• Revenue reached $895 million, up 53% YoY and beating guidance of $865 million.

• DAUq grew 12% YoY to 130.3 million (though this was flat QoQ).

• WAUq grew 16% YoY to 515.2 million (though again flat QoQ).

• ARPU grew 36% YoY and 11% QoQ to $6.87, with both international and U.S. ARPU growing nicely.

Despite these results, Reddit stock crashed by over 30% in just one day.

The reasons behind this crash aren’t in the headline numbers. Rather, most of the concerns stem from management commentary and guidance.

Both require some nuance.

User Growth: The Elephant in the Room

One of Reddit’s primary concerns is user growth, or rather the lack thereof.

Growth had already been slowing in the quarters prior, so another slowdown shouldn’t have come as a surprise.

Yet Reddit managed to reignite logged-in user growth, albeit by a small degree.

Logged-in DAUq reached 53.5 million this quarter, up 7% YoY and 1.7% QoQ.

While DAUq and WAUq continued to grow at a decent rate (12% and 16%), investors are increasingly focused on the slower growth of this subset of users.

These represent Reddit’s most engaged and monetizable users, with about 80% of revenue coming from this subset.

Logged-in users are more valuable for several reasons:

• Being logged in leads to more relevant content and communities, enriching the user experience.

• Being logged in allows for better ad targeting.

• Being logged in generally leads to higher engagement, motivating users to come back every day.

What has been worrying investors, however, is the stagnation in U.S. DAUq, including both logged-in and logged-out users.

This figure reached 53.5 million in Q1 2026, 53.2 million in Q2 2026, and declined to 50.7 million in Q3.

In other words, while logged-in DAUq is still growing, overall U.S. DAUq is declining. This suggests that growth is increasingly shifting towards international markets, where users are currently less valuable from a monetization perspective.

A big driver of this is Google.

As I flagged in my deep dive, Reddit’s dependence on Google was one of the core reasons its shares were, and still are, in a drawdown. Google has historically driven significant traffic to Reddit.

Since the beginning of this year, as Google has heavily pushed AI Overviews and AI Mode into search, traffic growth has stalled.

This is a strategic change from Google that is very unlikely to reverse.

The impact is mostly at the top of the funnel, but management acknowledged that “some of the impact flows through to logged in and app traffic, since search referral is a common source of re-engagement for users.”

According to management, search traffic dropped by more than 1 million referrals per week in October.

To make matters worse, management notes that if this trend persists, “total DAUq could fall in Q4 by a low double-digit percentage from our current base of 130 million DAUq.”

This will put further pressure on user growth, including the most valuable subset of users: logged-in users.

Search referrals were already volatile in Q2, but the problem has now accelerated.

As a result, management seems to have quietly abandoned its ambition of reaching 1 billion DAUq globally and 100 million in the U.S. This has been Reddit’s target for a while, but all of a sudden, we’ve heard nothing about these plans.

Not a great sign from management.

And so, Reddit’s strategy has shifted away from growing top-of-funnel users. Rather, the focus is now on converting logged-out users into logged-in app users and monetizing those users more effectively.

Converting Users Into Logged-In DAUq

Despite what the stock price suggests, there is still a serious opportunity here.

Read more

Today’s portfolio update marks a fresh start.

As you (hopefully) are aware, I sold my entire portfolio last month.

This was not a decision taken lightly, but one that was necessary. The reason behind it was the purchase of my first ever home.

A house is more than just an investment; it’s also a place to live. For that simple reason alone, I put higher value in a home than in stocks.

But that doesn’t mean I’ll no longer invest in stocks.

In fact, I’ve already re-established the portfolio, starting with Amazon.

I’ve been vocal about Amazon over the past month or two, with an update in September and an earnings update just yesterday.

Owning Amazon today means owning a company that could have two businesses generating over $1 trillion in revenue in the next 5-10 years.

The incremental returns Amazon is generating today by investing heavily into AWS and retail are unmatched, especially at the company’s scale.

For these reasons and more, the rebuilding of the portfolio has begun with Amazon.

Two shares, to be exact. I purchased them this week, before earnings.

As a result, the portfolio is slightly in the red. But it matters not. Amazon is a $700 share in a few years, in my opinion.

That’s not an exact price target, because I don’t invest with exact price targets in mind. Rather, it’s my way of saying there’s a lot of margin of safety embedded in the current valuation, given Amazon’s high-return investments.

In my last portfolio update, I mentioned there wouldn’t be a portfolio at all for 2-4 months. It happened a little sooner than expected, but I don’t expect to accumulate shares quickly just yet.

In other words, the portfolio remains small for now. A house is quite expensive, even (or especially) after the purchase.

But what matters for you as a reader of my work is not the size of my portfolio, but its percentage returns and composition. You’ll still get the same value out of this newsletter.

Alongside Amazon, I have a few other names I wish to accumulate as soon as possible.

Current Watchlist

Basic-Fit ($BFIT)

I’ve written extensively about Basic-Fit, which was a core position in my portfolio before I sold it.

Basic-Fit looks like a similar investment to Amazon when looking solely at cash flow dynamics.

Across almost its entire history, the European fitness chain has been highly focused on reinvesting all its cash flows into new fitness clubs.

As a result, cash flows have never been high.

But the returns on these investments (the addition of fitness clubs) are high, despite not being clearly visible in the income statement. This is the result of clubs needing time to mature, as well as other smaller factors.

Today, however, we’re at an inflection point.

Free cash flow is turning positive, partly as a result of a shift in strategy. Basic-Fit is investing less into building out clubs and more into franchising.

This allows the existing club base to mature while adding a second, more capital-light, revenue stream.

Based on my own estimates, about 8% annual FCF growth is priced into the stock, which I believe is a relatively low hurdle.

ROIIC is consistently above 20%, and today, Basic-Fit trades at a P/FCF of about 13x.

I will be looking to add Basic-Fit back to the portfolio, preferably below €30 per share.

Nintendo ($NTDOY, $7974)

Like Basic-Fit, Nintendo was also in my portfolio, though as a smaller position because it was a much newer position that I was still building out.

And like Basic-Fit, Nintendo is at an inflection point.

With the Switch and now the Switch 2, Nintendo has become a business revolving around a continuous family of devices. The Switch family allows players to use the same operating system and play the same games across devices.

Nintendo’s userbase now carries over with every new device, making the business more predictable and resilient.

On top of that, Nintendo owns world-class IP, with the likes of Mario, Zelda, Pokémon, Donkey Kong, Kirby, Animal Crossing, and more.

Software sales are recurring and high-margin, while hardware sales lock users into the Nintendo ecosystem.

Nintendo still trades at what I believe to be an attractive valuation. Like Basic-Fit, I will be looking to add Nintendo back to the portfolio.

Alphabet ($GOOGL, $GOOG)

Read more

How obvious the Amazon thesis looks right now.

While Amazon is investing aggressively in a once-in-a-lifetime opportunity, investors are abandoning ship instead of recognizing what’s happening, with shares down post-earnings.

Q3 2026 earnings were brilliant. But even more compelling were management’s remarks about the future, particularly the opportunity ahead for AWS.

But first things first:

• Amazon reported $201.9 billion in revenue, up 12% YoY, within its guidance range of $197 to $202 billion. Excluding the Prime Day timing impact (Prime Day shifted from Q3 in 2025 to Q2 this year in most countries), revenue would have grown 16%.

• Amazon reported $28.1 billion in operating income, up 61% YoY and well above guidance of $22.5 to $26.5 billion.

• TTM operating cash flow reached $168.3 billion, up 29% YoY.

• AWS continued accelerating, reaching $46 billion in revenue, up 39% YoY. AWS is now a $184 billion ARR business, up from nearly $170 billion just a quarter ago. At the same time, the AWS backlog grew from $496 billion to $578 billion sequentially.

These were the headline numbers.

But behind them is a fantastic story unfolding.

AWS is on track to become a trillion-dollar revenue business.

AWS: A Trillion-Dollar Opportunity

To reach that scale, management is investing enormous amounts of money. Amazon reiterated its 2026 capex guidance of $220 billion.

Over the trailing twelve months, the company has already spent $215.2 billion in capex. Meanwhile, Amazon “expects to spend meaningfully more in 2027.”

As you may know from my previous Amazon update, I stand fully behind this strategy. Not just because of a hunch that AI is a generational tailwind, but because the numbers support it.

As of Q2 2026, Amazon’s 2-year rolling TTM ROIIC was 23.8%, while its 3-year rolling TTM ROIIC was 35.3%.

In other words, Amazon’s recent investments are generating highly attractive incremental returns.

Moreover, today’s capex spending, and even last year’s, isn’t fully reflected in the profit numbers yet, since it takes about two years for AWS investments to start monetizing.

From 2027 onward, we’ll start to get more visibility into the returns on today’s investments.

And AWS is accelerating.

Management shared some incredible details about what they expect from AWS going forward: at least 40% annual revenue growth in both 2027 and 2028.

This is new information of the kind Amazon is usually reluctant to share. It seems management is trying to reassure investors by providing more detailed guidance than usual.

Let’s put 40% annual growth at AWS’ scale into perspective.

AWS currently operates at an annual revenue run rate of $184 billion. For simplicity, let’s assume it reaches a $200 billion run rate by the end of the year.

At 40% annual growth, that implies $280 billion in revenue in 2027 and $392 billion in 2028.

Assuming a 40% operating margin, AWS would generate $157 billion in operating income in 2028 alone.

Moreover, management stated that it believes AWS can become a trillion-dollar revenue business within five to ten years.

If we assume AWS reaches $1 trillion in revenue in eight years, revenue would need to grow by approximately 17% annually after 2028.

All of a sudden, a trillion dollars in revenue doesn’t seem implausible.

This extreme growth is enabled by Amazon’s capex investments. The company is spending these enormous sums because it’s seeing an incredible opportunity ahead.

“As I have said a few times, we will always, despite how much capacity we are providing, the big chunks of capacity that people take, the hundreds of megawatts or sometimes up to a gigawatt of capacity, those big chunks are reserved well in advance because you cannot leave a gigawatt of capacity hanging around. We have very good visibility into 2027 and 2028 and some of 2029 too, with respect to what customers want in those really big chunks of capacity.”

—Andy Jassy, CEO Amazon

Of course, strong demand alone doesn’t guarantee attractive returns. That’s why I pay close attention to ROIIC.

In Q3, invested capital increased from $854.4 billion to $985.4 billion in just one quarter.

TTM operating income increased 37% YoY to $104.4 billion.

These figures imply a 2-year rolling TTM ROIIC of 22.9% and a 3-year rolling TTM ROIIC of 31.9%.

Both figures remain extremely healthy, reaffirming that Amazon’s strategy is the right one to pursue.

Moreover, even if you believe the capex spending remains a source of uncertainty, there are several indications that the risk is lower than some investors think:

• The AWS backlog is enormous. Amazon’s AWS backlog stands at $578 billion, providing substantial evidence of future demand.

• Capex payback periods are shortening. Management explained that the time it takes for investments to pay back is decreasing. In other words, those two years it takes AWS investments to start monetizing are becoming shorter, providing greater visibility into returns on capital.

• AWS assets are becoming more valuable over time. Alongside shorter payback periods, Amazon is extending the useful lives of its AWS assets. We don’t know the specifics, but this is another positive indicator for the economics of its investments.

AWS is a trillion-dollar opportunity, and the market is punishing Amazon for pursuing it.

And here’s the thing: Amazon is much more than just a cloud business.

Don’t Forget About Retail

Read more

Not only does it help you identify high-quality businesses, but it also helps you better understand valuation.

ROIC measures how effectively a business uses invested capital to generate profits. The higher the ROIC, the more effectively a business uses its capital.

Say a business has $100 million in invested capital and uses that capital to generate $20 million in annual profit. That means the business has a 20% ROIC.

Profit, within the context of ROIC, is usually defined as net operating profit after tax (NOPAT), though operating income often also gets the point across. Invested capital is typically calculated by summing operating assets and net working capital.

Understanding Business Quality

ROIC is useful first and foremost because it helps you determine whether a business creates value.

In the context of investing, what is value?

Is it simply profit? Is it growth? Is it an annual dividend?

Financially speaking, a business creates value when its return on invested capital exceeds its cost of capital.

This means that if our business with a 20% ROIC has a cost of capital of 10%, it creates value because its ROIC exceeds the cost of capital.

But if our business generates an ROIC of just 9%, it destroys value, despite still generating a profit.

