone, GAF and Johns Manville.

Is this a good business?

Those that know me know that I have a building products fetish. Carlisle ticks pretty much all of the boxes for what makes a good building products company.

I have a very simple mental framework for building products that I think works ~90% of the time. It’s maybe a bit too simple, but it works for me as a good starting point. It’s two questions:

1. Is the industry consolidated or fragmented?

2. Is the cost-base variable or fixed?

It sounds pretty obvious, but it works. Consolidated and variable cost is important because this is a cyclical and seasonal industry. If you are a fixed-cost biz, there will naturally be a lot of price competition. And fragmented industries just usually aren’t good for anyone. But being consolidated isn’t good enough. Insulation (fiberglass bats) is a great example. There are only 4 players, but it’s brutally competetive and the returns on capital over the last decade have been pretty average. OSB is the same - swings in price and profits are huge.

Shingles, paint, plumbing fixtures, commercial ceiling tiles, HVAC systems (both residential and commercial), electrical systems (both residenital and commercial), composite decking - all variable cost and all consolidated 3-4 player markets. All high return on capital businesses.

If you’re in a fragmented and/or fixed cost business, you’re probably not making much money. Windows, cabinets, flooring and some specific low-end siding is like this.

Carlisle is both consolidated and variable cost, so that’s a nice starting point.

Barriers - Not impossible but pretty high

One of the underappreciated niceties of the building products industry is that the supply chain is hard to crack - really hard. Go have a look at your house and tell me how many Chinese manufactured products there are. I bet you won’t find many. And then tell me how many products were manufactured by companies that were founded in the last 30 years. Probably zilch if not very few. It’s not because Chinese products aren’t good, and it’s not because they aren’t cheaper.

There are three key barriers in building products. The first is getting accepted by contractors who install the product. The second is getting the distributors to hold your product. The third is getting all the other parties on-board - those being the engineers, the roofing consultants, the architects.

Typically a roofing project will look something like this:

1. Owner has a roofing project, either new-build or re-roof.

2. If new-build, they hire an architect who hires a general contractor, and they will hire roof consultants/engineers (or may have so in-house, if large enough) to design the technical specs of a roof. That spec will then be written, and a vendor will be written into the spec (generally a preferred vendor of the architect). For a re-roof, it looks similar, but there is not a GC/architect involved, it usually goes straight to a consultant or a subcontractor who will involve engineers/consultants.

3. Then the GC hires a subcontractor. The subbie is given the specs. They work with their local distributor to procure what they need. They will also work with a manfacturers rep (a sales person for the roof vendor) who can help them find the exact products that they need for the job. Then the subbie procures and installs the roof.

The key issues are such:

• The key decision-markers are the architects and consultants, but they are hard to sway because they’ve been working with Carlisle/Firestone for decades and had lots of success. They also are not motivated by saving a few bucks here or there. In-fact, often they are incentivised on a % of project cost, so they don’t really care if you’re gonna come in 5-10% lower on a particularly product.

• Architects and Engineers reputations are based on the efficacy of their work. If they spec in a product that nobody has heard of and it goes bad, it reflects poorly on them. It is double-bad (for them and for the subcontractor) if that particular vendor goes out of business and isn’t around in 20 years when they need a warranty claim. Carlisle and Firestone aren’t going out of business any time soon. Chinese Roof Company LTD who’s barely sold any product might not be around in 20 years. That’s not a chance worth taking.

• For the subbies - this stuff is complex. Every roof project is engineered and specified - no two roofs are the same. So you need to get properly trained and certified on a particular company’s products. That is a time-consuming and expensive task. Getting your whole crew certified on a new product is expensive and time-consuming. Also, you’re probably gonna be less efficient at installing the new product than you would be with Carlisle’s system, which they’ve been installing you 30 years. Roofing labour is perhaps some of the tightest in the country and it’s getting worse every year. They don’t have spare time to spend getting certified on products that they won’t install.

• Most distributors are on exclusive relationships. I.e. Carlisle tells SRS in California that we’ll only let you and maybe one other distributor in the state sell our product - that means every subbie that gets written a spec for Carlisle (lots of them) has to come to you/a select few to get product. In return, you can’t carry competitors products (or it might be that you can only carry us and Firestone, or us and GAF, or something like that - it’ll depend on region, the distributors market share in that region and the vendors’ market share in that region). Generally, the top vendors are split between the top distributors in any particular region. Breaking into that is tough. And why are you, as a distributor, going to risk your business (and working capital) on an unproven membrane from an unproven vendor. You probably aren’t.

So, the architects and consultants have no incentive to spec in cheap product. The subbies aren’t trained or familiar with this new product, so there is no incentive for them to try influence the spec and risk their reputation on an unproven new product with no track-record (and take the warranty risk). And if they aren’t getting the product written in spec, then they aren’t incentivised to go out and get the training. So, the distributor won’t hold the product if it isn’t being specc’d, and then if the distributor isn’t holding the product, there is even less incentive for the subbie the get trained on/flip-the-spec on a new product because it isn’t easily available anyway.

