We line these releases up on price and performance and ask where each sits on the curve. Through that lens Muse Spark reads as a clear positive: competitive on capability, materially cheaper, able to take share in developer and enterprise workloads. So we nod, mark Meta as more competitive than we thought, add a few billion to the bull case and move on. That is not wrong, but it is just small because price-for-performance is a power user’s scorecard, and we are power users. We go find the best model and switch when a notably better one appears. But we are a rounding error. Most people have never used a chatbot the way we do, and most never will go looking for one.

And here is the part that matters: people cannot tell you what they want if you ask them. What a person actually wants hides in what they do, not in what they type into a box. So the immodest question underneath the modest one is not which model wins, but what a winner even looks like. Why have we, who imagined Samantha and JARVIS for a century, lowered our ambition to a better answer engine? What people actually may want, whether they can name it or not, may just be a presence. The kind of thing that captures attention rather than charging for it (which just so happens to be the only business Meta has ever known how to run). Something that knows you rather than answers you. And a presence cannot just be the best model, because that title changes hands every quarter and locks in no one. It has to be something that cannot be copied or exported to a competitor’s website.

So let yourself imagine what that could look like. Something that has been quietly present long enough to have learned you. It talks to you the way you actually want to be talked to, blunt on the days you need it blunt, gentle on the days you don’t, because it has watched a thousand of your days and knows the difference. It picks up the joke you were making before you finish it. It knows which friend you go quiet around and why, remembers the thing you said last winter that you have already forgotten, and knows never to surface it at the wrong moment. It does not sound like a product. Over enough time it sounds like the version of you that you could never manage to write down.

To get to this, all you need is a good-enough model sitting on the one thing that cannot be reproduced: years of being present in a real life. It is a relationship that accrues, and it gets less like software the longer it runs. A better model releases next quarter, cheaper and sharper on every benchmark, and it changes nothing, because you do not leave the thing that knows you for the thing that scores higher any more than you would trade a friend who finally, actually gets you for a stranger with a better résumé. This is the type of lock-in no competitor can buy their way into, because it isn’t a feature. It’s time.

And you cannot start accruing that time in a lab or a search bar. You can only have it if you were already living where people live, which is a very short list. Meta is the one name where every piece lines up. It already lives in the places people actually spend their lives: WhatsApp, Instagram, Messenger, the threads and group chats that, for most of the planet outside the US, simply are daily life. It holds the deepest record of revealed behavior anywhere, not what people searched but what they did, who they talk to, what they buy, going back more than a decade. It has no existing franchise whose economics an aggressive agent would blow up, so it can build the most aggressive personal agent it can imagine with nothing to cannibalize. And it has the cost structure to scale frontier-adjacent intelligence to three billion people, funded by an ad machine that already prints, while everyone else has to charge for the tokens to pay off their data centers. The model may have been the last piece it was missing. Muse Spark looks like the signal it is arriving.

Strip all of the above away and you still own one of the best businesses on earth at an undemanding multiple, throwing off enough cash to be fine on its own. That is the floor. What you get for nearly free on top of it is a call option on Meta being the one company structurally able to build the thing people have wanted from this technology since before it even existed. You do not have to believe it happens. You only have to notice that almost no one is pricing the possibility, and that the possibility is not crazy.

Maybe Meta fumbles it. It has before. But that’s not the point. The point is that we are all grading the model now, but no one seems to be asking how far, truly far, this could actually go in Meta’s hands.

Disclaimer: Nothing in this article constitutes investment advice. These are my personal views and reflect my own portfolio decisions. Do your own due diligence. I am long the stocks mentioned in this article as of the date hereof.

Subscribe now

Meta Platforms (META)

Meta is facing one of the largest dislocations hiding in plain sight. It has traded down in sympathy with other companies that have dedicated an increasing amount of capex to AI-related spend. The returns on this spend is uncertain, so the market has sold off these companies or ascribed them lower multiples as a result.

However, Meta is different from its high-capex peers for a few, interrelated reasons, which I believe leads to an outsized opportunity. To be clear, my thoughts on Meta here aren’t to say that the other hyperscalers will fail to do well, they are instead aiming to show that META stands apart right now on a relative basis.

Meta’s AI spend goes directly into making its core products better, and therefore the ROIC on that spend is much clearer than what the market thinks.