This makes sense intuitively: why invest in a business earning a 9% return on its capital if the risk-adjusted return required by investors is 10%?

But high ROIC alone isn’t enough. As an investor, you should be looking for growth alongside high returns on capital.

All else equal, a company with a low ROIC needs to invest significantly more than a company with a high ROIC to achieve the same growth rate.

Taking this one step further, returns on incremental invested capital (ROIIC) need to be high for an investment to be attractive.

This is perhaps more important than ROIC. If a company is investing its cash flows in low-return projects despite having a high ROIC, you should probably not invest in that business.

ROIIC is calculated by dividing incremental profit by incremental invested capital from the prior year.

Interpret this as: how much additional profit did last year’s investments make?

Of course, this is a rough and imprecise estimate. Investments take time to become productive, and the returns they generate can vary significantly from year to year. Still, ROIIC can provide valuable insight into a company’s growth prospects.

If you’re wondering how I think about ROIIC in practice, and how I use it to make investment decisions, check out my recent Amazon article.

A company can report a high ROIC because of investments made years ago, while its new investments earn substantially lower returns. This is what you could call a “quality trap.”

Many mature companies with strong moats still generate high returns on capital.

Yet because of their maturity, the investments they make today are often small and/or yield low returns.

On the other hand, a company can report a low or modest ROIC today while investing heavily in projects that will generate attractive returns in the future.

An inflection in ROIC can be a great way to spot undervalued stocks, particularly when the market hasn’t yet recognized the improvement in those underlying economics.

Think of Amazon: ROIC today looks very low, but the ROIIC on the investments the business is making today are about 20-40%.

How ROIC Drives Free Cash Flow

Read more

Revenue growth had stalled, competition was intensifying, and investors began envisioning a doom scenario, sending shares down almost 75%.

Today, this enormous drawdown looks nothing more than a small, temporary dip.

But Netflix didn’t recover out of nowhere.

Management made several important changes. The company launched a global password-sharing crackdown, introduced advertising as a second revenue stream, and focused on operational efficiencies, pushing its operating margin above 30%.

But once again, Netflix is going through a rough period, with its stock now down about 50% from its all-time high.

This drawdown reflects stretched expectations, uncertainty surrounding the Warner Bros. acquisition, engagement concerns, and slowing revenue growth.

Yesterday, Netflix released its Q3 2026 earnings report, and shares fell another 7% in after-hours trading, approaching $60 per share.

Let’s dive in.

Netflix’s Q3 2026 Earnings

Netflix reported $12.87 billion in revenue, up 11.8% YoY.

Operating income reached $4.46 billion, up 37% YoY, while its operating margin hit a record 34.6%.

On the surface, these are excellent results. Revenue is growing at a double-digit rate, and profitability continues to improve.

The concern, however, is that growth is slowing.

Looking at Netflix’s revenue growth since 2021, I see four distinct phases (each marked by an arrow).

• Phase 1: Growth deceleration. Subscriber growth stagnated, revenue growth weakened, and investors began questioning Netflix’s long-term prospects. The stock fell almost 75%.

• Phase 2: Growth reacceleration. Netflix addressed its subscriber growth problem, largely through its password-sharing crackdown. Revenue growth picked up again.

• Phase 3: Monetization. Netflix benefited from the resulting subscriber growth while expanding advertising, gaming, and live programming. Pricing increases and operating leverage provided additional tailwinds.

• Phase 4: Growth deceleration. Revenue growth is slowing again, raising questions about what will drive the next leg of growth.

We are now in phase four.

These periods are dangerous for investors because the direction of growth is uncertain. Buying today essentially means betting on reacceleration, or at least a halt in the slowdown.

Read more

We’ve all heard this quote perhaps a hundred times. Warren Buffett’s wisdom gets shared every day, and for good reason.

If there’s one investor you can learn from, it’s Buffett.

And he’s right. It is better to buy a wonderful business at a fair price than the other way around.

A “fair” business implies an average business with little to no moat, one that struggles to create value for its shareholders. At least, that’s my interpretation.

Buying such a business at a wonderful price, say at a P/E below 10x, can easily turn into a value trap.

On the other hand, buying a wonderful company, one with a sustainable moat that creates value year after year, is a much better proposition, even at a higher multiple.

But that doesn’t mean we should ignore valuation entirely.

Which is what many investors do. They justify paying excessive prices for great companies by using quotes like Buffett’s as an excuse.

And often, here’s what happens afterwards:

These are just 3 examples of typical high-quality companies that have significantly underperformed the market over the past five years, partly because investors paid too much for their quality.

Read more

By September 2025, shares had reached $271, nearly an 8x increase in just 18 months.

Although Reddit has existed for over two decades, it has only recently begun unlocking the commercial potential of its user base.

After returning as CEO in 2015, co-founder Steve Huffman began professionalizing the platform. In 2018, Reddit started investing more seriously in advertising technology, targeting, and measurement tools.

The results have been impressive. Revenue grew 69% in 2025, while reaching $805 million in Q2 2026, compared with just $108 million in Q2 2021.

Advertising is the primary growth engine, but Reddit also has an opportunity to monetize its enormous archive of human conversations through content licensing to AI companies.

Yet despite this potential, shares have fallen roughly 40% from their September 2025 peak and are down approximately 34% year to date.

In this deep dive, we’ll examine the drawdown, Reddit’s business model, and whether the current share price offers an attractive investment opportunity.

Why Are Shares Down?

The 2026 drawdown has four primary causes:

• Concerns about ad effectiveness: In January, investment bank RBC reported mixed advertiser feedback, with some smaller brands seeing limited returns from Reddit ads. This raised questions about Reddit’s ability to attract larger advertising budgets and improve monetization.

• High growth expectations: At a price above $200, expectations were extremely high. Despite strong revenue growth in Q1 and Q2, investors became concerned about slowing user and revenue growth.

• Dependence on Google: AI-powered search could reduce traffic referrals to Reddit, while reports that YouTube is increasingly being prioritized as a source of data for AI, further began concerning investors.

• Slowing U.S. user and revenue growth: U.S. DAUq declined slightly QoQ in Q2, from 53.5 million to 53.2 million. For Q3, management guided for revenue of approximately $865 million, implying 48% YoY growth. Although strong, this was a deceleration from the 60%+ growth investors had become accustomed to, and below analyst expectations of roughly $886 million.

Reddit’s Business Model

Reddit’s platform consists of a network of over 100,000 communities, with 130 million Daily Active Uniques (DAUq) and 515 million Weekly Active Uniques (WAUq).

As the company calls it, a digital city.

Users visit particular corners of Reddit based on their interests. Some discuss hobbies, others seek advice, and many look for recommendations before making a purchase.

This creates an interesting proposition for advertisers. Reddit connects businesses with audiences based on their interests, while discussions often reveal what users are considering buying.

Management estimates that “around 40% of conversations on Reddit are commercial in nature.” Reddit also reports that “84% of shoppers also say they feel more confident in their decisions after researching on Reddit.”

Reddit sells advertising inventory that gives businesses access to audiences with varied interests and potential purchase intent.

The downside is that Reddit users tend to be skeptical toward ads and may not purchase immediately after seeing one. This can make advertising returns harder to measure.

Still, advertising remains Reddit’s primary revenue source. In Q2 2026, it generated $762 million, representing 95% of total revenue and growing 64% YoY.

Ad revenue is driven by two variables: the number of impressions served and the price advertisers pay for each impression. To keep growing, Reddit needs to increase both.

Growing Impressions

Read more

Collectively, these investors, much like an ant colony, interact to determine stock prices.

The resulting price reflects the collective judgment of the group.

And the wisdom of the crowd is real: investors as a group (the market) tend to outperform individual investors.

Put differently, if you’re investing in individual stocks, you need to beat the wisdom of the crowd to outperform.

That’s why outperforming the stock market is so hard.

But it is possible.

The key lies in valuation.

Not in building the most sophisticated DCF model, or in applying the most appropriate earnings multiple.

The key lies in understanding market expectations.

“The issue is not which horse in the race is the most likely winner, but which horse or horses are offering odds that exceed their actual chances of victory… Under this mindset, everything but the odds fades from view. There is no such thing as “liking” a horse to win a race, only an attractive discrepancy between his chances and his price”

—Steven Crist

First Principles of Valuation

Before we can understand market expectations, we need to understand the fundamentals of valuation.

A stock, like a bond, is a claim on future cash flows.

Therefore, there are three fundamental questions we must ask:

1. How much cash will the business generate?

2. When will it generate that cash?

3. How risky are those cash flows?

A DCF model attempts to answer all three questions. It is nothing more than an attempt to estimate future cash flows and discount them into today’s dollars.

If valuation hinges on just these three questions, it shouldn’t be complicated.

And it isn’t. At least fundamentally speaking.

The difficulty lies in uncertainty.

Unlike bonds, nobody knows exactly how much cash a business will generate, and when it will generate that cash.

This is why a DCF model is often considered too complicated, or too uncertain, despite its simplicity.

You need to estimate inputs like revenue growth, operating margins, and the length of the forecast period. Even small deviations in these assumptions can produce drastically different valuations.

As a result, many investors ignore the DCF model entirely and turn to valuation multiples instead.

But multiples are simply a shorthand for the same valuation process.

They save you time, yes, but they also conceal the assumptions that ultimately determine value.

“That many investors are comfortable with multiples but uncomfortable with discounted cash flow reflects cognitive dissonance. The simplicity of multiples is a sign of inaccuracy, not accuracy. As Keynes said, “It is better to be vaguely right than precisely wrong.””

—Michael Mauboussin

Subscribe now

Expectations Investing

If a DCF model requires so many uncertain inputs, and multiples are merely a shorthand, how then should we approach valuation?

The answer lies in understanding expectations.

Every stock price already reflects a set of expectations about a company’s future.

And those expectations ultimately revolve around the same three fundamental questions:

1. How much cash will the business generate?

2. When will it generate that cash?

3. How risky are those cash flows?

The difference is that, instead of trying to forecast everything ourselves, we can start with the market price and work backwards.

This brings us to an important distinction between fundamentals and expectations.

A company’s fundamentals tell us how the business is doing, while its stock price tells us what investors expect the business to do.

Those are not the same thing.

A great business can be a terrible investment if expectations are too high, while a mediocre business can be a great investment if expectations are too low.

Consider two companies.

Company A is expected to grow its free cash flow by 20% annually over the next decade. Company B is expected to grow by just 5%.

Here’s what actually happens: Company A grows FCF by 15% annually, while Company B grows by 7%.

Clearly, Company A is the superior business when looking at growth.

But Company A disappointed expectations, while Company B exceeded them. All else equal, Company B would likely deliver the better investment return, despite being the slower-growing business.

Returns depend not only on what happens, but on what was expected to happen.

Reverse-Engineering the Stock Price

This is where the DCF becomes particularly useful.

Instead of using a DCF to estimate what a stock should be worth, we can reverse the process and ask what must happen to justify its current price.

This is what I try to do in my own investment research.

For instance, in my previous post, I estimated that Basic-Fit must grow FCF by about 8% annually over the next decade to justify its current share price.

That is the expectation embedded in the stock price, representing the collective judgment of the market, or the wisdom of the crowd.

It gives us a hurdle to evaluate.

We can then investigate whether that hurdle is too high, too low, or roughly fair.

Based on our own research, we need to develop a compelling thesis explaining why the market’s implied expectations are too pessimistic (or too optimistic if you want to go short).

In Basic-Fit’s case, I believe 8% annual FCF growth is low for several reasons:

• Revenue is growing much faster than 8%.

• Capex is declining, with the shift toward franchising potentially reducing capital intensity further.

• Basic-Fit has substantial “hidden” earnings power that is gradually being revealed as its fitness clubs mature.

• Operational efficiencies and economies of scale should continue to support margin expansion.

The full process is straightforward: first, estimate the expectations implied by the stock price. Then, identify where those expectations might be wrong. Finally, determine whether the difference is large enough to create an attractive investment opportunity.

The goal isn’t to predict the future with extraordinary precision, which is exactly what a standard DCF model tries to do.

Instead, the goal is to understand what the market is already pricing in, and whether reality is likely to turn out differently.

Valuation is less about finding the right number and more about understanding what has to be true for the current price to make sense.

Until next time,
Lucas
Summit Stocks

Disclaimer: the information provided is for informational purposes only and should not be considered as financial advice. I am not a financial advisor, and nothing on this platform should be construed as personalized financial advice. All investment decisions should be made based on your own research.