No one of these things in and of themselves makes it impossible for entrants to come in, but when you put it all together, it’s very difficult to take share. It’s a classic chicken and egg problem. Not to mention that actually making this stuff is hard if you have no knowledge of the roofing business. And building a plant can take 3-5 years - it’s a huge capital investment. The roof contractor business is local and fragmented - the biggest player has 15%. Certifying all the roofers across the country will take many years, likely decades.

Now, yes, there have been a few entrants into the industry since Carlisle and Firestone - those are GAF, JM and a small plant in 2021 from IKO. What do these all have in common? They are all roofing manufacturers that have pre-existing relationships/trust with subbies, distributors and architects/engineers for commercial asphalt. And even then, it took them a while to get a foothold.

So could IKO be a little bigger in 5-10 years? Sure. What will the roofing industry look like in 10-20-30 years? Well, the membranes might be made of different materials and maybe market shares will go up or down a little, but I suspect it won’t look a whole lot different. You can’t say that about many other businesses.

Competition - Pretty rational, getting better

It is a consolidated industry. History of competition has been pretty rational, evidenced by the company’s smooth-ish margins through time. There have been bumps - in the 2013-16ish period, a fair bit of new capacity got added to the market after the industry returned to strong growth post-GFC which it needed to absorb, so pricing was soft. When I say soft, I mean down ~4%ish in total over 4 years.

This was also somewhat excacerbated by the fact that GAF and JM were tying hard to get a foothold in the market that Firestone and Carlisle had dominated for so long. And as I mentioned in the barriers section, it’s hard to get in. So, to get in, you need to use kickbacks and incentives to move product. For JM and GAF, it was less about absorbing fixed costs, and more about getting acceptance in the market.

What compounded this was the fact that JM and GAF both had declining asphalt businesses, so they had to be aggresive to get a foothold in the growing synthetics market before it matured and the asphalt business went to zero. Hence, GAF and JM have not always been totally rational. You can see in Carlisle’s margins that this has never been an overly competitive industry, but there have been periods of softer margins, particularly around capacity additions, where JM and GAF have forced softer pricing to move product. As you can see below, there was a soft pricing period. Note: The company no longer discloses pricing.

So, my read is that 5-10 years ago, this business was pretty rational but still competitve as it was growing fast and volumes was the name of the game. Today, some things have changed, which I adress next.

There are a couple of points I want to make here.

1. Management and ownership changes at the major manufacturers should be conducive for better price discipline.

2. Increasing ‘maturity’ of the industry is a structural positive for rational behaviour.

3. Industry economics support high-returns and rational behaviour.

Management and Ownership changes at the major manufacturers

In the early 2010s, Carlisle started talking about price discipline more, which started with walking away from a bunch of business in Canada. Though margins tend to fluctuate with raw materials, over-time, the company has increased margins leading up to Covid. This is a mixture of increased rationality in the industry and cost-discipline.

Carlisle has always been the leader in the industry; nothing has changed from that perspective. They are playing to hold price and margin - their position on the situation is clear (from 1Q22 earnings call below).

However, it’s never been Carlisle that was the problem - it was their competitors who put pressure on pricing, as discussed earlier. There are strong signs that this is changing also.

In 2021, Firestone (#2 in the industry) got bought by Holcim. Firestone was historically, according my checks, a more aggresive player in the industry. Already, we have seen discipline from Firestone in the marketplace, pushing price increases in-line with Carlisle and also raising prices on their ancillary system products to be more in-line with Carlisle. Whilst it’s too early to say how they will act in a downturn, we can look to Holcim’s earnings calls for a suggestion on how they might behave.

From 4Q22:

From 3Q22:

So Holcim’s position on margins is clear - they’ve put a bunch of capital to work on US Roofing (including a number of other acquisitions in Malarkey, DuroLast, Adhesives business from ITW and SES Foams to build out the portfolio) and they want to show a return on that. Carlisle has ~28% EBIT margins compared to Firestone at ~19% in roofing for 2022. Firestone wants to get to Carlisle level margins, and it won’t do so by cutting price.

Holcim has also made consistent comments about the ‘pricing power’ in the roofing business. From the 2022 annual report. They understand the business that they are in.

There has also been a big change at #3 player, GAF. From 1998-2018, Carlisle’s roofing business was run by a man named John Altmeyer. John is a legend in the roofing industry, and played a big part of pushing Carlisle towards the ‘system-sale’, focussing on value over price.

In early 2021, John was made chairman of GAF’s commercial roofing business. Then, in January 2023, he was made CEO of GAF commercial.

Many speculate that with GAF’s new leadership, it will be more rational. Here is a great Q&A with John from RoofingContractor Magazine, where John talks about his ambitions for GAF. Below is one important snippet.