When most people think about AI capex, they picture companies like Microsoft, Alphabet or Amazon building massive cloud infrastructure and racing to commercialize new AI products. To be clear, I am not an expert on every nuance of what each of these businesses is doing. But at a high level, the returns on their spending depend on things largely outside of their control, and that is why I think the market is discounting hyperscalers generally.

Microsoft has staked a significant portion of its cloud growth narrative on OpenAI and other enterprise customers consuming excess compute, many of whom are burning cash at a high rate and facing fierce competition of their own. Amazon is spending incremental billions on AWS infrastructure betting it can maintain or grow market share against Alphabet and Microsoft, both of whom are spending just as aggressively to take that share away. Alphabet is spending billions on Gemini, a product that is directly competing against OpenAI, Anthropic and others for the same users and budgets. These examples aren’t exhaustive, but show how a large portion of the hyperscalers’ success depends on external parties behaving in specific ways, on markets that haven’t fully formed and on competitive races that are just beginning.

Meta’s capex story is structurally different. The infrastructure it is building feeds directly into products that over three billion people already use every day. The vast majority of each dollar spent on AI compute translates into a better feed, a more accurate ad auction, a more compelling Reel, a less-invasive ad. These improvements don’t require anyone outside the company to change their behavior for Meta to capture the value.

And because the AI infrastructure directly optimizes the core ad product, improvements show up in revenue on a short feedback loop. Management can determine what is working in real time, and we can see the returns appear in next quarter’s numbers. Meta has internalized the entire value chain, which makes its capex dollars both more legible and more defensible compared to its peers’ spending and therefore an outsized opportunity.

Meta isn’t facing a sunk-cost problem. The moment it believes the return on spend is no longer there, it can turn off the spigot without significant risk.

When a company builds out proprietary cloud infrastructure and stakes its entire sales positioning on a grand roadmap, it becomes very difficult to reverse course even when the economics start to look shaky. The capital commitments create their own momentum, and walking them back carries cost: monetary, competitive and reputational (look at the blowback Apple received when it stepped back from its AI efforts). For example, Alphabet cannot meaningfully pull back on AI spend in the near term without every “Search is dead” headline coming back to life and the market reading it as throwing in the towel on Gemini. Microsoft cannot slow Azure AI investment without signaling that its cloud growth story is shakier than advertised. Amazon cannot blink on AWS infrastructure without handing share to two rivals who will absolutely not blink in return. Each of these companies may well find that the returns justify the spend. However, in the meantime, they are on a trajectory, both from a competitive positioning standpoint and stakeholder optics standpoint, that is difficult to alter regardless of what the near-term returns actually start to look like.

Meta is not similarly locked in. Its AI spending serves a core business that generates enormous free cash flow independent of any particular AI outcome. If Zuck decides tomorrow that the return profile on incremental GPU spend or hiring the next AI Rockstar has deteriorated, Meta could pull meaningfully back without cratering its forward revenue base or signaling strategic failure. The social media flywheel keeps spinning, the ad auctions keep clearing and the user base keeps engaging, all while shareholders see a reversion to higher historical EBIT margins and free cash flow conversion. What you are left with is a company that can chase returns aggressively when they are there and pull back rationally when they are not. That optionality directly leads to my view that Meta’s shares are underpriced vis-à-vis other hyperscalers.

Even with very conservative assumptions, Meta’s valuation gives us a strong margin of safety.

Even if you believe the AI spend delivers only a fraction of what management expects, you are still buying one of the most profitable businesses in the world at a price that requires no heroic assumptions to justify. My conservative case assumes revenue growth decelerates from 22% today to roughly 10% by 2030, capex stays elevated, margins hover near historic lows and Reality Labs continues to burn $19 billion a year with zero recovery. Exit at 25x earnings, which is modest given the market has averaged around 22x over the past five years and 20x over the past ten, and META should trade at a premium to that (still below its own historical median). Under those assumptions, you are still seeing 10%+ IRRs. And those assumptions are probably too harsh: in a world where revenue growth slows materially, capex should come down and margins should recover, neither of which my conservative case gives credit for.

The AI ad uplift, WhatsApp monetization, Meta glasses and any improvement in Reality Labs are not in these numbers. In the base case, you can easily see 20%+ IRRs for this established mega cap — the thesis is not that complicated. Mega cap tech is one of the most under-owned asset classes in the market right now. Sometimes the best opportunity is hiding in plain sight. What Alphabet was in 2025 is what Meta is now.