Negative FCF stems from low profitability, high (re)investments, or both.

If you’re not an expert on reading financial statements, you might not know that the cash flow statement begins with net income, which is then adjusted for non-cash items to derive operating cash flow.

This means that if a business starts with low net income, it’s difficult to end up with high FCF.

From there, the cash flow statement continues. With operating cash flow in hand, management has a few decisions to make: will they reinvest that cash flow (which shows up as capital expenditures or acquisitions), return it to shareholders via dividends or buybacks, or use it to pay down debt?

If reinvestments are high, there isn’t much left for investors: free cash flow will be low.

We saw this play out perfectly at Amazon last year. Net income was fairly high at $77.8 billion. After big non-cash adjustments like D&A, operating cash flow ended up at $139.5 billion.

But since Amazon decided to reinvest nearly all of that cash ($131.8 billion), little was left for debt repayments, share buybacks, or dividends—which is why you barely see those on the cash flow statement.

The point is: when profitability is low and reinvestments are high, free cash flow will inevitably look poor.

For some businesses, however, there comes a point when profitability starts to increase just as investments decrease. When these happen simultaneously, you get a sort of twin-engine effect.

This is what’s happening at Basic-Fit.

Leave a comment

Basic-Fit is the dominant market leader in the European fitness industry. The company operates value-for-money fitness chains mostly across the Benelux, France, Spain, and Germany.

In the Netherlands, Basic-Fit is twice as large as its closest competitor. In Belgium, it is 3x larger; in France, 4x. In Spain and Germany, it is the number two and number one player in the market, respectively.

This scale is one of Basic-Fit’s core advantages, allowing the business to offer some of the lowest prices in the industry.

Today, the company has 6.1 million members across 2,192 clubs.

In addition, all of Basic-Fit’s KPIs look beautiful, growing gradually and consistently (if we ignore the COVID-19 disruption, of course).

Yet, while these metrics have compounded at 17-21% since 2016—leading to a total increase of 411% to 556%—shares have increased by “only” ~101% during that same period, with lots of volatility.

In fact, over the past five years, Basic-Fit shares are still down about 31%, despite a strong 65% run-up since early 2025.

Part of the reason shares continue to trade sideways rather than tracking Basic-Fit’s KPIs upward is free cash flow, or rather, the lack thereof.

Basic-Fit uses underlying EBITDA less rent as a proxy for profitability. You can roughly calculate free cash flow by subtracting capex, interest, and taxes from this metric.

Below, I’ll show you how calculating free cash flow the “traditional” way compares to this workaround.

Looking at the traditional method on the left: we start with operating cash flow, subtract capex, and (because of IFRS statements) subtract lease repayments and interest expenses. This gets us to a 2025 free cash flow of -€117 million.

Using underlying EBITDA less rent, we end up with -€119 million.

In my opinion, a €2 million difference is negligible.

The point is, free cash flow was negative last year. It was negative in 2024. And it was negative in 2023.

Basic-Fit hasn’t had a single year of truly strong FCF. This worries investors, hence the relatively poor share price performance.

But if you paid attention to the calculation above, you’ll notice that underlying EBITDA less rent is already high at €348 million in 2025 (a 24.5% margin). This means that if capex declines for whatever reason, Basic-Fit will end up with positive FCF, and potentially lots of it.

Of course, there are still interest and tax expenses, but those (especially the latter) are inevitable.

This is the profitability-plus-investment dynamic I mentioned earlier. And here’s where it gets even more interesting.

Subscribe now

In the first half of 2026, management reported €204 million in underlying EBITDA less rent, a 36% YoY increase. For the full year, the company now expects €430-460 million, representing a 28% YoY increase at the midpoint.

This is profit growth acceleration at the exact same time Basic-Fit’s capital intensity is declining.

Driven by a new strategy, Basic-Fit is opening fewer clubs, acquiring smaller tuck-in fitness chains, and shifting focus toward franchising, a capital-light avenue.

In other words: higher profitability, lower investments.

In H1 2026, FCF jumped from negative €57.4 million last year to positive €24.8 million this year, an increase of €82.2 million.

Now, I’m estimating €150 million in FCF for the full year, a number I derived in my previous update.

Using that €150 million as a starting point, it’s clear market expectations are low. To justify today’s share price, just 8% annual FCF growth is required over the next decade.

That’s a low hurdle. At the end of the forecast period, free cash flow will have to reach only €320.1 million.

Remember that underlying EBITDA less rent will already be significantly higher than that this year.

Moreover, Basic-Fit’s recent investments are proving highly accretive. Its 3-year rolling TTM ROIIC is 23.9%, while the 2-year rolling TTM ROIIC is 25.3% (based on underlying EBITDA less rent).

With the shift to franchising, ROIIC should remain high, as less capital is required to grow the business.

Taking the midpoint of underlying EBITDA less rent guidance (€445 million), we should see a 3-year rolling 2026 ROIIC of 21%, and a 2-year rolling 2026 ROIIC of 22.2%.

While it’s tricky to predict exactly what the economics of franchising will look like, if the ROIIC on the core business is already this high, franchising ROIIC should be even higher. The main difference is simply that franchising takes less capital to deploy.

That is key.

This trend of rising profits and declining investments will persist as the core business grows steadily alongside the franchise model.

At the same time, the market expects very little.

As mentioned, just 8% annual FCF growth is priced in. Based on a €150 million FCF estimate, the company trades at a P/FCF of just 13x.

Put differently: Basic-Fit is nearing an FCF inflection point driven by surging profit growth—which I covered in detail in my previous post—while investments are simultaneously expected to decline.

Basic-Fit remains one of my highest-conviction stocks.

Until next time,
Lucas
Summit Stocks

Disclaimer: the information provided is for informational purposes only and should not be considered as financial advice. I am not a financial advisor, and nothing on this platform should be construed as personalized financial advice. All investment decisions should be made based on your own research.

This month’s portfolio update is going to look quite different from the previous ones.

Put bluntly: I’ve sold my entire stock portfolio.

This isn’t because I’ve lost conviction in the businesses I own. In fact, my long-term conviction in these companies hasn’t changed at all.

I still think Amazon is bound for outperformance, and I would love to own companies like Basic-Fit and Nintendo again.

I wrote about Amazon just last week, so I still very much believe in the story.

The actual reason why I sold my entire portfolio is very simple: I’ve bought a home, my first ever home.

And in my case, a substantial portion of my liquid capital is now needed to fund that purchase.

While I would’ve (much) preferred to keep my stocks, I essentially had no choice.

Taking on additional debt to fund the home wasn’t an option, and trying to keep my portfolio invested with money that I need for the house obviously isn’t possible either.

Because of that, it actually made the decision somewhat easier. Without selling my stocks, funding the house wouldn’t have been possible.

This wasn’t an and/and situation.

Either way, it’s a significant change for me, particularly because investing has been such a large part of my life over the past few years, and because I’ve built a newsletter (now with 1,900+ subscribers) around the portfolio and my investing journey.

I think it’s important for you to distinguish between an investment decision and a personal capital allocation decision.

Selling Amazon, for example, doesn’t mean I suddenly think Amazon is a poor investment. Again, I believe Amazon remains poised for outperformance.

Selling Amazon means I currently have a more important use for the capital.

Leave a comment

What does this mean for Summit Stocks?

For the time being, the portfolio will obviously look very different since there won’t be a portfolio.

There won't be a list of holdings to update or a portfolio return to report. Instead, I'll be starting from a clean slate and rebuilding the portfolio with new capital over time.

I think this will make the next phase of the portfolio very interesting.

This forced reset allows me to think about how I want to structure conviction and concentration, and it allows me to rethink how I report my portfolio to you.

For instance, will I keep writing monthly updates? How many stocks do I want to own initially? Will I repurchase every stock I used to own, or will some be left out?

I'll have to put fresh money behind my highest-conviction ideas rather than simply carrying forward positions that I've accumulated over the past few years. Every position will once again have to earn its place in the portfolio.

For now, I expect that over the next 2-4 months, there won’t be a portfolio for me to report to you.

Perhaps I’ll be able to purchase a small number of shares earlier than that, but don’t expect too much. I’ve learned that a home costs a ton of money besides the actual purchase itself.

One thing that is 99% certain is that the first addition to the new portfolio will be Amazon.

My thinking as of today is that I want to accumulate a sizable position in Amazon before adding other positions. But that’s still up for debate.

Despite this enormous change, my investment philosophy itself hasn’t changed.

I’m still looking for quality businesses where market expectations are significantly lower than what the business is able to achieve. I do think I will be more strict about what qualifies as quality going forward. I won’t be investing in just any company simply because market expectations are very low.

I will also be stricter about the gap between market expectations and actual performance, as I have been over the past half year or so already.

In quality, I continue to look for high reinvestment opportunities, high returns on incremental invested capital (ROIIC), and market-leading positions.

The portfolio is unfortunately empty for now.

But that does give me time to look outward to other positions again, while continuing to keep an eye on the stocks in my watchlist (which now include my prior portfolio holdings).

I will continue to post earnings updates when they are due, I will continue to write deep dives, and I will continue to write about stocks I follow in general.

In other words, you will continue to receive 1-2 posts every week like nothing has changed.

That is all for this month’s portfolio update.

Wish me luck ;)

Subscribe now

Until next time,
Lucas
Summit Stocks

Disclaimer: the information provided is for informational purposes only and should not be considered as financial advice. I am not a financial advisor, and nothing on this platform should be construed as personalized financial advice. All investment decisions should be made based on your own research.

Without positive FCF, there is no value.

This is because the value of a business is comprised of all future cash flows discounted to today’s value of money.

If a business hypothetically earns $100 per year for the coming 5 years, and then disappears, an investor demanding a 10% annual return values that business at $379 (see image below).

This is basic finance.

And so, when a company’s FCF turns negative, investors grow anxious.

This is what we’re seeing with hyperscalers today, in particular Amazon.

Amazon is notorious for investing more than it earns. Throughout its entire history, FCF has been either negative or lower than what you’d expect from a company of Amazon’s size.

For context, FCF reached a record $33 billion in 2024.

Impressive when viewed in isolation, but this is against $638 billion in revenue. At the end of 2024, Amazon’s market cap was about $2.35 trillion.

This implies a 2024 FCF margin of just 5.2% and a yield of just 1.4%.

Now, Amazon’s FCF has turned negative once again. As a result, each time the stock reaches a new all-time high, investors turn around, take another look at FCF, and decide to sell.

Amazon has now faced two consecutive quarters of negative FCF, with many more (supposedly) coming in the near future.

This looks bad.

After all, value is ultimately created by future FCF, as I showed earlier. And so, when FCF is negative, there is no current cash flow to value.

But here’s the important distinction, phrased by none other than Michael Mauboussin:

“If a company can make an investment that earns a return in excess of the opportunity cost of capital, an estimate of the return that investors require, then future FCFs more than make up for the investment cost, and the value of the business goes up. In other words, a company with opportunities that promise sufficient returns should invest in all of them because those investments create value.”

Particularly important in this quote is the following: “…future FCFs more than make up for the investment cost…”

If a business is investing in projects that earn high returns, that business should invest in those projects.

But an investment costs money. In Amazon’s case, lots of money.

This temporarily depresses today’s FCF in return for much, much higher FCF in the future.

Here’s Amazon’s historic and future (using analyst estimates) ROIC and FCF:

As you can see, ROIC remains plentiful, well above the cost of capital, while FCF faces a trough in 2026-2027 before surging upwards.

You should also see that future FCFs more than make up for the negative period Amazon currently faces.

This is because today’s investments are expected to generate extremely attractive returns, well above the cost of capital.

As a shareholder, this is what you want.

I’d rather have management allocate the cash flows I have a right to to projects generating, say, a 30% return than have management return that cash to me.

And the assumption that today’s investments generate extremely attractive returns is no longer just an assumption.

This is supported anecdotally by Amazon’s management itself, but more importantly, by the numbers.

We know that returns on capital must be high for a business to create value. But perhaps more importantly, returns on incremental capital must be high as well.

Since invested capital is a sunk cost, it matters more whether today’s investments are also generating high returns.

In Amazon’s case, today’s investments largely consist of AWS capex.

In 2024, Amazon’s invested capital base increased by $82.6 billion. Since investments made in a single year don’t immediately yield results, we look at the incremental income generated a year later, or perhaps two years later depending on the investment, to estimate ROIIC.

In 2025, operating income grew by $11.4 billion. This results in a 2025 ROIIC of ~14% ($11.4 billion divided by $82.6 billion).