John is going to re-invigorate GAF, and it will likely become a tougher competitor going forward. However, I view this as a signficant net-benefit for the industry. GAF, being focussed on customer service and product R&D with internal targets (presumably including profitability), instead of focussed on winning business by being the cheapest, it a positive development for everyone in the industry.

JM is a question-mark, and I really don’t have any insight into them. IKO similar, and they’ll probably be a bit more aggresive, but they are so small that it shouldn’t impact much. To have the big three all being publically vocal about price discipline is a good sign.

It’s too early to say how the industry is going to behave, because all of these changes have occured since 2021, and the industry has been in post-covid boom-mode since. A minor positive is that the 4Q22/1Q23 large channel destock that we’re seeing has not impacted pricing at all (confirmed via conversations with channel).

As a quick aside - and I realise that these sorts of anecdotes aren’t always helpful - but a similar thing did happen in the residential roofing industry. Through the 00s, the roofing industry consolidated considerably.

Over this period, Residential Roofing went from a low-margin (~5-10% EBIT pre-GFC) business, to one that now does 22% (according to OC filings in 2022). It has a lot of similarities with commercial in terms of drivers (high re-roof), channels and cost-base (oil-linked, variable cost). GAF also plays in both markets. It’s just an anecdote, but there’s precedent for building products industries that undergo consolidation and see higher margins as a result in the future.

So, my view is that changes at GAF and Firestone are structural and supportive of higher margins going forward.

Increasing ‘maturity’ of the industry

I already touched on this earlier, so I will be brief. The industry has become more mature. Capacity additions now add ~3% instead of 5-6% that they may have 10 years ago, and volume growth is far more modest now that the ‘asphalt tailwind’ has by-and-large come to an end. Hence, capacity additions going forward will both (1) have less of an impact on overall S/D (and be more managable), and (2) be more measured and modest.

When you have expectations for an industry growing at 7-8%, and especially when you need to grab volume to protect your shrinking asphalt business and establish yourself, you’ll naturally push harder on volumes to get a foothold in the industry. This also makes sense because more volumes can be absorbed via share-gains from asphalt.

Today, volume growth will be far more modest - it’s more a function of construction and re-roofs. I think this is driving the changing mindset of the manufacturers. It’s no longer a sexy high-growth industry. It’s a modest but predictable growth industry, which should bring forward more rational behaviour.

Industry economics are supportive of price discipline

I’ve already touched on this, so again, I will be brief. This is a mostly variable cost business - 85% of COGS are raws - and capital costs are only ~2-3% of sales, which is ~7-10% of EBITDA. This is not an industry which has significant fixed-costs to cover when volumes fall. It also has a natural counter-cyclical cost structure through oil-linked commodities. Hence, when volumes fall, it isn’t economic to use price to get volume for volumes sake because decremental margins on volume are so much lower than price.

The pushback might be: Yeah, well, the industry has shown periods of price compression in the past, so is it irrational?

Well, I would say that - actually - that’s not neccesarily the case. In the past, price weakness has come in times of volume growth, and weak price has been used as a function of ‘growth’ (and in JM/GAF case, getting acceptance in the market), not neccesarily to maximise short/medium-term profits. It was also used to protect their share from the shrinking asphalt business. And that isn’t neccesarily irrational - it might well be the long-term sensible decision. Actually, if you look at GAF, it probably was - now they have a foothold, even though it wasn’t so profitable to get here, and now they can focus on improving the business and improving margins.

On a pure price v.s. volume debate, now that things are more stable, I’d expect that the manufacturers will be incentivised to act rationally. Of course, theory doesn’t always translate into practice. However, it’s better to have economic theory on your side and not against you, such that it is in high fixed-cost industry like fiberglass insulation, which has no hope of holding pricing when capacity utilization slips.

The other point - the cost of TPO per square foot is ~$0.85c (I called a distributor and got a quote). The total cost of a roofing project, including all ancillaries, labour, contractor profit, overheads, insluation, etc. is $10-15. So the end project is somewhat insulated from the increasing price of TPO. Also, TPO is still cheaper than PVC and asphalt by a factor of 50-100%, so there is not really substitution risk.

Below is an example BOM from a project I found online in 2017. The Genflex TPO membrane (line-2 - Genflex is Firestone) is only a fraction of the materials, even before labour.

And then lastly - distributors don’t want pricing to go down, that’s bad for business as it devalues their inventory. Contractors don’t really care that much, from the checks I’ve done. So this isn’t a super transperant industry like automotive where you go to a dealer and get a price - it’s a bit more nuanced than that in terms of its overall impact on the end customer. This, I think, aids the industry’s pricing-power.

In summary, there have been some pretty significant changes in the industry which have taken place over the past few years, which have been over-shadowed by increasing margins on large price increases. We need to see how they play out, but I think just automatically assuming that we go back to 22% pre-covid margins is probably not right. Can the industry hold 31%? Maybe not. But even if we give a few points back, Carlisle is way too cheap.