Carvana (CVNA)

I usually avoid writing about Carvana, so let me be direct: this is one of the highest-conviction positions I hold, and the recent selloff has not changed that. The short reports making the rounds are, in my view, noise. They are mainly recycled skepticism dressed up in new conspiratorial packaging, aimed at a company that has been the subject of short-seller attention for years and has consistently made those bets look foolish.

Here is what actually happened when the latest 70+ page report dropped: within minutes, people on X were declaring fraud confirmed and the end of Carvana. Let that sink in when you think about where the shorts are coming from. You cannot read 70 pages in a few minutes, let alone think critically about them. These were people who had their conclusion before they read a single word. Meanwhile, serious longs, who have bet massive portions of their net worth on the name and have followed the company closely for years, sat down and went through each substantive claim methodically. Several publicly addressed these claims with actual evidence, and in the process caught clear factual errors and other misrepresentations. That is the difference between people who understand a business and are thoughtful investors, and people who are performative investors.

Volume is not rigor. Anyone can stitch together screenshots of publicly available court documents, half-baked charts with messed up scaling and random large fonts and insinuations (“well isn’t that a coincidence?!”) into a document long enough to look intimidating. What you cannot fake is the depth of knowledge that comes from actually studying how this company creates value, how its unit economics work and why the moat is real.

That’s enough on the shorts. As mentioned, they are noise. You can refer to Austin Swanson’s article to see a more detailed rebuttal, in addition to sell side analyst reports fact-checking key claims.

What I want to focus on instead is the business itself.

The original thesis was always that:

• this is a vertically integrated platform that is systematically reinventing one of the most inefficient and crappy consumer purchasing experiences that exists;

• incumbents in this space are structurally disadvantaged compared to Carvana; and

• Carvana’s model gets even better as it scales.

Carvana is eating other large competitors’ lunch and is growing at 40%+ year into a market in which its share is still a rounding error.

No competitor is growing volumes anywhere close to Carvana’s pace, and the gap is only widening. Carvana has spent years putting physical infrastructure into the ground and building a skilled workforce that would take any rival years to replicate. As the business scales, Carvana passes the gains back to customers through better buying prices, higher offers on trade-ins and increased selection, which drives more volume, which creates more scale and which funds more gains to pass back to the consumer, leading to even more volume, scale, gains and pass backs. This is a clear example of a flywheel.

Arbitraging Clarity

Let me give you a framework that I think about a lot: “arbitraging clarity”.

It comes with the territory when investing that the highest potential opportunities often come with the most perceived hair. Regulatory concerns, nascent competitors, AI disruption fears, accounting questions. Pick your poison. Every great dislocation I have ever seen has had some version of this. The hair is the reason the opportunity exists in the first place. The confusion and noise, it turns out, is something you can arbitrage.

This is the framework behind every position I discuss in this piece and the basis of every position in my portfolio as of today.

An example I keep coming back to is Luckin Coffee. Actual fraud. Confirmed, admitted and textbook accounting fabrication. The company got delisted. The obituaries were written. It was over. Except it wasn’t, because if you had put down your fucking spreadsheet and gone outside you would have found thousands of real coffee shops (coffee stalls) full of real customers growing at an insane rate and completely indifferent to the fact that some executive went rogue and told everyone that they grew sales at 50% instead of the 30% that they actually grew. This is not to say the accounting doesn’t matter or that we should ignore potential irregularities. What I am instead trying to say is that the business did not care about the scandal, and that’s what mattered in the long run. The coffee didn’t lose its flavor. The business just kept winning. There were actually people who made 25x+ returns following the scandal — they were the ones who arbitraged the gap between how scary it looked in that moment and how simple it actually was when you cut through the near-term noise.

That gap between perceived complexity and actual clarity is where the money is. Always. Numbers live inside computers and on paper. Businesses live in the real world. A spreadsheet and a 10-K is a simplified representation of reality, not reality itself.

With Meta, the market looks at a massive step up in capex mixed with AI fears and sees a company burning cash on an uncertain bet. What I see is a company spending money directly into a core product used by three billion people every single day with returns that show up in next quarter’s revenue. The feedback loop is real and working. The hair is the capex number. The clarity is that every dollar spent makes the ad auction more accurate and the feed more compelling for users who are not going anywhere.