Decent, but not spectacular.

Over the trailing twelve months, Amazon’s operating income increased by $17.6 billion.

Twelve months prior to that, invested capital had risen by $98.6 billion.

In other words, TTM ROIIC is roughly 18%.

A rising trend.

Subscribe now

But these are rough estimates, and we know based on management’s remarks that AWS investments take about two years instead of one to monetize:

“Data center capital is spent starting two years before we can put servers into them to start monetizing. Once a data center opens with servers plugged in, we start generating significant revenue right away and then get to monetize these data centers for 30-plus years without having to spend that startup capital again.”

—CEO Andy Jassy

Thus, we should perhaps estimate ROIIC on a 2- or 3-year rolling basis.

This puts more emphasis on the actual return cycle of AWS investments, while also smoothing out ROIIC because investments can be lumpy from year to year.

If we calculate ROIIC on a rolling three-year basis, this is what the formula looks like:

You take the most recent income metric, subtract income from 3 years prior, and divide that by invested capital from a year ago minus invested capital from 3 years prior.

Based on that formula, Amazon’s 2-year rolling TTM ROIIC is 23.8%, while its 3-year rolling TTM ROIIC is 35.3%.

These are very respectable numbers.

All of a sudden, it makes sense why Amazon is pursuing this megatrend, spending $220 billion just this year.

“We now believe we will spend approximately $220 billion in cash CapEx in 2026… Even at that amount, we will still not have enough capacity to meet all the demand we have in 2026, and I believe this dynamic will also be true in 2027. In fact, the demand we already have for 2028 is striking… We long believed AWS could become a few hundred billion-dollar revenue business and now believe it'll be at least double that, and very possibly be a trillion-dollar annual revenue business for us in time, with very appealing accompanying free cash flow and return on invested capital.”

—CEO Andy Jassy

Jassy’s statement about AWS becoming a trillion-dollar business is not far-fetched, because, in just twelve months, Amazon’s invested capital has increased by an enormous $359.2 billion!

Remember that in 2024, invested capital grew by “just” $82.6 billion.

If we use a 30% ROIIC on that capital, roughly the midpoint of 23.8% and 35.3%, that translates into $107.8 billion in incremental operating income that is yet to arrive.

Because, as mentioned, it takes time for invested capital to yield results. This doesn’t happen right away.

And here’s what’s even better: Amazon isn’t done investing yet.

This is implied by guidance and management commentary on 2027 and beyond.

If Amazon keeps investing at this rate, and we see another $300-400 billion increase in invested capital, the results that follow will be extraordinary.

We could see operating income of over $300 billion in a few years (based on TTM operating income of $93.8 billion plus over $200 billion in incremental operating income).

A $300 billion operating income estimate implies Amazon is currently trading at just 9x that number.

Today, Amazon trades at 29x its TTM operating income.

If we assume that remains constant, and we assume Amazon reaches $300 billion in operating income in 2029, we could see a 2029 share price more than 3x higher than it is today (at least $750).

This implies an annual return of 48%!

Even if the multiple compresses to a market average of 20x, we see annualized returns of 31%!

And I know what you’re thinking: lots of assumptions and lots of napkin math.

I’m not saying these projections are accurate.

The point is to illustrate how today’s enormous capital expenditures can translate into substantial future earnings if Amazon continues to earn attractive returns on that capital.

And this way of thinking about Amazon’s investments does match what I’ve written above.

One thing is very clear: Amazon is investing an incredible amount of capital into what appears to be a high-return avenue, which should lead to exceptional incremental operating income over the coming years.

And that brings us back to where we started: free cash flow.

Yes, Amazon’s FCF is negative today. But that doesn’t mean the investments destroying FCF today are destroying value.

Based on our ROIIC estimates, Amazon’s investments are creating value, and lots of it.

If those investments continue to earn returns well above the cost of capital, the future FCF they generate should more than make up for today’s FCF deficit.

This is why I continue to hold Amazon as my largest position (by far) and don’t intent to sell, even when today’s share price is struggling to hold all-time highs.

Until next time,
Lucas
Summit Stocks

Disclaimer: the information provided is for informational purposes only and should not be considered as financial advice. I am not a financial advisor, and nothing on this platform should be construed as personalized financial advice. All investment decisions should be made based on your own research.

Relative to what exactly?

Market expectations.

I talk about this quite often because it’s one of the important concepts to understand if you’re looking for outperformance.

Say you’re valuing a high-quality businesses. That’s what we all want to own, so naturally, that’s where you’ll be looking.

But therein lies the problem.

The high-quality business you’re analyzing is also recognized as high quality by the market.

So while the average company might trade at 20x earnings, this business trades at 30x.

You start building a DCF model, and before long, because the valuation initially doesn’t work, you start to tweak the numbers. Growth gets a little higher, margins improve a little faster, and you also lower the discount rate just a bit.

Partly because you want to own the stock so badly, and partly because you give the business the benefit of the doubt because of its high quality.

This tends to work out poorly.

Look at typical high-quality companies such as Hermès, Copart, or Moody’s. They are vastly different businesses, yet all three are widely recognized for their quality.

And despite that quality, all three have underperformed the market over the past five years, with Hermès gaining just 3%, Copart declining 21%, and Moody’s gaining only 25%.

The lesson isn’t that high-quality businesses make bad investments. It’s that quality alone does not create alpha.

This is why you cannot look at valuation in isolation.

Valuation is relative.

When a company trades at a high P/E multiple, the market is already pricing in some combination of strong growth, high returns on capital, long reinvestment runways, or some other form of exceptional performance.

And if the company delivers exactly that, you may still earn a good return. But you shouldn’t expect much alpha from it.

You were simply right alongside the market.

At most, you’ll earn a return similar to a given discount rate (at least, according to theory), usually around 10% .

To generate alpha, you need a gap between what the market expects and what you believe will actually happen.

Subscribe now

That means looking for companies where expectations are low enough that the market is underappreciating the future.

This could mean finding a mediocre business that is better than the market thinks.

But the sweet spot is a great business that the market temporarily misunderstands, underestimates, or simply prices too conservatively.

This is where quality and valuation come together.

So-called quality investors and value investors aren’t, and shouldn’t be, separate branches. They intertwine and rely on one another.

You cannot be a great quality investor without considering valuation, and if you only consider valuation, you’ll just end up with a bunch of value traps because you ignored quality.

You can, and should, still invest in high-quality companies. But you need to ask a different question than “Is this a great business?”

Instead, you should ask: “How much of that greatness is already reflected in the price?”

It’s your goal to find the highest-quality businesses, where the gap between what the market expects and what is actually likely to happen, is biggest.

The first step in pursuing alpha is understanding this concept. The second is to use a set of tools to measure market expectations, from which you can analyze whether those expectations are too high, too low, or just right.

The most important tool is the reverse DCF model.

Reverse DCF

The market provides us investors with one particularly valuable datapoint: the stock price itself.

Instead of calculating intrinsic value ourselves, we can reverse-engineer the market’s valuation assuming that today’s stock price represents fair value.

By taking today’s price as a starting point, we can find out what the business needs to deliver to justify that price.

At its simplest, the outcome is the free cash flow growth rate required to justify today’s price, given a certain discount rate.

I always use a discount rate of 10%. This represents my minimum required return, or the hurdle a stock needs to clear for me to consider it attractive.

Consider Adobe.

I think we’ve all become familiar with Adobe by now. The stock is down more than 60% from its all-time high, while revenue and profits have continued to grow as if nothing happened.

Intuitively, this already suggests that market expectations have come down significantly relative to the underlying business.

Had revenue and profits fallen alongside the stock price, this wouldn't necessarily be the case.

Either way, expectations have clearly come down. At its peak in 2021, Adobe traded at a P/E of about 67x. Today, it trades at around 14x.

At the same time, Adobe generated less FCF in 2021 ($5.82 billion when adjusted for stock-based comp) than it does today, despite trading at a much higher price back then.

At its all-time high of ~$688, roughly 20% annual FCF growth was required to justify the share price using a 10% discount rate.

Moreover, most of the value was derived from the period beyond the first decade, the terminal period. This reinforces just how demanding the expectations were. The market wasn't simply expecting strong growth over the next few years. A substantial portion of the valuation depended on Adobe continuing to generate significant cash flows far into the future.

Today, this has completely flipped.

With more FCF and a much lower share price, essentially zero FCF growth is required to justify today's price.

On top of that, most of the value is now comprised of the cash flows generated during the first decade, the explicit forecast period.

This is what makes a reverse DCF so useful.

We can now take this 0.13% annual FCF growth requirement as our hurdle.

If we believe Adobe can grow FCF materially faster than 0.13% over the coming decade, then we have identified a gap between market expectations and our own expectations.

That gap is what creates the potential for alpha.

The larger the gap, all else equal, the greater the potential for our returns to exceed our 10% hurdle rate.

And since a discount rate of 10% is a safe proxy for market returns1, this leads to outperformance.

Understanding what the market expects from a business, and identifying if and where those expectations differ from your own, is a tremendously valuable concept to grasp as an investor.

Until next time,
Lucas
Summit Stocks

Let me know if you enjoy shorter posts like this one👇

Leave a comment

Disclaimer: the information provided is for informational purposes only and should not be considered as financial advice. I am not a financial advisor, and nothing on this platform should be construed as personalized financial advice. All investment decisions should be made based on your own research.

AppLovin ($APP), a mobile app advertiser, has undergone a remarkable transformation over the past several years.

The company, founded in 2012, went public in 2021 as a combination of several mobile gaming studios and advertising software. By the end of 2022, the stock had collapsed from ~$112 to just ~$9, a more than 90% drawdown, as growth disappeared and investors questioned the future of mobile gaming.

By 2025, AppLovin had vastly improved its advertising engine, sold its gaming studios, continued to generate extraordinary margins from its advertising platform, and became one of the market’s biggest success stories.

From 2024 to the end of 2025, shares were up roughly 1,800%! That’s an 18-bagger in the span of just 2 years.

But now, after another sharp and volatile decline, investors may have another opportunity to pick up a quality business with incredible potential at a discount.

Today, shares are down ~56% from their all-time high.

AppLovin’s long-term opportunity rests on a simple idea. The company has built a very effective advertising engine for mobile gaming. If it can replicate that success in e-commerce and other industries, while continuing to grow in gaming, the market opportunity may be much larger than what is priced in today.

In this deep dive, I will explain why AppLovin shares are in a drawdown, what is noise and what isn’t, and what the current valuation implies for investors.

What You’ll Read Today

1. Why the Shares Are Down, and Why the Opportunity Exists

2. The Business, the Moat, and the Growth Pillars

3. Valuation and Verdict

   ---

   Subscribe for free to discover high-quality investment opportunities. Upgrade to paid for deep-dive stock research, portfolio updates, and full access to Capaxe.

   ---

Why the Shares Are Down, and Why the Opportunity Exists

AppLovin is down approximately ~56% from its 2025 peak. But before discussing this decline, it is important to understand that AppLovin’s stock had already increased by 1,800% since 2024 and roughly 7,300% (!) since its 2022 low.

After falling to roughly $9 in December 2022, the shares entered one of the strongest rallies in the market. By September 2025, AppLovin had reached an all-time high of approximately $745 per share.

AXON 2.0, AppLovin’s AI advertising engine (more on that later), had turned AppLovin’s advertising platform into a remarkably profitable growth engine, advertising margins had climbed toward 80%, free cash flow had exploded, and investors were beginning to believe anything was possible.

At this all-time high, free cash flow would have had to grow by 25% per year for a decade to justify that price. That’s an enormous hurdle, especially when we consider the law of large numbers: free cash flow would need to reach $35 billion in 10 years if we assume 25% annual growth, up from “just” $3.7 billion in 2025.

When a stock has risen more than 70x, expectations become difficult to satisfy. Everything has to go right.

But the current decline has not been caused by one catastrophic event. Instead, investors have become increasingly concerned about the durability of AppLovin’s growth engine, while a handful of external and sentiment-driven factors have added to the pressure.

It is worth walking through these events in order.

2022: the growth collapse

The first major selloff began in early 2022.

AppLovin had reported 92% revenue growth in 2021, although organic growth was 70%.

However, Q4 2021 organic growth had already slowed to approximately 25%, and management’s 2022 guidance implied just 27-38% revenue growth.

AppLovin’s two businesses were moving in opposite directions. The Apps segment (AppLovin used to own the games they advertised on) was expected to grow only 4-11%, while Software Platform (the actual advertisement platform) revenue was expected to grow 100-123%.