Quality of financials

It’s obviously high-margin and high ROIC. In 2022, the roofing business did ~$1.2bn in EBITDA. It did ~$135m (3.5% of sales) of capital spending (including a new TPO plant and ISO facilitiy, so an elevated year - historically it averages 2-3%). The company no-longer discloses asset base by segment, but in 2020, it had a total asset base of ~$2bn. Hence, it has an operating return on assets of likely well north of 50%. Theoretically, in normalised WC environments, FCF conversion should be ~90-95%.

There are a few attractive financial charecteristics of the business:

1. Variable cost-base. 85% of COGS are raw materials. You get a double-whammy effect in downturns of buying less volume and cheaper prices (generally, oil prices fall in recessions). This is how Carlisle protects its margins in downturns.

2. Capital light. It doesn’t need much capital to grow - that’s evidenced by the 2-3% of revenues spent on capex over time, despite MSD-HSD growth rates.

3. Leverage is modest at ~1.4x EV/EBITDA.

4. 70% of the roofing business is re-roof, and that is medium-term non-discretionary. You can kick the can down the road with a patch here or there for a few years if you’re really tight on cash, but you ultimately have to replace a broken roof. A key catalyst for this is often the warranty - if a roof goes out of warranty, you don’t want to hold onto it when it gets smashed by a storm, or inevitably leaks and causes internal damage to the structure of the building. So, although sales volumes might fluctuate a bit in a down-market, you can hold onto your hat knowing that in 3-4 years sales will be fine and indeed probably a bit higher, and re-roof projects don’t go away - they simply get deferred. Such was the experience in the GFC, as volumes recovered quite well and rather more quickly than other building products (see above volumes around GFC period).

Unattractive charectieristics of the business:

1. Roofing has a fair bit (~30%) of exposure to new-commercial, which is cyclical. The waterproofing business, which is a smaller percentage of profits (~15% today) has ~50% exposure to residential and ~50% exposure to new-build. This is much more cyclical (and will be down this year in the double digits). Hence, the total business is cyclical. It’s just not as cyclical as people might think. But yes, rising interest rates is generally not good for the business in the short-term. Nor is tightening of lending standards. Hence why the stock is down nearly 20% in a week.

2. I can’t say with certainty that the industry will remain rational. I believe, through my work/checks, that it will be (obviously that’s why I own it). But there is a risk that IKO or JM get really aggresive. This isn’t a monopoly and you can’t say with certainty how pricing will play out. That’s why there is the oppourtunity, I guess.

In summary, yes - I think this is a very nice business. It generates a lot of cash, it generates a little more in good markets and in bad markets it generates a little less. It operates in a stable competetive set, and I think I can understand that pretty well. It sells a product that the world needs and is perhaps one of the least tech-disruptable products out there. Software might eat the world but it won’t keep rain from seeping into your building envelope. It has some strong non-economic drivers, meaning that I can say with, again, a reasonable degree of confidence that volumes will be higher in 5-10 years (on average) than they are today, because storms will continues to ruin roofs, and we need to replace them.

It’s not microsoft, but it’s a nice business, and I can say with much more confidence what this company will look like in 10 years than I can say about Microsoft.

Growth - Go Green or Go Home

There are two key structural drivers of growth for the company.

1. Growing re-roof demand

It’s pretty simple. Over time, the stock of roofs has increased, meaning that replacement demand has a tailwind behind it. The typical life of a roof is 20-30 years - some say 20-25. As you can see below, a considerable percentage of the roofing stock was constructed 2000-2012, meaning that in the 2020/25-2030/35+ period (based on 20-25 year average life of a roof), Carlisle will benefit from a tailwind of re-roofs. It’s impossible to quantify but management believes it is supportive of ~MSD growth for the roofing business.

2. Green Buildings

I’m not going to write a green buildings thesis here (I might do so in another post). Just go read about it. This is a good resource to start. There is significant fiscal support and private sector support (net-zero pledges) behind decarbonizing buildings. This is a decent article, ableit a bit more focussed on residential. The AIA 2030 commitment is good to read about too. Point is - if we want to get to net-zero, we need to decarbonize buildings.

There are lots of ways to take emissions out of buildings. If we think about building emissions as just those that are emitted at the building (not including the emissions in the steel, cement, etc.), the the biggest culprit is HVAC. HVAC systems are not just energy-hungry, but they also generally rely on fossil fuels. One of the easiest ways to save energy in the building is to use the HVAC system less intensely, and to do that, the building envelope needs to be more efficient. That could mean more insulation, a white roof that reflects light, a tighter air barrier to keep the cool air in, more coatings/adheisves to seal potential gaps, etc.

Over the years, Carlisle has transformed itself from a “roof membrane vendor” to a “buildings envelope company”. They’ve grown both via acquistion and organically to be the largest PolyIso insulation manufacturer in the US (the most popular type of roof insulation), and they’ve also added ancillary products such as air barriers, spray-foam insulation, water barriers and roof coatings. All of these aid in making the envelope more efficient. A few good snips below that illustrate the company’s products.