With FTAI, a report dropped and the crowd decided the thesis was broken. What the crowd missed is that the facilities are real. The engines are real. The managers are competent. Customers are flying real planes that need real repairs and they are paying FTAI real money over others to do it. You can go visit the shops. You can talk to the customers. The business is not a spreadsheet story it is a physical, operating, cash generating enterprise that was sitting there in plain sight while everyone was busy reading a slide show arguing whether the TTM Adj. EBITDA numbers should be +/- X% based on how you account for captive engine sales and exchanges (which, by the way, all of the serious longs already knew about and could think through/explain, if you had asked them).

With Shift4 Payments, the market sees a levered balance sheet, a founder who left to go to space, a crowded payments sector and a recently-closed and uncertain large acquisition. What I see is a cheap, dominant player in hotels and sports venues, a best-in-class capital allocator buying back stock aggressively while the stock lags and an acquisition that just opened up an entirely new international runway. The hair is obvious. The business underneath it is just as obvious if you bother to look.

With Carvana, a 70 page short report drops and within minutes bears on X have declared fraud confirmed. What they are missing is that the lots are real. The cars are real. The customers are real. The employees are real. The flywheel is real. You can drive to a Carvana lot right now and count the fucking inventory if you want to. You can talk to the hundreds of employees on the reconditioning lots that are processing actual cars every single day. You can go on Bloomberg and check the trading price of the ABS and talk to the sophisticated debt investors that keep coming back to Carvana for more and more. The noise is loud, but the business underneath it is so plainly obvious that it is almost embarrassing.

Portfolio Update

Since my last portfolio update, I (1) completely exited my position in Evolution AB at a 30% loss from cost basis, (2) bought more shares of Shift4 Payments and (3) started a new position in Meta. I lost confidence in my ability to underwrite EVO’s path from here, given deterioration in its core EU market as well as lack of clarity from management on how they intend to handle their persistent struggles and where they themselves see the business going. As a result, I decided to cut the whole investment and reallocate to FOUR and a start new position in META.

My current portfolio looks like this:

Disclaimer: Nothing in this article constitutes investment advice. These are my personal views and reflect my own portfolio decisions. Do your own due diligence. I am long the stocks mentioned in this article as of the date hereof.

Going into year-end, my overall view of my holdings is positive — the laggards are not materially impaired, and the outperformers are not close to their long-term fair value.

Looking forward to 2026 on select names:

FTAI Aviation: We go into 2026 with a humming core business, now with a massive call option.

1. FTAI has hit on all cylinders and surpassed management’s prior market share projections. It is now well on its way to achieving a 25%+ market share over the next few years.

2. FTAI’s capital-light transition has proved successful. The company has massively grown EBITDA this year and has already made the capital expenditures needed to service 25% of the market. It has done this while shrinking hard assets on its balance sheet. Further, FTAI completed its upsized strategic capital initiative. I would not be surprised if it expands the initiative further at some point given its success so far.

3. FTAI has stellar management. In the background, it is always thinking about ways to extract the most value from its CFM56 platform. Its recent announcement to convert past useful flying-life engines into modular power turbines for data centers further evidences this. Napkin math implies that if this call option pans out, the company could generate an incremental $800m to $1b of EBITDA.

International Workplace Group: 2026 is set to be an inflection point for the unit economics of the franchise business model.

1. IWG has continued to implement its franchise-model business plan. As I’ve discussed previously, we should see early franchise cohorts reach maturity by the end of 2026. FCF should materially rise as a result.

2. Although the company didn’t post surprising upside results this year, as many had anticipated it would do, I have become increasingly confident in management. The company hired a new CFO, has completed its prior share repurchase program and launched a new one, showing the company’s willingness to adjust capital allocation to optimize the business — previously, management had implied share repurchases would take a backseat to debt paydown. This looks to be a more shareholder-friendly approach that I am optimistic about as IWG navigates a pivotal year for its franchise business.

Shift4 Payments: 2026 will be an opportunity to demonstrate organic growth ability and commitment to optimal capital allocation.