For a company thought of as a hyper-growth stock, this was a major change in expectations. But the growth slowdown was only the beginning.

Apple’s App Tracking Transparency framework had made mobile advertising less effective, while the normalization of consumer behavior after COVID put further pressure on mobile gaming.

Advertisers became more focused on ROAS, while game developers reduced user acquisition spending.

AppLovin’s Apps business deteriorated rapidly.

The company subsequently cut its Apps revenue guidance, reduced user acquisition spending and began reviewing its gaming portfolio.

Then, in August 2022, AppLovin controversially sought to acquire Unity in an approximately $20 billion all-stock transaction. The proposal was rejected by Unity, and AppLovin eventually withdrew the offer.

The combination of collapsing growth, a weak mobile gaming environment, Apple’s privacy changes, deteriorating technology valuations, and concerns about management’s strategy sent the shares into a freefall.

By December 2022, AppLovin had fallen to approximately $9 per share.

2023: AXON changes the story

The first signs of a turnaround appeared in 2023.

The key catalyst was AXON 2.0, a major upgrade to AppLovin’s advertising technology.

To understand why it mattered, you have to look at the environment AppLovin was operating in. Apple’s privacy changes had made traditional user-level tracking much less effective, making it harder for advertisers to identify which users were most likely to convert.

AXON’s approach was different. Instead of depending primarily on device-level tracking, it could use the enormous amount of behavioral and conversion data flowing through AppLovin's ecosystem to make increasingly accurate predictions about which users were most valuable.

AXON 2.0 significantly improved those predictions. The result was better targeting, higher returns for advertisers, and ultimately, more demand flowing through AppLovin’s platform.

This also created a powerful feedback loop. More advertisers generated more data, better data improved AXON, and better performance encouraged advertisers to spend more.

The results spoke for themselves: AXON 2.0 was introduced in Q2 2023, after which revenue immediately accelerated to 21% growth in Q3 and 36% in Q4.

This was despite the Apps business continuing to drag the business. Software Platform revenue accelerated to 65% growth in Q3 (just 28% in Q2) and 88% in Q4.

Following its August 2023 earnings report, the stock jumped more than 25% in a single day.

AppLovin ended 2023 at approximately $40 per share.

The slow realization that AppLovin was drastically accelerating set the stage for the enormous rally that followed.

Subscribe now

2024: the market discovers the new AppLovin

The transformation became increasingly obvious in 2024.

Revenue reached approximately $4.7 billion, up 43% YoY, while net income increased to approximately $1.6 billion. Advertising EBITDA margins approached 80%, and free cash flow reached approximately $2.1 billion.

Advertising revenue (AppLovin had changed the name Software Platform to Advertising) grew 75% YoY to $3.2 billion, while Apps grew just 3% YoY.

AppLovin was producing the combination investors dream about:

High growth + enormous margins + extraordinary free cash flow.

The market began assigning a serious premium to AppLovin, treating the business as an AI advertising platform.

Shares increased almost 800% in 2024 alone.

At the same time, management began testing whether AXON’s capabilities could extend beyond gaming.

The first major target was e-commerce.

If AppLovin could use the same technology that helped a game developer acquire valuable users to help an online merchant acquire valuable customers, the addressable market would be dramatically larger.

This became the foundation of the next leg of the bull case.

2025: the second leg of the rally

The transformation continued in 2025.

AppLovin sold its ten mobile gaming studios to Tripledot, completing its transition from a company that owned games to one that fully monetized other people's games.

The slow growth of the Apps business would no longer drag AppLovin’s overall growth down: 2025 revenue growth (excluding Apps) accelerated to 70% YoY, reaching $5.48 billion.

The sale simplified the business, increased focus, and made the economics of the advertising platform much easier to see.

At the same time, e-commerce continued to scale.

AppLovin began expanding its advertiser base, developing self-service advertising tools, and integrating its technology with platforms such as Shopify.

Investors began to believe that the gaming flywheel could be replicated in e-commerce.

The result was another extraordinary rerating.

By September 2025, AppLovin had reached approximately $745 per share.

By this point, however, expectations had become dangerously severe, with 25% annual FCF growth embedded in the stock price (as shown earlier).

AppLovin’s P/FCF (based on 2025 FCF) was about 65x.

This is important context for understanding what happened next.

2026: the market starts questioning the durability of growth

Read more

My Greggs thesis is simple: as one of the largest food retail chains in the U.K., the business is continuously capturing market share and expanding its footprint toward 3,500+ stores.

In order to do so, Greggs has been investing heavily in its supply chain, essentially expanding capacity ahead of demand (which remains lower than usual due to a tough macro environment).

A result of these heavy investments is temporarily low free cash flow. My thesis therefore depended on a free cash flow inflection once Greggs finishes its investment cycle.

Today, the end is in sight: 2025 was the peak capex year, with Greggs deploying its cash into opening new stores, as usual, as well as expensive logistics facilities.

Specifically, capex reached £263 million in 2025 (or £288 million including intangible assets), while 2026 capex is now expected to reach just £180 million.

At the same time, operating cash flow continues to increase. These two factors form the perfect cocktail of a sustainable free cash flow inflection in the medium term.

The market hasn’t priced this in. Shares remain down 38% over the past 5 years; YTD, they’re up just 11%.

In this Greggs update, I discuss the H1 numbers and what those numbers mean for my thesis. If you haven’t yet, make sure to read my 2025 update first.

The Headline Numbers

In 2025, Greggs was plagued by a handful of macro issues in the U.K., causing the business to grow slower than usual.

2025 revenue grew “just” 7% YoY, compared to 11% in 2024 and 20% in 2023, as high inflation (5.6% for Greggs in 2025) caused a decline in volumes in the food-to-go market.

So far, 2026 continues to be a “difficult consumer market,” as Greggs calls it, yet growth was resilient.

H1 2026 revenue grew 7.2% YoY to £1.1 billion, with same-store growth of 2.1%. Overall cost inflation was 2.2%, which is expected to hold steady for the full year.

This, combined with continued store expansion, has led to revenue growth.

Greggs opened 34 net new shops, including newer concepts like “bitesize” stores. More concepts are on their way, such as “Greggs Express” stores, which serve as self-service, vending machine-like units.

Finally, Greggs opened its first international store earlier this year. The store is located in Tenerife’s airport, and “sales to date are very encouraging and are hitting all the hurdles that we [management] have set.”

That is encouraging indeed.

Greggs now has 2,773 locations, up from 2,739 locations at the end of 2025. The company is still aiming for 3,500 locations in the U.K. alone by opening at least 100 stores annually.

Profitability is also moving in the right direction.

Cost inflation was relatively low, as mentioned, and Greggs is well on its way to reaching its £11 million annual structural cost savings. In H1, £7 million has already been saved.

At the same time, revenue growth has led to operating leverage, and the H1 2025 period itself was relatively poor, allowing for easier comps.

These factors combined allowed Greggs to grow H1 operating profit by 23% to £87 million.

However, management expects full-year profit to be flat YoY due to operating costs from Greggs’s new facilities increasing by about £10 million in the second half of the year. Specifically, the Derby facility built last year will open in H2, thereby increasing operating costs.

Moreover, the Kettering facility opens in H1 2027. As a result of Derby’s annualized costs plus Kettering, management expects limited profit growth in 2027 too, with 2028 marking a return to profit growth.

Even so, the first half of the year was strong and is proving the thesis right.

Operating cash flow growth was solid, growing 18% YoY from £94 million in H1 2025 to £111 million in H1 2026.

At the same time, capex reduced from £175 million to just £77 million, meaning free cash flow increased from negative £81 million to £34 million.

For the full year, management now expects £180 million in capex instead of the prior estimate of £200 million, which should result in even higher free cash flow this year. The reduction is the result of a slight lowering of the shop opening target (from 120 to 100-110) as well as supply-chain facilities in Derby and Kettering being delivered below their respective budgets.

With £288 million in capex in 2025, Greggs is now officially past the peak of its investment cycle.

From this point onward, we will start to see more of the benefits of the investments made and less of the depressed cash flows.

Subscribe now

Early Positive Signs From The Investment Cycle

Not only has Greggs been investing heavily in its supply chain, the company continues to open stores.

Positively, management is getting better at opening stores, as well as innovating with its store formats.

The company has become more selective in its store openings (hence the lower 2026 target), but the quality of the openings has improved.

“You see we've nudged that [store openings] down slightly this year, but the quality of the openings and the performance of them is better, and that's important. And if you actually model out the impact of taking fewer shops at a higher sort of ROI versus trying to take more shops to leverage your capacity, it's better to go slightly slower with higher quality returns because you're still having to deploy all that shop capital.”

—Greggs CFO Richard Hutton

New stores are currently above their target of a 25% mature cash ROI within 2 to 3 years.

Stores continue to be incremental rather than simply cannibalizing the existing estate, which has become a more important topic as the store network has grown within the U.K.

62% of new stores opened in H1 were in areas without another Greggs within a mile, compared with 53% in 2025. Where a new store does overlap, average sales transfer has remained below 5%.

And more importantly, Greggs's app data shows that customers who gain access to a new store tend to visit Greggs more frequently rather than simply shifting their visits between locations.

As long as new store returns remain as they are today, the 3,500-store target remains a meaningful runway for high-return capital deployment. In other words, this is where cash flows should go, which is precisely what management is doing.

At the same time, Greggs is very soon moving from building capacity to using it.

The Derby and Kettering facilities were built to support a much larger store estate, with increasing automation intended to improve the efficiency of the supply chain. As these facilities move into operation, Greggs should increasingly benefit from the infrastructure it has already paid for rather than continuing to commit large amounts of capital to building it.

This creates an attractive combination: Greggs can continue investing in growth at high returns by adding new stores while requiring less capital to support the existing business.

Put differently, for its next ~800 stores (to reach 3,500), Greggs won’t need to invest heavily in its supply chain again.

This has a positive impact on returns on capital. Management's medium-term objective is to restore ROCE toward 20%, a very achievable target given that those investments are now behind us.

Over the next few years, we will see 3 things happen simultaneously: Greggs will continue to grow revenue, returns on capital will improve, and free cash flow will surge.

On top of that is the potential for improvements in the U.K. food-to-go market, which could serve as an additional tailwind.

The implications for the valuation are quite substantial.

Subscribe now

Valuation and Verdict

My early thesis assumed Greggs had the potential to generate over £200 million in FCF in 2027.

Since the share price has barely moved, this implies a P/FCF of about 10x. I think the situation today has become slightly less attractive in the short-term given the limited profit growth expected next year.

Yet the medium-term opportunity remains as attractive, if not more.

In H1 2026, Greggs generated £111 million in operating cash flow. Assuming H2 operating cash flow is similar to H2 2025 (£179 million), we get 2026 operating cash flow of roughly £290 million.

Since management doesn’t expect profit growth in 2027, we can conservatively assume operating cash flow should remain roughly stable just below £300 million.

Management expects about £180 million in 2026 capex and £160 million in 2027 capex.

This implies 2026 FCF of £110 million and 2027 FCF of £130 million. That then implies a 2026 P/FCF of 17x and 2027 P/FCF of 15x.

It’s quite clear that my early FCF estimates for 2027 were on the optimistic side, but keep in mind that the estimates I’m proposing now are on the conservative side.

If macro improves, or if we do see profit growth in 2027, there is clear upside.

Moreover, if we take £110 million FCF as a starting point, Greggs must grow FCF by just 9% annually over the next decade to justify today’s price.

That isn’t demanding, considering revenue growth is in the high single digits despite the tough macro environment.

My conviction remains strong, and I believe my Greggs holding in the portfolio of ~5% is still appropriate.

This is not an extremely high-quality business that demands a core position within the portfolio, but Greggs still presents a unique opportunity with FCF inflecting as the business comes out of the investment cycle.

Until next time,
Lucas
Summit Stocks

Disclaimer: the information provided is for informational purposes only and should not be considered as financial advice. I am not a financial advisor, and nothing on this platform should be construed as personalized financial advice. All investment decisions should be made based on your own research.

Hi there,

August was a steady month, where the portfolio returned 2.5% despite already increasing by 11.2% in July.

The August return was roughly in line with the return of the S&P 500.

This solid performance comes despite my largest holding Amazon declining 5% in the month, after the stock jumped significantly in July following yet another strong earnings report.

Performance can be attributed mostly to my software holdings: Salesforce, Adobe, and Duolingo. (I also own Topicus, another software name, but Topicus stock wasn’t able to benefit from the software momentum.)