The biggest and simplest one is that the intensity of insulation is increased - i.e. instead of putting X amount of insulation per squarefoot of roofing membrane, you’re now going to put 1.2 or 1.3x. This trend has already been happening and will continue in the future. Here is an example of how PolyIso intensity has grown over the years.

Spray-Foam is also a secularly growing area due to its increased efficiency relative to traditional bats and/or plastic boards. It’s growing in both residential and commercial applications. The industry has been consolidating over the past 5 years as well. R-Value requirements in buildings are going up, driven by codes that will only intensify, and spray-foam is a great way to achieve higher R.

Beyond the product content story, there is a higher-level trend for retrofits. It’s very hard to quantify, but the push to make existing buildings more efficient could increase general activity levels. A 15-year old building with an asphalt roof might accelerate the roof replacement to a white reflective TPO membrane with 2x polyiso insulation, for example - a decision which is not driven by the current condition of the roof, neccesarily. ROIs on this stuff are uncertain and I’m cognizant that the EU’s push to retrofit existing buildings has by-and-large not been succesful (very hard to get the ROIs). At the margin, it may help, and policies/incentives/carbon prices in the future might help.

So, all of this is very hard to measure. The point is that content will grow over time for energy saving products, and it will grow at above GDP levels.

Putting together the re-roof oppourtunity + increased content from energy efficiency, it seems totally plausible that the company can achieve its ~MSD targets going forward.

Other growth drivers

There are some other smaller drivers which might add a point here or there. Those include:

• Further share gains from Asphalt. Asphalt has been a share donor for the past 50 years, but has now shrunk to ~10-15% of the market. The share will slowly continue to bleed over time, but it won’t be the same tailwind that it was in the past. But the company could still take a few points of sales in the coming years.

• Labour efficiency enhancing products. Labour in the roof contractor market is tight and only getting tighter as not many youngsters are coming into the industry. Carlisle has a number of time-saving products which are sold at a premium, but help the contractor net-save time/money. The biggest one is the 16-foot TPO plant. You can read about it here. It is still ramping and my recent checks suggested it hasn’t properly hit the market yet. But the company can drive some mix benefit through R&D.

• EU/Global expansion of EPDM membranes. I was recently at a trade show in Sydney and was surprised to see Carlisle there. They haven’t sold a single square-foot here (going back to difficulty getting into the supply chain), but it’s a possibility. They have a position in Europe and recently expanded their Germany facility.

• Metal Roofing - its is only 10% of sales but growing faster and has a much higher (~2x) ASP.

Numbers - valuation and forecasts

Share Price: $224.32

Market Cap: $11.4bn

Debt: $2.2bn

EV: $13.6bn

2023e EPS: $21.20

2023e P/E: 10.6x

2023e EBITDA: $1,559

2023e EV/EBITDA: 8.7x

Forecasts

Below are my forecasts for CCM:

For CCM, I have 2023 slightly higher. It is a mixture of carry-over pricing from 2023 and some slightly negative volumes. That is mostly due to -25% volume in 1Q23e, as there is a big destock currently going on. The rest of the year is modest LSD growth, as channels suggest that backlogs are still healthy and expecting sell-through growth. Then, in 2024, I have volumes down 4%, except for the reversal of the destock in 1Q which won’t persist. So whilst it looks like I have volumes +2.5% YoY - that actually means volumes down ~4% on a normalised basis. See below.

I then have price going -3% in 2024. If we have a recession, raws will come down and I suspect that the manufacturers will have to give a bit of price back. Who on earth knows how much they’ll give back. The worst ever pricing year that this business had was 2009 when prices were down ~3.4%. I think am being overly conservative here - I don’t actually think prices will fall that much, but want to illustrate that even with conservative assumptions the case still stacks up.

I then have pricing flat on goes forward basis, and volumes returning to a ~4% CAGR (right at that MSD guide, and not giving any benefit from bounce-back in demand post-2024). I think my volume forecasts are pretty conservative, unless the world ends.

Also, it’s worth noting that whilst volumes have been robust in recent years, they have only grown at ~4.1% CAGR from 2019-2023. And on my assumptions, the volume growth from 2019-2027 will be ~3.8%. So I think that is fair.

Then, on margins, I have price dropping through at -100% incremental in 2024e and volumes at a 0% incremental. This resets margins to 28.3% (down from 31.6% at peak), and they gradually grow a little bit from there in the future at ~10pts/year. In 2020, the biz did ~26% margins.

One quick thing to note - margins look like they jumped a lot from 2019-2020. There are two reasons. (1) oil prices. (2) re-ratement of the waterproofing business into its own segment. I estimate that on a comparable basis, 2019 EBITDA margin was probably ~23% and 2018 ~20%.