1. Although FOUR’s stock performance was poor, it had a great year in terms of business fundamentals. The company closed its Global Blue acquisition, which opens up the European market and adds product offerings (primarily VAT refunds) that complement the business well. With Global Blue, FOUR has unlocked a longer runway of organic growth. M&A notwithstanding, shareholders should see the incremental benefits of this in years to come.

2. FOUR’s management team is best-in-class. Isaacman remains the largest shareholder and has entrusted key individuals to manage the business day-to-day. The new CEO has been with the company for many years and was instrumental to its performance. He aligns with Isaacman’s view of capital deployment, and it’s clear from his time in the role so far that he doesn’t plan to change that winning formula. I see the business continuing to have the same benefits as one would see in a founder-led company, but now with a cleaner single-share class structure.

3. FOUR is poised to hit the buyback button aggressively so long as its stock price lags. Some may argue that FOUR’s debt burden makes this imprudent. However, again, management has demonstrated great capital discipline over many years — I have no reason to believe they cannot or will not be managing debt and equity properly to maximize risk-adjusted results.

Disclosure: I am long these stocks. This is not investment advice.

My portfolio currently looks like this:

New Positions:

FTAI Aviation ($FTAI)

FTAI started as an aircraft and engine leasing company, but it’s shifting its focus to a fast-growing, high-return maintenance, repair and overhaul (MRO) business. FTAI works mostly on CFM56 and V2500 engines, which are widely used in Boeing and Airbus aircraft. These engines are still in high demand and will broadly stay in use for at least the next couple of decades. FTAI currently has around 5% market share in the fragmented MRO market for these engine platforms, providing plenty of room for growth.

The company’s edge comes from how it repairs engines. Instead of doing a full engine overhaul like most MROs, FTAI swaps or fixes engine modules using ones it keeps in inventory. This saves customers time and money, and helps FTAI earn higher margins than do traditional MROs.

FTAI’s leasing business supports its MRO work. It supplies used modules for repairs and creates demand for the swaps, helping the MRO side grow faster. Earlier this year, the stock fell sharply after a few short reports came out, but at the same time, FTAI announced a multi-billion dollar capital deal that will move much of its leasing business off its balance sheet and into a new partnership. This lets FTAI keep expanding the leasing side with less employed capital, all while focusing on its higher-return MRO business. The sell-off and new capital initiative created a good entry point.

FTAI currently trades at around 10x NTM EV/EBITDA. Organic growth to around 10-12% marketshare, fueled by the strategic capital initiative, should be more than enough to generate a high teens return from here assuming no multiple expansion. If you factor in a multiple rerate to 20+ EV/EBITDA (closer to other quality MRO businesses), per-unit EBITDA growth due to the implementation of PMA parts (which have higher margins) and general scale with the potential to blow past the 10% market share target, you have something that can generate well in excess of 20% IRR for the long-term.

Vistry Group PLC ($VTY.L)

Vistry is one of the biggest homebuilders in the UK and has received a lot of attention from hedge funds and fintwit. The stock dropped after the company issued a few profit warnings and slightly missed its 2024 goals. The issues are both one-off and not that big in the grand scheme on an absolute-pound basis.

What’s more important is that Vistry is changing how it runs its business. Instead of holding land and in-process homes on its balance sheet like a traditional builder, it’s moving to an asset-light model. This means working through partnerships wherein partners help cover build costs upfront — to the tune of at least 50%. This lowers Vistry’s capital needs with respect to both in-process projects and the amount of land it needs to keep on hand to support future projects. This new model not only results in higher ROIC, it should also make the company less exposed to housing market cycles. Furthermore, the macro backdrop is favorable — there is a chronic housing shortage in the UK that the current government is pushing hard to fix.

Whether the steady state 40% ROIC and £1 billion in capital release that management projects with the new partnership model actually pans out has become irrelevant with the recent sell offs. Following the sell-offs, Vistry traded at around tangible book value, in line with traditional homebuilders. Any modicum of success on the partnership side should result in significant returns from here. If the partnership model evolves even close to management’s estimates and Vistry becomes valued on an EV/EBIT basis, the upside case can be in the 40% IRR range.

Shift4 Payments ($FOUR)

I noted that FOUR was on my close watchlist in my last post. After continuing to look into it, I decided that it was worth the investment.