August marked a month where the SaaSpocalypse narrative seems to have largely collapsed.

Salesforce sparked the reversal in the software stock downtrend, thanks to yet another strong earnings report, with solid revenue growth, even better cRPO growth, and continuous accelerating AI-related revenue.

Sentiment has shifted, and investors are—all of a sudden—bullish again.

Funny how narratives can shift in an instant. We’ll see if the narrative turns once again. All I know is that I’m confident in my software holdings, even before the change in sentiment.

I don’t just say this for the sake of sounding smart; I put my money where my mouth is.

This year, I added:

• Topicus shares at C$92;

• Adobe shares at $222;

• Salesforce shares at $169;

• And Duolingo shares at $113.

This is how we investors can benefit from unjustified changes in narratives. If we look back at Salesforce’s fundamentals today, we don’t see a business that’s still struggling. We don’t see a business whose stock deserved to decline ~58% (!) from its all-time high.

Of course the market is forward-looking, but the point remains valid.

Despite the portfolio increasing nicely over the past few months, I still believe it remains exceptionally well-positioned for strong returns going forward.

In this portfolio update, I want to dig a little deeper into sentiment changes, after which I will cover the portfolio in more detail.

Sentiment Is Not Fundamentals

The last few months have been a good reminder of just how quickly sentiment can change.

Prior to August, software stocks were among the least-loved areas of the market. The dominant narrative was that AI would fundamentally disrupt the SaaS business model, as I’m sure you’ve heard many times during that period.

If AI agents could perform tasks that previously required software applications, why would businesses continue paying for those applications?

That wasn’t the only argument, but it was one of the strongest.

Strong enough to drive significant declines across the sector.

Then, seemingly almost overnight, the narrative changed as a handful of software firms reported strong earnings, yet again.

But this wasn’t their first strong earnings report in a while. These software businesses have been reporting business-as-usual quarters all the time, even when the SaaSpocalypse narrative was at its worst.

The businesses didn’t change this month (at least not for the worse); the perception of those businesses did.

That’s because stocks are ultimately priced on expectations.

And when expectations change, valuations can change dramatically, even if the underlying business has barely moved.

This is also why narratives play such an important role in markets.

When investors are uncertain about the future, they construct a story to make sense of it. AI is going to destroy SaaS. AI is going to make software more valuable than ever. Interest rates will remain higher for longer. The consumer is weakening.

Some narratives are proven correct in time. Others aren’t.

The problem is that once a narrative becomes dominant, it can become difficult to distinguish the narrative itself from the underlying evidence.

This is where herd behavior comes into play.

Investors don’t form their opinions independently. We constantly observe what other investors are doing, what analysts are saying, which stocks are rising or falling, and which narratives are gaining traction.

In fact, the stock price itself is a powerful force that serves a narrative. When a stock declines, investors who believe in a bearish narrative will say “I told you so,” and the feedback loop only strengthens.

A stock falls because investors become more concerned about its future. The decline itself then becomes evidence that those concerns were justified. More investors become cautious, the stock falls further, and the narrative becomes even stronger.

The same process works in reverse.

And none of this requires investors to be irrational. If everyone around you appears to have reached the same conclusion, ignoring that conclusion by thinking independently is what seems irrational.

It’s why contrarian investors can feel sick to their stomach whenever they purchase an extremely beaten-down stock.

This is why market narratives are so powerful. They influence not only what investors think, but what they do. And once enough investors act on the same narrative, their actions influence the price, which then reinforces the narrative.

The problem for you to solve as an individual investor is which narrative is wrong, and which is right.

Whenever a stock falls significantly, it would be far too convenient to conclude that the market is simply being irrational.

Sometimes the market is right.

If a company’s growth is slowing, its competitive position is deteriorating, or its long-term economics are weakening, a declining stock price may be telling you something important.

The challenge isn’t to always automatically oppose the crowd when a stock or group of stocks falls.

It’s to determine what the crowd, or the market, is pricing in.

By reading the expectations embedded in a stock price, you gain valuable information about what the crowd thinks about future growth.

Opportunity emerges when sentiment changes for the better.

That doesn’t mean you need to predict when sentiment will change.

Trying to predict when investors will decide that software is attractive again or when the market will embrace one of my other holdings would turn investing into a guessing game.

Instead, I want to understand the relationship between sentiment and expectations, and fundamentals and valuation.

Sentiment determines what investors are willing to pay today.

Fundamentals determine what the business can ultimately become.

The gap between the two is where investment opportunities can emerge.

And this is precisely what I was trying to take advantage of when I added to my software holdings while sentiment was at its worst—while fundamentals weren’t.

Now that sentiment has improved for software, the market now expects higher growth for software going forward.

The important question then is whether these businesses continue to compound their intrinsic value faster than the market expects.

With that in mind, let’s take a closer look at what actually happened across the portfolio during August.

Portfolio Update

Read more

Prior to Salesforce’s earnings last week, the market was already starting to show more love toward SaaS stocks.

Their depressed valuations, combined with a continued lack of evidence supporting the SaaSpocalypse, had been slowly convincing the market that software businesses aren’t being disrupted.

This was the perfect scenario for Salesforce to inflect: another strong earnings report would further convince the market.

And convince the market is exactly what Salesforce has done: shares jumped 23% in a single day, now back above $250.

In just a single month, shares are up almost 40%. Since the June low, shares are up ~70%!

The market is realizing that enterprise software is not being replaced by AI. Enterprise customers aren’t building their own AI CRM tools, AI-native startups aren’t outcompeting incumbents, and software isn’t being commoditized.

I discussed this in more detail in my previous update, but for today, let’s take a look at this week’s earnings and what those earnings mean for the thesis.

The Headline Numbers

Put simply, the Salesforce thesis revolves around a wide-moat software business that generates highly predictable revenues and cash flows and is still able to grow at low double digits.

On top of that is the AI opportunity, mostly in the form of Agentforce. Agentforce is becoming Salesforce’s next growth engine, as the core products, such as Sales, Service, and Marketing, are growing at a single-digit rate.

For Q2 2027, Salesforce’s revenue grew 11% YoY to $11.3 billion, within guidance of $11.27 to $11.35 billion.

cRPO grew 14% YoY to $33.5 billion, ahead of management’s expectation of 13%. This is meaningful acceleration that supports the thesis that AI is actually contributing.

Agentforce ARR exceeded $1.5 billion this quarter, growing 240% YoY, a sequential acceleration from a quarter ago, when it grew 205% YoY.

More key numbers related to AI stood out:

• Salesforce processed 3.2 billion Agentic Work Units (AWUs), a discrete task accomplished by an AI agent, this quarter, up 97% QoQ. This compares to 111% QoQ growth in Q1 and 57% QoQ growth in Q4 2026.

• Slack delivered its fastest quarterly Net New Annual Order Value (NNAOV) since it was acquired, thanks to 150% QoQ Slackbot user growth to 1 million active users

Salesforce’s AI-related revenue continues to gain momentum, which is necessary for the thesis, as we’ll discuss soon.

In terms of guidance, the company initiated Q3 2027 revenue guidance of $11.42 to $11.5 billion, implying growth of 11% to 12% YoY. cRPO is expected to grow at a very solid pace of 14% YoY, compared to 11% YoY growth in Q3 2026.

Full-year 2027 revenue guidance was raised to $46.1 to $46.4 billion, up from $45.9 to $46.2 billion a quarter ago.

All in all, strong headline numbers.

Core Business Shows Strength

Don’t worry, we’ll inevitably discuss AI and how it relates to Salesforce soon.

But first, I want to shed some light on the Salesforce core business: the traditional CRM services.

In my previous update, I explicitly said “the core business, meaning no AI shenanigans, has slowed down.”

I suggested that, to continue growing at a healthy rate, Salesforce’s core business wouldn’t be strong enough. Agentforce then emerged as a potentially meaningful growth driver, as well as Slack’s growing importance.

That is still true.

But this quarter provided evidence that the core business may yet be stronger than first expected.

Growth of Agentforce Apps accelerated from 7% YoY in the previous quarter to 8% YoY.

More importantly, the underlying indicators of core demand were remarkably strong.

Management noted that net new annual order value (AOV) growth is now the strongest it has been in 4 years, while H1 net new AOV growth significantly outpaced overall AOV growth. This is keeping Salesforce on track for the organic revenue reacceleration management expects in the second half of the year.

At the same time, attrition is near its lowest level ever.

A recurring-revenue business needs to do two things well: retain its existing customers while adding new revenue on top of that.

Salesforce is doing both.

This, as has been case over the past several quarters, directly contradicts the SaaSpocalypse narrative:

“Our seats were supposed to decline. Instead, Agentforce Sales, Service, and Slack all saw YoY growth. We were told customers would abandon us, but attrition is near its lowest level ever… We were told customers would pay less for our products, but they are upgrading, and A1E and A4X bookings more than doubled QoQ. We were told contract lanes would shrink. Well, they improved across all segments, and we were told agents accessing our system would kill apps. Well, actually, usage is exploding. Agentic use of our apps surged sixfold. We were told the frontier models would eliminate the need for CRM. Instead, they depend on it.”

—Salesforce CEO Marc Benioff

In my previous update, I laid out 4 core risks introduced in the SaaSpocalypse narrative:

• Customers may build their own tools.

• AI-native startups could outcompete incumbents.

• Traditional SaaS pricing models could come under pressure

• AI may make software interfaces less important, commoditizing software

None of these risks have materialized, and based on the fact that seats are growing, customers aren’t leaving, customers are spending more, and AI usage within Salesforce is exploding, I don’t see any of these bear cases becoming reality.

That doesn’t mean the core businesses will all of a sudden reaccelerate drastically. Marketing and Commerce are still weaker businesses, and Service hasn’t been growing quickly this year either.

But that isn’t really the point.

My previous thesis assumed the core business could remain a durable, sticky software business while new AI products add another layer of growth.

Moreover, expectations embedded within the stock price have been low as shares have declined.

Shares traded at ~$192 at my previous update, which required Salesforce to grow FCF by just 5% annually.

In other words, even high-single digit growth (e.g. only the core business) would be sufficient to generate meaningful stock returns.

But with AI as a catalyst, returns can still be extraordinary for Salesforce shareholders.

Subscribe now

AI Is Salesforce’s New Moat

One of the original concerns surrounding AI was simple.

If AI makes software easier to build, why would companies continue paying Salesforce for CRM applications? If an AI agent can perform the work directly, perhaps the application layer, the software interface, becomes irrelevant.

But this assumes that the interface is the most valuable part of enterprise software.

Luckily for Salesforce, it isn’t.

The hard part is everything underneath the interface: customer data, metadata, permissions, security models, business rules, workflows, and context that allow an action to be executed correctly.

Salesforce owns this infrastructure, with its CRM solutions serving as the source of truth for enterprises.

This is where the Headless strategy, as discussed in my previous update, becomes particularly interesting. Salesforce increasingly doesn’t need to be the interface through which users interact with their data. Instead, it can become the underlying enterprise system that AI interfaces connect to.

We are already seeing this happen.

This quarter, agentic use of Salesforce through MCP and CLI calls surged sixfold, as customers increasingly accessed Salesforce’s data, metadata, and workflows through AI rather than its traditional UI.

And the Claudeforce announcement with Anthropic is the clearest example of what this could look like.

Claude provides the reasoning and intelligence, while Salesforce provides the data and infrastructure.

Users can access live Salesforce context within Claude, update pipelines, review deal health, take governed actions, and much more, without leaving Claude.

This way, employees no longer need to navigate “static UI,” as Salesforce calls it, with software becoming a “system that powers every interface.”

“Here, the UI is the AI—allowing you to build custom apps dynamically and answer any enterprise question. Probabilistic intelligence alone doesn't run a company, and deterministic systems don't reason. By fusing Claude's extraordinary reasoning with the trusted data, workflows, and governance every enterprise runs on, we're delivering a dynamic interface that thinks, reasons, and acts. This is how every business will run.”

—Salesforce CEO Marc Benioff

This is almost the opposite of the SaaSpocalypse narrative.

Claude isn’t replacing Salesforce. It sits on top of Salesforce. And Salesforce doesn’t need to build the world’s best AI model to benefit. It can let Anthropic, OpenAI, Google, or whoever ultimately wins the model race compete at that layer, while Salesforce remains the trusted enterprise system underneath.

The same dynamic is playing out with Slack.