I’d also note that the business has improved its cost base over this time as well - its isn’t just price - it has opened new, more efficient plants, and acquired businesses which have raw material synergies. So my assumptions are for ~5 points of structural margin improvement out of the ~8.5 points that we’ve seen so far. In-fact, the company thinks that EBITDA margins will be up in 2023e over 2022, but I’ve modelled them slightly down just to be conservative.

Here are the assumptions for the other businesses. I don’t go into a huge amount of detail but you can email me (or leave a comment) if you want to see my model. I have modelled out CIT and CFT. I assume that they got sold at ~the same multiple as Carlisle group company, which means that it doesn’t change the multiple whether you have them in there or not. Getting rid of them is not so much about unlocking value as it is about simplifying the story for investors. There is some upside I think to the multiples but it isn’t a huge needle mover.

And then here is group earnings.

The last assumption is buybacks. I keep ND/EBITDA constant at 1.2x and funnel everything into buybacks. I grow the share-price at 15% P.A. Obviously if the shares increase by more than that, it will impact terminal EPS. But if the shares go up by 15% P.A. I’ll still be happy.

I use 16x P/E on $28.9 of EPS in 2027 to get the target price of $462. There are some small dividends of ~$3.5-4/share. That gives you an IRR of ~21% by capitalising it at Dec-26 with 16x NTM 2027e EPS. There’s upside to this if you think margins can hold the line a little higher.

Let’s get bearish, baby!

Okay - lets get bearish margins and say we go back to 22%. What does that look like? Keeping everything else the same, this is how our EPS and valuation comes out:

I’ve also gotten brought the target multiple down to 14x P/E.

Well, we still get a respectable 11% IRR.

Let’s get bullish, baby!

Now lets see what we get if we hold margins at 31% and the market gives us the nice ESG 18x P/E multiple.

That gets us a nearly 27% IRR.

So - base case = 20% IRR, bullish = 27% IRR, and if I’m wrong on margins and multiple, we still make 11% IRR. The disruption risk in the roofing business is nearly nil, and the ability for companies to oversupply it is also low, given that we have visibility over volumes for 3-5 years, unless we have another GFC (but we can just pull down lines and right-size the biz, like they did in the actual GFC).

That’s a nice risk/reward for the long-term investor.

Summary

I think Carlisle is a very nice business which is not appreciated quite as such. There are still question marks around where margins shake-out, but I think the risk/reward is very attractive for this highly cash-generative business with structural tailwinds.

Hence, I am long.

The last statements that we got was March 31, 2022. Cash and Cash equivalents at that point was $335m. So in the last year, the company has burned ~35% of its cash - a little more than the ~$65m that I had estimated. I presume that there’s lots of severance etc. in that, and the company also had a bunch of money is short-term securities which may have hurt them.

Anways, nothing too material, but a rare bit of disclosure from the company and confirmation that it continues to incinerate shareholder funds with what seems to be no prospect of ever generating profits.

Shrinking to greatness

The company was founded in 1860 as a cork manufacturer and had early success expanding its cork products into different industries such as building materials, packaging and machinery. By the end of the 1900s, Armstrong was a disparate collection of industrial products. In 2000, Armstrong entered chapter 11 bankruptcy protection for legacy asbestos liabilities which created a significant restructuring of the portfolio, including the sale of many assets. It emerged in 2006 with just four building products business lines – global ceilings, global wood flooring, global resilient flooring and global cabinets.

Soon after re-listing, the GFC hit and volumes fell drastically across the business. The global flooring and cabinets businesses were significantly affected, though due to the strong cashflows from the ceilings business, Armstrong sailed through relatively unscathed. The fact of the matter was, however, that Armstrong had a collection of poor business, and one very good one. Cabinets was subscale, competitive and unprofitable. Flooring is a brutally competitive industry with global imports and huge fragmentation. The ceilings business, however, was making good returns and required very little capital.

As markets started to slowly recover in 2012, Armstrong began to move forward strategically. It first sold the cabinets business, and then spun-off flooring (which has since declared bankruptcy). Management made a bet on international ceilings throughout this which didn’t pay off. The market structure was just too fragmented, and the company realized that they wouldn’t reach the same returns profile as the US. The international business’ profitability was revealed at the 2016 investor day – it comprised 33% of sales and contributed just 7% of operating earnings. It turned out that whist the ceilings business was profitable, it was actually comprised of a very profitable North American business, and lackluster international business.

In 2017, it concluded that consolidation was required internationally to improve the market structure, and that it wouldn’t be the consolidator. The international operations were sold to Knauf for ~1x sales. By the end of 2017, the company had a single, sole-focus – North American Ceilings. Armstrong has essentially ‘shrunk to greatness’, and it has shown in its stock price, moving from $40 to $100 shortly after the sale was completed.

Quick Business Overview

The business comprises of two ceilings segments today – mineral fiber and architectural specialties.