FOUR’s land grab M&A model combined with its vertically integrated product offering (from POS to backend processing) is unique in an otherwise crowded space. It has already established itself as the dominant player in two major verticals (hotels and sports venues) and continues to compete well in the much larger restaurant segment. With its upcoming acquisition of Global Blue, it is well positioned for further international expansion as its TAM continues to organically expand as more payments around the world move from cash to digital/card.

The current valuation is a bargain because the implied organic growth of the business more than justifies it — if FOUR builds on its historical streak of best-in-space capital allocation and is able to continue expanding inorganically growth through M&A, there is massive long-run upside.

Disclosure: I am long these stocks. This is not investment advice.

I don’t know if I’ll do this monthly, quarterly or continue this at all…we’ll see how useful this is — both to myself and to anyone else reading this. I’m experimenting with things to write about other than actual stock pitches, which obviously do not come along consistently.

A couple interesting ideas I looked at recently

Abacus Global Management ($ABL)

ABL’s main business is to acquire life insurance policies from older individuals for cash (before the policies expire) and then either hold them (i.e. in hopes of obtaining the payout) or sell them for a spread (e.g. to an investor looking for uncorrelated returns on a portfolio of life insurance policies or the original carrier looking to close out the policy early and release regulator-required reserves). It also operates a few related ancillary businesses including asset management, portfolio servicing, etc.

ABL is qualitatively a very unique business.

At the micro level, the business itself is seemingly a win-win for all stakeholders. In the ordinary course, 9 out of 10 life insurance policies expire without payment — by selling their policies to ABL, older individuals can obtain a guaranteed cash payout that they might not otherwise have realized. On the other end, alternative asset managers/investors can hold these policies for uncorrelated returns or insurance carriers can buy back and cancel these policies to clear up reserves for a net gain. ABL maintains a decades-long track record and a network of relationships with wealth advisors that look to have created a durable funnel for new policy acquisitions.

At the macro level, the life insurance market is massive and a very small portion of policy holders are aware that they can settle their soon-to-expire policies for cash. ABL, as the only life settlements player that is publicly traded, is positioned well to grow into this largely untapped market.

I’ve passed on this idea for now because I really can’t understand the quality of its earnings with the information it has provided in order to come up with a valuation that I am confident in.

• Management’s non-GAAP metrics include substantial unrealized gains on policies. On one hand, there is good reason to believe unrealized gains will at some point become realize given that FMV is based on management’s models refined over decades. However, on the other hand, if I were to apply any notable discount or adjustment to remove the effect of non-cash gains, the current valuation of the company doesn’t look too amazing in my view.

• Although the business is expected to grow significantly, I expect significant dilution over the next few years, and I am unsure how fast earnings can grow to outweigh this drag. Although management is paid on the low end, it is not unreasonable to think SBC will increase and weigh on the business over the next several years — indeed, ABL unilaterally accelerated the CEO’s massive equity compensation package recently. There are also a number of warrants outstanding — if you believe in the investment case for ABL, then the strike price of the warrants is below the intrinsic value of the business and regardless of the fact that management can deploy that capital at 15% ROIC, the issuance of those shares would still be net dilutive in most scenarios.

This is a weird one because I have little doubt that this is a profitable business that ABL is in a great position dominate. I think it’s an “up and to the right” story, but I just can’t put my finger on the slope of that line right now.

Shift4 Payments ($FOUR):

FOUR has a compelling business model driven by successful M&A and entry into various POS verticals. Currently, it has the technology to successfully use M&A as a customer acquisition method. Direct competitors in the POS space (e.g. Toast, Lightspeed) are largely focused on their verticals and will first need to build out their product offering to be in a position to replicate FOUR’s M&A strategy. Upstream players (e.g. the various payment processors), as some have pointed out, either haven’t had the greatest track record in M&A or the economics of this space make such a strategy nonsensical financially due to their massive size. So, at a high level, FOUR’s growth model looks fairly defensible for at least the medium term.

The valuation isn’t unreasonable. What entices me about this name is that it could really well be one of those “wonderful businesses at a great price” bets. But I don’t want to let the prospect of buying what looks to be a great business at a great price to cloud my thinking here. How will the business fare once the founder/CEO leaves? Is there real, underlying organic growth? Does that even matter, really? Should I go in relying on M&A to continue, organic growth be damned?

I probably need to do more work on this one to build conviction. But, I generally like what I see here and will keep it on my close watch list.