Slack already made Salesforce’s CRM products stickier by connecting customer data and workflows to where employees communicate and collaborate. AI is now changing Slack’s role again.

Slack is becoming the interface where humans and agents work together.

Slackbot users grew 150% sequentially this quarter to 1 million active users, while Slack delivered its fastest quarterly Net New AOV growth since Salesforce acquired it.

“Slackbot is our fastest adopted AI product ever. Five months after launch, it has 1 million active users, up 150% QoQ.”

—Salesforce CEO Marc Benioff

Premium Slack upgrades have also tripled since Slackbot launched, while Slackbot has driven 8.1 million hours of annualized productivity gains at Salesforce itself, more than doubling QoQ.

This is the value AI delivers when combined with underlying enterprise data.

The workflow is changing from:

To:

The Salesforce UI disappears, but the platform remains.

From this perspective, investors can now think about Salesforce in five critical layers:

• Data 360 provides the context.

• Salesforce provides workflows and business logic.

• Agentforce provides the agents.

• Claude provides frontier reasoning.

• Slack provides the environment where humans and agents interact.

AI is therefore making the underlying Salesforce platform more valuable because of the enterprise context and infrastructure necessary for AI to actually do something.

This reinforces Salesforce’s moat in an AI-first world while simultaneously fueling growth.

But none of that matters if Salesforce’s valuation doesn’t make sense after a ~70% run-up in price since June.

Subscribe now

Valuation and Verdict

Before discussing the valuation, I want to briefly mention the $25 billion share repurchase program commenced in March this year.

In Q1 2027, Salesforce repurchased a total amount of shares worth $27.3 billion. This was done at an average share price of $198.34.

In other words, significant value has already been created considering that shares now trade about 30% above the repurchase price.

Salesforce’s diluted share count is now 10% lower than it was a year ago.

Of course, the repurchase program was funded with debt, but that essentially implies management at the time believed the expected return from repurchasing undervalued shares exceeded the cost and risk of the debt.

So far, they have been proven right.

In terms of valuation, Salesforce’s TTM adjusted FCF now stands at $11.5 billion, up from $11.1 billion a quarter ago. That translates into an adjusted (for stock-based comp) EV/FCF of 21x, up from 17x in my previous update and 16x in my February update.

Despite this increase in valuation, I still believe shares are undervalued.

Based on $11.5 billion in TTM adjusted FCF, the company needs to grow free cash flow at less than 8% annually to justify today’s price.

Management still targets $63 billion revenue by 2030, which implies low-double-digit revenue growth over the coming years.

That makes 8% FCF growth a relatively low hurdle.

Yet the margin of safety that was present during my previous update has now largely disappeared.

Salesforce is no longer a strong buy. I now consider Salesforce a buy below $300, and a strong buy below $200 (as was the case at my previous update).

Above that price point, FCF growth must grow double-digits to justify today’s price, which I consider quite a high hurdle.

Subscribe now

Until next time,
Lucas
Summit Stocks

Disclaimer: the information provided is for informational purposes only and should not be considered as financial advice. I am not a financial advisor, and nothing on this platform should be construed as personalized financial advice. All investment decisions should be made based on your own research.

Duolingo’s earnings report last week reaffirmed management’s user growth-over-monetization strategy.

DAU and MAU growth accelerated, while user retention reached an all-time high. Importantly, management now expects DAU growth to remain above its previous 20% guidance for the rest of the year.

But as in Q1, bookings growth decelerated to just 8% YoY while revenue grew 18%, also a stark deceleration from 42% growth seen in Q2 2025. Both are expected to slow further in Q3.

But this is what Duolingo told us to expect. And despite weak growth, both metrics beat guidance.

Duolingo’s new strategy wasn’t part of my original thesis, but as shares have come down so drastically, I think there is considerable value in Duolingo despite the slowdown.

2026 is a year in which Duolingo focuses on scaling first, and then monetizing later. During 2025, early indicators showed that Duolingo’s user growth was slowing. Management explained that this was caused by monetizing its users too aggressively.

Thus, Duolingo has taken 2026 as a year of experimentation regarding monetization versus user growth.

The company is sacrificing short-term monetization gains to build a larger and better product. This sacrifice is seen most clearly in Duolingo’s 2026 bookings guidance, which now stands at just 10.9% growth, compared to 33% growth in 2025.

If you’ll recall, 2026 bookings growth would’ve been around 20% had management not pursued this strategy.

Evidence from this earnings call suggests that Duolingo’s strategy is starting to work.

The Headline Numbers

Duolingo reported revenue of $298.5 million in Q2, up 18% YoY and slightly ahead of management’s guidance of $295.5 million.

Bookings increased just 8% to $289.1 million, although also ahead of the $283.5 million guidance.

As a reminder, bookings are an important financial metric, as they are a leading indicator of future revenue. Revenue is recognized over the subscription period, while bookings are recorded upfront when a customer pays.

Therefore, if bookings growth slows, revenue growth is expected to follow that pattern.

This year, however, I think investors shouldn’t focus on the financial numbers, given the new strategy.

The more important numbers are on the user side—and they were strong.

DAUs increased 23% YoY to 58.7 million, accelerating from 21% in Q1 and coming in ahead of management’s expectations.

MAUs grew 10%, reaching approximately 141 million. A solid improvement from Q1, when MAU growth stalled at just 6%.

Paid subscribers increased 17% to 12.7 million, which was slower than the 21% growth reported in Q1.

For Q3, management expects:

• 8.9% bookings growth to $307 million

• 11.1% revenue growth to $302 million

Another big slowdown for revenue, but somewhat of a stabilization for bookings. The important point is that full-year guidance remains unchanged, implying a decent acceleration in Q4 for bookings.

Based on the results, Q4 bookings should reach about $380 million, implying 13% YoY growth, while Q4 revenue should reach $315 million, implying 11% growth.

While no investor likes a slowdown in revenue, it misses the point of what Duolingo is doing in 2026.

The more interesting question is what this means for the stock. My updated valuation still implies a potentially very attractive annualized return, despite the shares being up ~35% since my previous update.

User Growth Is Starting To Reaccelerate

Read more

Amazon reported one of its strongest quarters in years last week.

AWS grew at its fastest pace in 18 quarters. Operating income crushed management’s own guidance. And despite already planning to spend an extraordinary $200 billion this year, management raised its capital expenditure guidance to $220 billion.

The market reacted positively: shares surged ~15% after earnings.

Finally, it seems the market is realizing that Amazon’s enormous AI investments are not destroying shareholder value, but creating it.

For several quarters, management has repeated the same message: today’s free cash flow is temporarily depressed because Amazon is investing ahead of demand. Data centers require years of construction before they generate revenue, while servers are only purchased once customer demand is already visible.

This quarter, investors finally seem to be listening.

What may have changed their minds was the following:

“We long believed AWS could become a few hundred billion-dollar revenue business and now believe it'll be at least double that, and very possibly be a trillion-dollar annual revenue business for us.”

—CEO Andy Jassy

When I first invested in Amazon in 2024, my thesis rested on four observations:

• First, Amazon's revenue mix was steadily improving as higher-margin businesses such as AWS, advertising, subscriptions, and third-party seller services became a larger share of revenue.

• Second, Amazon's scale economies shared flywheel remained one of the strongest competitive advantages ever created.

• Third, Amazon possessed extraordinary optionality, repeatedly turning internal capabilities into entirely new businesses. AWS, Fulfillment by Amazon, advertising, and third-party seller services all followed a similar playbook.

• Finally, Amazon's culture, established by Jeff Bezos and continued by Andy Jassy, consistently prioritized maximizing long-term intrinsic value over optimizing quarterly results (as it is doing today).

Two years later, I believe every one of those observations has strengthened.

Patience and conviction are paying off.

The Q2 2026 Numbers

Amazon reported revenue of $200.6 billion during the quarter, an increase of 20% YoY. Operating income reached $27.5 billion, growing 43% and comfortably exceeding management's guidance of $20 to $24 billion.

AWS once again proved its worth.

It increased 37% YoY to $42.2 billion, accelerating from 28% last quarter and marking the fastest growth rate in 18 quarters. Annualized revenue now stands at nearly $170 billion.

Even more remarkable was the continued expansion of Amazon’s backlog, which reached $496 billion, compared to $364 billion just three months ago.

Advertising revenue accelerated to 26% growth. Third-party seller services accelerated to 16%. Even online stores, Amazon's most mature business, accelerated to 15% growth.

This is important.

Most of the discussion surrounding Amazon revolves around AWS. But Amazon is not becoming a stronger business just because AWS is performing well.

Amazon is becoming a stronger business because almost every major business segment is improving simultaneously.

Finally, management also raised full-year capital expenditure guidance from $200 billion to $220 billion, bringing expected investment to an even higher level.

One of Amazon's strongest quarters in years. More importantly, management provided clarity (again) on why today's AI investment cycle should generate attractive free cash flow and returns on invested capital over the coming decades.

Leave a comment

The Market Finally Starts To Recognize The AWS Investment Cycle

Higher capex means lower free cash flow today, and investors naturally become skeptical when free cash flow turns negative.

In the first half of 2026, Amazon generated negative free cash flow of $24.9 billion. Adjusting for stock-based compensation, that figure was closer to negative $35 billion.

Concerning when you do not know the context.

But free cash flow only tells you when cash leaves the business, not why.

Amazon spent much of the earnings call explaining precisely where today’s cash is going and why management believes these investments will generate exceptional returns over the coming decades.

Amazon’s AI infrastructure consists of two very different investments: data centers and servers.

Data centers require enormous upfront capital and often begin construction nearly two years before generating a single dollar of revenue. Once completed, however, those facilities can remain productive for more than 30 years.

Servers, which are placed inside the data centers, follow a different economic model.

Amazon purchases them only once customer demand is already visible. On average, management expects those investments to break even in less than three years, while the servers themselves remain useful for five to six years. Even more importantly, most AI capacity is already being contracted under agreements lasting at least five years.

As Andy Jassy explained:

“For servers and networking equipment, on average, it takes a little less than three years to break even… The servers currently have a useful life of at least five to six years, and most of our AI capacity these days is being contracted for at least five-year terms. That means that we're driving significant free cash flow on the servers and networking equipment in the two to three years after we break even.”

Today's free cash flow is being depressed because Amazon is building assets that have not yet started producing revenue.

We have seen investment cycles like this before, just at a smaller scale.

The first wave of AWS investments initially depressed profitability before eventually generating significant cash flows.

Management is explaining that AI represents another version of that same opportunity, just at a much larger scale.

So far, the evidence supports that view.

What we are seeing in AWS today is direct proof that Amazon’s capital spending is generating real economic returns.

AWS has now accelerated for the fifth consecutive quarter, culminating in 37% YoY growth this quarter, its fastest pace in 18 quarters.

During the last period AWS was growing this quickly, it was less than half its current size. Today, annualized revenue has reached approximately $169 billion, compared to $123 billion just one year ago.

What is even more remarkable is AWS’s backlog, which reached almost $500 billion, growing at a triple-digit-rate.

Just as importantly, management continues to emphasize that demand remains well ahead of available capacity.

Amazon expects AWS to remain capacity constrained throughout both 2026 and 2027, while Andy Jassy described demand for 2028 as already being "striking."

Amazon is not just throwing cash against the wall, hoping for a positive outcome.

The bottleneck is not customer demand. It is Amazon’s ability to build data centers fast enough.

Over time, we should only see AWS continue accelerating, at least in absolute dollar terms.

This is because the capex Amazon is spending now will also (obviously) have future returns. According to management, data center capital “is spent starting two years before monetizing.”

In other words, the 37% AWS growth we are seeing today is largely driven by infrastructure investments made several years ago.

We have barely begun seeing the returns from last year’s capex, and we certainly haven’t seen the returns from this year’s record investment program.

Obviously, this is a simplified way of thinking about AWS. Capital is deployed continuously, projects come online at different times, and demand itself evolves.

But the conclusion remains the same.

Much of Amazon’s future earnings power is, as you are reading this, still under construction.

What will AWS growth look like in 2027 or 2028? Nobody knows.

What I do know is that Amazon has repeatedly demonstrated an exceptional ability to forecast long-term demand.

Considering management already sees strong demand extending into 2028, I remain highly optimistic about AWS’s future.

Subscribe now

Optionality Continues To Be Underappreciated

AWS may seem Amazon’s most important business, but it is far from the only reason I own Amazon.

One of the original pillars of my investment thesis was Amazon’s optionality.