Mineral Fiber (~$900m in sales (73%) with 39% EBITDA margin)

Mineral fiber tiles are the most common type of material used in commercial drop ceilings today. The North American mineral fiber ceilings tile has been long dominated by just 3 players – Armstrong (~60%), USG (~25-30%) and CertainTeed (~10-15%). The tiles themselves aren’t particularly difficult to make – a mixture of clay, Kraft paper, resin and mineral wool is mixed together, heated and pressed – though achieving the acoustics, fireproofing, insulation factor and aesthetics is a bit more difficult.

Barriers to entry in the industry are rather high. Ceilings are generally a specified product by acoustical engineers and architects. Armstrong has long-standing relationships with architects, designers and contractors across the country, and the brand is synonymous with ceilings. These parties tend to be rather loyal, and Armstrong has all sorts of retention programs based on loyalty rebates, leads, etc. to keep their ecosystem strong. Specifications can be influenced at any point along the chain. The contractor is a big influence point, as they could change the spec for a competitor if that is the tile that they prefer to hold, which is generally driven by the distributor.

For this reason, distribution also plays a big role. Armstrong’s presence in installed base (for replacement volumes), as well as its efforts to drive specifications means that Armstrong is the most demanded tile every year. This gives them the power over distributors to enforce exclusivity. Hence, the best distributors hold Armstrong. Contractors want to install the products that their distributors hold, and don’t want to install multiple brands as they are incentivized by loyalty/volume discounts. This not only means that Armstrong is far less likely to have the spec ‘flipped’, but contractors also act as influencers in their favor.

The barriers to entry are observably high. Rockfon, a European company, tried to enter the US market in 2014 and hasn’t managed to take any meaningful share despite have 30% of Europe. They sued Armstrong for exerting monopoly power and were ultimately unsuccessful. One can read about that here, which gives insight into the importance of Armstrong’s distribution network.

Whilst historically Armstrong has perhaps invested in its products a little less than it should have, it has still always been a cut above competition. The coy. invests heavily in the salesforce. Channel checks paint a picture of a company that people like dealing with. Their service is instant and exceptional. It has the widest range of SKUs for architects, creates leads for its best contractors (sometimes the salesperson is in a project early with the architect, and can pull a contractor in) and does its own logistics, meaning that product delivery to distributors is fast and reliable. Armstrong also runs a ‘just-in-time’ manufacturing schedule, meaning that its distributors don’t have to hold significant amounts of inventory, which is a big benefit for them. This all combines to create a flywheel of influence from Armstrong which is hard to penetrate.

The consolidated nature of the industry combined with the low value of the product relative to the overall project, as well as the labour cost to install it, gives Armstrong considerable pricing power which has been examined over time. Armstrong has also innovated in its metal grid products as well as ceiling tiles to make them easier and faster to install over-time, along with digital tools that make ordering and estimating faster more accurate, reducing time and waste. This has given them additional ability to push price.

Architectural Specialties (~$350m in sales (27%) with 17% EBITDA margin)

Architectural Specialties is a collection of aesthetic ceilings products that would be seen in the atrium or reception. It includes, felt, wood, metal etc. and have a much higher value than mineral fiber tiles. This part of the business has largely been put together via acquisition over time it is a very fragmented market but highly synergistic with Armstrong’s existing distribution and architect relationship network, as well as manufacturing scale/expertise.

The market is rather fragmented with lots of $10-30m revenue players. As the projects are so customizable, it isn’t uncommon to find a reasonable amount of differentiation in aesthetics between different players. Armstrong believes that the category as a whole is growing at around 5% as these more aesthetic products have been growing in popularity. Whilst manufacturing is generally not integrated, management find that there are lots of ways that they can help the sub-scale companies’ profitability whether it be sourcing or upgrading equipment. Armstrong believe that it is a $2b market, with a long runway of growth ahead in a fragmented market.

Armstrong has grown this from a small niche of $80m of sales in 2015 to $350m of sales this year. It has grown largely via acquisition, though prices paid have approximated ~7x EBITDA post-synergies, with some organic growth too. The ability to continue to roll-up the industry is high, and the company is targeting 1-3 deals a year. It also expects to grow the business to $500m by 2026 entirely organically, with any M&A being incremental to that figure.

The Bull Case – a decent business with a free option on growth

There are a couple of key points to the bull case, and I’ll go into them in a bit more detail below.

1. Armstrong’s base min fiber business is a high-quality franchise with a strong market position. It is very cash generative. It should be able to grow a bit, with a free option on further growth.

2. Architectural Specialties is a decent business which can grow quite a bit, and Armstrong is the logical consolidator of the space.

3. The stock is cheap even if you only believe half of what management thinks it can do, and downside seems somewhat limited. It has a good track record of returning capital, so you should get paid.

Point 1 – Min Fiber should grow a little bit

Since splitting from flooring, Armstrong has become a lot more focused and invested heavily in its products. Over recent years, it has introduced more premium products in mineral fiber with better acoustical and environmentally friendly properties. Below is an illustration of the company’s vitality index.