Snapshot

As of the end of March 2025, my portfolio breakdown was as follows:

I have a big SPY position that came to be because of restrictions on my retirement accounts and on general stock trading at my prior job. I am now unrestricted, so I hope to shift more of my SPY and cash into individual investments as I find them. I’m in no rush.

I have already written in more detail about my investments in Evolution AB and International Workplace plc. The only other position I haven’t discussed in detail is Carvana. I have owned this position for many years (my first purchase dating back to 2019) and held through the rise, fall and recent rise again, adding at times along the way. It had become an extremely large portion of my portfolio, so I trimmed it during the back half of last year and now hold a position that I feel comfortable holding onto. I still think there is ample runway for the company. The bull and bear cases for CVNA have been beaten to death, and there are many who have written on it elsewhere if you are interested. I have a small Ethereum position that I’ve held for many years, which I’ll let run. At this point, I think about it as schmuck insurance in crypto.

Disclaimer: None of this investing advice.

The purpose of this post isn’t to explain the business qualities (e.g. moat, etc.). My thoughts aren’t too different from what has been explained in detail by other bulls elsewhere throughout the years and reflected in the business’s underlying performance.

Despite the clear signs of business quality, the stock trades at multi-year lows in both price and valuation. I think it’s clear that, outside of a few near-term hiccups, the key drag on the stock is regulatory risk. I’ll make that the focus of this post, primarily so I can look back at how I thought about it in a few years’ time.

***

The misunderstanding of both the game theory of regulatory regimes and the nature/level of risk associated with regulation in iGaming has made this company fall out of favor, along with the general uncertainty regulation as a concept draws from investors. There are two key risks, neither of which are great enough to justify the discount in EVO’s valuation in my view:

1. The bulk of EVO’s revenues (60%) coming from unregulated markets.

2. Operators providing EVO games to jurisdictions in which they are not licensed to operate.

On number 1, there are a few potential outcomes in each unregulated market that I see:

• Each market slowly adopts regulations for online gaming. In this situation, EVO would be the incumbent supplier of games and quickly adapt to new regulatory frameworks and make required investments.

• The status quo remains.

• Outright ban/implied ban on all online gaming activity, as we have seen in China.

The most likely outcome is either the first or second scenario. Online gaming is a massive industry in Asia and has operated for many years under the clear view of governments. Either governments are too busy with other things to care, or they do not care. If they cared enough to say that gambling should be banned, then they would not have allowed it to flourish at this rate for so long. So, the likely case is that unregulated markets continue to remain unregulated and, in a situation in which governments do decide to act, regulation is more likely than outright bans.

The Philippines, which just recently adopted regulations, provides a representative case study of the first scenario playing out: EVO has continued to operate in the market, has built out a new studio and has hired local employees — all of which is being done without much consternation and while complying properly with new laws. As unregulated revenues become regulated, it is fair to say that EVO’s sky-high margins would decrease. However, if you take bears at their word that unregulated revenues is a problem which is reflected in the low multiple at which EVO currently trades, then you can already start to see a situation where even though the margins from unregulated markets may decrease as they transition, the higher multiple on those now “better” revenue streams should have an offsetting effect.

In the second scenario, unregulated revenues would continue to drag the valuation of the company. However, these revenues would nonetheless come in and support continued buybacks and returns of capital to shareholders. Sure, we may not benefit from significant multiple appreciation in this scenario, but the value of the underlying cash flows would still be there.

The third scenario, as mentioned above, is unlikely. First, a ban of a highly active online market is not practicable outside of places like China where the government can and does implement totalitarian control over internet activity. Indeed, a ban in any another Asian country would result in a further entrenched black market and prevent that government from generating jobs and revenues — it is not a logical course of action, so I do not view it as the most likely route to be taken. I would also note that each jurisdiction acts independently — even if one jurisdiction takes the ban route, there is no reason to believe others would follow given that we’ve already seen the Philippines taking the other approach successfully so far. Further, given EVO’s global footprint, the loss of one market should not materially affect EVO’s long term revenue potential overall.

I think the likely path for EVO from here is that it will shake off its near term hiccups (cyberattacks, UK regulatory review, labor issues) and continue to grow revenues through a combination of entry into and maturation of new markets (e.g. United States) and benefits from the global tailwinds to online gaming activity.

It is reasonable that its status as a “sin” stock may mean that it won’t ever command an eye-watering multiple despite the amazing financial metrics; however, the combination of FCF generation, buybacks and a rerate closer to a more reasonable multiple as EVO’s revenue mix shifts more toward regulated should spell great returns from here.

Disclosure: I am long the stock. This is not investing advice.

I’m mainly writing this to myself to look back on and organize my thoughts…and because I like writing. If you are reading this, maybe you’ll find it intriguing enough to go kick the tires yourself.

What is it?

IWG leases and manages a large network of flexible workspaces (think: WeWork). Historically, it took the standard approach of methodically buying “centers” and then earning revenue by filling these spaces with users.

For the past few years, IWG has been executing a “capital light” managed location/franchising strategy. Instead of directly leasing buildings, it has been partnering with building owners to convert their buildings into co-working spaces. The owner bears the capital costs, while IWG manages the space and drives usage in exchange for a management fee — IWG has stated that it takes an average of 10 months to open a managed location after it is formally signed up, and then another 18 months for it to reach mature earnings.

Why am I investing?

The basic investment case is that not only has the legacy owned and operated segment of IWG’s business recovered to its pre-Covid earnings, IWG now has a fast-growing franchise segment of which the initial cohorts are starting to come online and reach maturity.

If we do very conservative math on the legacy business, we reach a valuation that implies fairly decent upside from where the stock currently trades. If you add on the managed/franchised business, you end up with very high upside in 5x+ range. I’ll say a variation of the cliche (forgive me): this appears to be a “heads I win a lot, tails I win a decent amount.”

Why is this so undervalued?

Among other things, I listened through several years’ worth of investor calls of IWG, in order, starting with 2020, along with reading the corresponding reports/releases and investor presentations. Here are some observations:

• The CEO and management team haven’t done a great job explaining the business and setting expectations with investors. Probably not intentional, but they have consistently overpromised and underdelivered, so I can understand why people will continue to discount management until the numbers hit the bottom line. One example is that the software business was repeatedly touted as something that will grow 10%+ per year, slated for a future IPO, etc. Since these projections were given, the software business has gone flat. The CEO and management said it was due to development hiccups and basically stopped talking about it. When you listen to some of the earlier calls without any context, it felt like the CEO was telling you that the company happens to own a Stamps.com flyer. Another example is that some of the early projections for the managed/franchise pipeline uptake and timing were pie-in-the-sky before coming down to the current more reasonable (TBD) projections. Yet another was IWG floating the idea of a US-listing over the years without any actual plan or timing for said listing.

• The financials are hard to understand. I’m saying this as someone who has lived and died by Carvana and its supposedly opaque financials. IWG’s financials are hard to understand because of a combination of 1) IFRS’s annoying (and in this company’s case, largely inapplicable) treatment of leases and 2) management trying to develop a useful way to present IWG’s new business segment on the fly. The way IWG has presented its financials has not been consistent over the last few years, so it is difficult to easily see trends from one report or presentation to the next, or to understand which key metrics need to be focused on. I spent a lot of time trying to reconcile various metrics in order to model something out that made sense and was accurate. I assume this makes the stock less appealing or accessible to more investors.

Catalysts

So the stock has been dead money for many years. The business has been executing directionally (as mentioned previously, the company has overpromised and underdelivered, but if you take their under delivery and discount it even further in your own numbers, you still get something very good in terms of implied IRR). What’s different now? Well, they just started reporting in USD and have committed to going GAAP this year. That alone, I think, will substantially improve the visibility of the business for the broader market. On top of that, the managed cohorts are starting to come online and hit the FCF numbers — by the end of 2026, we’ll finally have some clean segment-level P&L that the market can start modeling out period over period. IWG also recently started paying a dividend and announced a buyback. Although all of these things together should help push stock performance over the next year or two, I think the ultimate rerate, if it is to come, will only happen if IWG lists in the US. The optimist in me hopes that management understands this obvious reality and is making plans for a US listing in due course, but in the meantime I feel comfortable holding this stock for a while while the business continues to perform.

Some Links for Further Viewing:

• 1MainCapital Letter Q4 2024 Letter

• Youtube (YAVP): Link 1, Link 2, Link 3, Link 4

• Youtube (FatAlpha)

• YAVP Writeup

• @secretlyaninja X Thread

Disclaimer: I’m long this stock. This is not investing advice.