These products come with much higher price points, as well. Given that ceiling tiles are very cheap, its possible to get significant value for new product innovations, even if they remain small in volume terms. Below is an illustration of the new products and their price difference to the traditional base model.

Armstrong uses average unit value or ‘AUV’ to illustrate the price/mix effects. As one can see, a significant increase in AUV was examined starting from 2017 when the new products started to work their way through the channel.

The company stumbled in 2020 as Covid impacted the business, but it bounced back in 2021 and 2022 with extremely strong AUV growth. 2021 and 2022 are anomalies due to significant pricing measures taken to offset inflation.

The pandemic, Armstrong would argue, has increased its opportunity to expand AUV. The company has introduced another new product line, Health Zone, which brings benefits to combatting viruses and so forth. The company believes that these products have been seeing a lot of attention in its education markets, where schools are looking for ways to make classrooms safer.

The company is so optimistic that it has upgraded its forecasts for AUV growth in its 2026 plan from 5% to 5-8%. As the company had already proven its ability to do 5% in AUV prior to the pandemic, the additional couple of percentage points seems like a free option. Below are the company’s 2026 targets.

Min Fiber volumes are still down somewhere in the region of ~10% since covid for a couple of reasons. Firstly, supply constraints have halted construction/remodel projects, particularly on the labour side. In addition, the office market has remained weak with return to work slowly coming back. Armstrong believes that it can achieve 1-2% in volume through to 2026 just from market recovery. In addition, it has also launched a number of digital initiatives to go after the long tail of volumes out there that it tends not to touch with its sales force, which it thinks can bring an extra 1-2% for a total of 2-4%, + 5-8% on AUV.

This feels a bit bullish considering that the market hasn’t grown volumes for almost a decade, but I’ll go into that in a bit.

Point 2 – A/S can grow and is reaching a scale where it’s starting to move the needle

Historically, A/S hasn’t really moved the needle. However, in 2020/21 the company did a number of deals which significantly increased the scale of the biz. I’m not going to spend too much time here because I’m not an expert, but I’ll just note a couple of things.

1. Armstrong’s strategy seems sensible matching its extensive sales/distribution network with amazing products.

2. The market is significantly fragmented, and Armstrong is the natural consolidator.

3. It is paying good prices for the businesses, esp. considering that private equity is probably not a natural buyer of these assets.

In 2022, the business should do ~$350m in sales. Armstrong believes that that can grow to $500m organically by 2026. Over time, that business has grown organically at HSD due to (1) growing category combined with (2) market share gains. Combine the fact that the business’ volumes are still down and recovering from Covid, and that $500m target (HSD growth) in 4 years doesn’t seem so unreasonable.

Point 3 – The stock is so damn cheap if you believe management

If you believe management’s 2026 targets, the stock is an absolute steal. I’ve built a rough model which illustrate the company’s targets below.

Here, I’ve used 7% AUV growth, 2% volume growth and 10% growth in A/S. I also have included no acquisitions. This rest the company to $1.8bn in sales. At the stated margins of 40% and 20% respectively (which were both achieved pre-covid and so shouldn’t be an issue) it gives group financials as follows. Also note that the company has a target of $2bn sales, which is presumably anticipating some incremental acquisitions.  

As one can see, this puts the coy. on 6.4x ex-cash P/E in 2026 at today’s price of $77. The company has a history of repurchasing shares and paying dividends. My assessment for how this stock plays out over the long-term are as follows:

Bull Case

Management does what it says it’s going to do and continues to show growth. The market gives it an appropriate 15-20x multiple that a company growing EPS at 15-20% with such high-quality cash generation deserves, and the stock is a 3-4x over 4-5 years.

Base/lackluster Case

Organic volumes and A/S growth are weaker than expected as the economy takes a while to recover and potentially goes through a recession. AUV continues along at ~4-5% as the healthy building initiatives aren’t as successful as hoped. Even using 1% volume growth, 5% A/S growth and 4.5% AUV growth, we get 2026 ex-cash P/E of 8x. The stock probably just maintains its multiple of 14x or maybe derates a bit to 12x. One should make 50-75% in 3-4 years. Nothing to balk at but you’ll take that.

Bear Case

Big recession. AUV growth goes away. Volumes fall and A/S doesn’t grow. Stock de-rates to 10x earnings. Over 3-4 years, the cash generated will repurchase stock to offset drop in earnings and multiple and the unlucky investor will probably just make no money and own a cash-chow on a 10% earnings yield. Not the worst bear case outcome ever.

Conclusion

Obviously so much more stuff can go wrong here. But this is a boring as hell business with very low threat of disruption from competitors or substitution. It also isn’t as cyclical as one might assume, given that 75% of the business is replace/remodel.

The way I see it: If management do what they say, I’m going to make a lot of money. If they kind just keep doing what they’ve been doing, then I’ll make good money. If they totally mess it up and/or the economy sucks for a long period of time, then I probably will just break even. These are the types of bets that I like to take.