Very few companies repeatedly create entirely new businesses from capabilities originally built for internal use.

AWS was first built for internal use. So was Fulfillment by Amazon. Advertising emerged from Amazon’s very own marketplace. And third-party seller services grew out of Amazon’s own retail network.

This quarter provided us another reminder that innovation continues.

Supply Chain Services officially launched this quarter, allowing other businesses to use Amazon’s logistics network without selling products through Amazon’s marketplace.

“Any business can move, store, and deliver everything from raw materials to finished products using the same supply chain that supports Amazon.

—CEO Andy Jassy

Large customers including Procter & Gamble, 3M, Lands' End and American Eagle Outfitters have already signed up.

It remains far too early to estimate how big this business could be, so I will not speculate. But history suggests Amazon is quite successful at selling an internal capability to external customers.

Amazon Leo also continues progressing toward commercial deployment, with nearly 400 satellites already in orbit and meaningful customer commitments secured before commercial service even begins.

Both initiatives are full of potential.

Amazon’s ability to repeatedly create entirely new businesses remains unique.

Retail Keeps Improving

Amazon's retail business also continues improving across nearly every meaningful metric.

Revenue growth accelerated across online stores, third-party seller services and advertising.

And the underlying KPIs are equally encouraging:

• Amazon Haul expanded its assortment twenty-fold since launch, now offering more than six million products priced below $10.

• Monthly active grocery customers increased more than 50% since the start of the year, while same-day grocery orders now contain more than three times as many items as before.

• Amazon Pharmacy doubled new customer additions during the first half of the year, while same-day prescription deliveries increased fivefold.

• The company delivered more than 40% more items through same-day or overnight delivery compared to last year.

• Alexa Shopping engagement nearly doubled, while total interactions increased fivefold.

This is a business that is seemingly unstoppable. Investors focus entirely on AWS, but Amazon’s retail businesses are also flourishing.

That matters because retail still represents an enormous business, much larger than AWS.

Even small improvements in profit margins create tremendous amount of incremental profit. Robotics should continue contributing to that trend.

Management expects to more than double its fleet of robotic arms during 2026, while North American and International operating margins continue their gradual upward trend.

Subscribe now

Valuation and Verdict

This past quarter once again reinforced the strength of the Amazon thesis.

The only thing that has really changed is that the market seems to finally recognize it.

AWS demand is stronger than ever. Optionality keeps expanding. And retail, while less exciting than AWS, keeps improving meaningfully.

Management continues to take a long-term approach. Rather than optimizing quarterly free cash flow, they continue investing aggressively where expected long-term returns justify today’s spending.

Traditional valuation methods remain mostly useless for Amazon, precisely because of the company’s long-termism.

P/E and P/FCF ratios send the wrong signals, while a DCF does not work because free cash flow is expected to remain negative in the short term. We have to look further than an arbitrary five or ten years for Amazon, but then we would simply be guessing.

I prefer to look at Amazon’s P/OCF ratio, which now stands at roughly 18x, compared to 20x in my previous update, despite the run-up in the share price.

This is because TTM operating cash flow grew from $148.5 billion to $161.4 billion QoQ.

A year ago, Amazon’s operating cash flow stood at $121.1 billion, implying growth of 33.3%.

But instead of asking whether Amazon looks cheap based on next year’s earnings or cash flows, as P/E or P/FCF ratios do, I think the better question is whether the business has become more valuable faster than the stock has appreciated.

In other words, has Amazon become relatively cheaper or relatively more expensive?

I believe in the former.

What I find most compelling is Amazon’s rapidly growing invested capital base, which now stands at $854 billion, up 22% YoY.

Over the past decade, invested capital grew at an average rate of just 9% annually. Amazon is deploying capital today at a pace far above its historical average.

Obviously, right now Amazon is not earning high returns on this capital.

Returns on capital naturally decline while billions of dollars are being invested in assets that have not yet begun generating meaningful revenue. Data centers are constructed years before they begin producing cash flows, while AI infrastructure only starts monetizing once capacity comes online. Today’s earnings therefore understate tomorrow’s earning power.

To illustrate the point, let's perform a simple thought experiment.

Suppose Amazon stopped expanding tomorrow and simply operated as a mature business.

This is not going to happen, but stay with me.

Imagine the existing capital base eventually earned a 20% return on invested capital.

That would imply roughly $171 billion of annual operating profit on today’s invested capital base. At Amazon’s current valuation, that equates to roughly 17x earnings.

Again, this is not a valuation model.

It is a way of visualizing Amazon’s earnings power.

Another interesting observation is how quickly Amazon’s capital base has expanded.

Just 2 years ago, invested capital was a little under $400 billion. Since then, Amazon has deployed roughly $458 billion of additional net capital (meaning after depreciation).

Even if we conservatively assume that only half of that incremental capital ultimately earns a 20% return, that would represent roughly $46 billion of additional annual operating profit that has yet to fully appear in today's financial statements, since AWS capex takes time to be monetized, as we discussed.

Amazon as a mature business today is already valued attractively. Amazon tomorrow, once the current investment cycle is behind it, is an incredibly attractive business that will undeniably generate billions more in cash flow than it does today.

Even after the run-up in share price, I consider Amazon a buy.

My conviction remains as high as ever, as reflected in the portfolio, where Amazon represents more than 30% of my holdings.

I put my money where my mouth is and believe Amazon will continue to handsomely reward investors.

Until next time,
Lucas
Summit Stocks

Disclaimer: the information provided is for informational purposes only and should not be considered as financial advice. I am not a financial advisor, and nothing on this platform should be construed as personalized financial advice. All investment decisions should be made based on your own research.

Hi there,

The portfolio has done well over the past few months, and July was no exception, with a return of 11.2%, compared to ~0% for the S&P 500.

One of the hardest parts of investing is watching your highest-conviction ideas get cheaper while everyone else is, or at least seems to be, outperforming.

Earlier this year, that is exactly what happened. While software stocks remained deeply out of favor, I continued adding to several of my highest-conviction positions, including Amazon, Topicus, Adobe, and Duolingo.

Staying disciplined and adding to high-conviction names in my portfolio while they were cheap has been rewarding so far. July brought strong earnings reports from my holdings, including Amazon, which grew AWS at 37% YoY, as well as Basic-Fit, which continues to show strong operating leverage.

In addition to that, the market rotated out of the most popular AI names this month (e.g. Micron, Sandisk, ASML), into hyperscalers and quality stocks in general.

As a result, since March, when both the portfolio and the broader market reached their deepest drawdowns, the portfolio has returned roughly 20%, compared to around 10% for the S&P 500.

While this is just a short period, I expect my stock picks to continue compounding and delivering strong results.

In this portfolio update, I want to briefly discuss the importance of conviction within this context, after which I will cover the portfolio in more detail.

Investing Is a Mind-Game

Conviction is a strong belief in an idea or opinion. Conviction is only conviction when that belief does not change easily.

The problem is that investors are bombarded with information every day, all of it encouraging them to change that conviction. You may immediately think of Wall Street headlines, financial shows, or journalists claiming to have found the next big winner.

But I believe the most influential, and most dangerous, piece of information is the stock price itself.

Businesses change slowly over time, while earnings are reported quarterly. Stock prices, however, change every second. Because prices update constantly while fundamentals do not, our brains naturally give stock prices more weight. We begin to confuse changes in price with changes in business quality.

In other words, we let stock prices dictate our sentiment, or even worse, our conviction.

Over the long term, stock prices and business fundamentals move together. But in the short term, it is not that simple.

When Alphabet traded at ~$150 in 2025, no one wanted to own it. Today, with the stock trading at ~$370, everybody wants to own it.

We may be seeing something similar with Amazon today, although only time will tell. Now that Amazon’s shares are climbing higher, people want to own the stock. When shares were closer to $200, very few did.

The point is that, as an investor, you own actual businesses, not stock prices. If your conviction is based on the strength of the underlying business, a falling stock price alone should not suddenly change your mind. A lower share price does not automatically mean a weaker business.

Over the past few months, I have been able to take advantage of this phenomenon. By staying disciplined, stepping back from the stock price, and focusing only on the fundamentals, I have been able to buy more shares of high-quality businesses at prices well below today's.

The portfolio is now in what I believe is an extremely healthy position, with more compounding and growth ahead.

Portfolio Update

As of July 31, the portfolio consists of the following holdings:

Read more

Basic-Fit ($BFIT), my third-largest holding, reported H1 2026 results this week.

This was the company’s first half-year result since introducing its new “Era 3” strategy earlier this year. (More on this in my previous update below.)

Over the past year, Basic-Fit shares have gained ~40%, recovering strongly from a prolonged slump. From the 2025 low of about €17, shares are now up more than 100%.

Investors have become increasingly convinced that management’s shift toward higher-quality growth is starting to pay off. The company has slowed its pace of club openings, prioritized free cash flow, and introduced franchising as a third growth pillar after acquiring Clever Fit last year, alongside organic expansion and bolt-on acquisitions.

My original Basic-Fit investment thesis could be summarized in four pillars:

• Basic-Fit is Europe’s dominant value-for-money fitness chain, by far.

• The European fitness market still offers a long runway for expansion, driven by tailwinds such as urbanization, rising lifestyle diseases, and lower penetration rates than the U.S.

• Basic-Fit’s reported profitability understates the company’s true earnings power because rapid expansion costs mask the profitability of mature clubs.

• Slower expansion, at some point (I did not know when at the time of my deep dive), should unlock substantially higher free cash flow and returns on capital.

Now, a fifth bullet has been added to the thesis:

• Franchising has the potential to become a high-margin, capital-light growth engine alongside the core business.

We are now in the era of slower expansion, which is why profitability is growing rapidly, with the shares following.

The latest earnings provide another opportunity to evaluate the thesis and see how the business is doing.

The H1 2026 Numbers

Basic-Fit reported revenue of €800 million for H1 2026, up 18% YoY.

Total membership grew by 1.5 million YoY to 6.1 million members. Excluding Clever Fit, membership increased 15% from 4.5 million to 5.2 million. In other words, even without the acquisition, growth was still very healthy.

Underlying EBITDA less rent grew 36% to €204 million. As a reminder, this is the rough operating cash flow metric Basic-Fit uses. After subtracting capex, interest, and taxes, you get an approximation of free cash flow.

Operating profit increased 67% YoY to €95.8 million, with an operating margin of 12%, compared to 8.5% in H1 2025. Free cash flow improved from negative €57.4 million to positive €24.6 million.

This is textbook operating leverage, and it directly reflects bullet three of my investment thesis:

“Basic-Fit’s reported profitability understates the company’s true earnings power because rapid expansion costs mask the profitability of mature clubs.”

By slowing the pace of owned club expansion, Basic-Fit has allowed its existing clubs to mature while also reducing expansion costs. A twin engine, so to speak.

Revenue growing 18% while operating profit grew 67% is a fantastic achievement and direct proof that the thesis is playing out as expected. More importantly, this gap should continue to widen as more clubs mature and expansion remains deliberately slow.

We are likely still in the early stages of this operating leverage story.

Another sign of this operating leverage is that the average member count per owned club increased by 229 YoY to 2,999 members. This means these additional members are contributing almost directly to free cash flow because most of Basic-Fit’s costs are fixed.

Adding more members to an existing club base is highly accretive to profits.

The number of clubs increased from 1,628 to 2,192, mainly because the Clever Fit acquisition added 491 franchise clubs. Excluding Clever Fit, the network expanded by 35 clubs, keeping the company well on track to reach its goal of 50 new clubs in 2026.

Basic-Fit has essentially already achieved and surpassed this target, as the company also announced the acquisition of wellyou, a German chain with 41 clubs. The acquisition is expected to close in Q3 2026.

Finally, full-year revenue guidance of €1.64 to €1.69 billion was reiterated, while EBITDA less rent guidance was raised to €430 to €460 million, from €415 to €455 million.

Management also reaffirmed that it expects a “significant improvement in positive free cash flow in 2026.”

All in all, the headline numbers are promising once again.

Subscribe now

Is the Investment Thesis Playing Out?

Since my first purchases of Basic-Fit in January 2025, the investment thesis has been playing out nicely.

The headline numbers have been strong and continue to improve. Reported profitability keeps rising, capital-heavy expansion is slowing to make room for franchising, and free cash flow is inflecting.

As the shares move in the right direction, so does the business.

Hidden Earnings Power Continues to Emerge

Basic-Fit’s financial reports have never been easy to read. Management uses several non-IFRS metrics, and some seem intentionally vague or misleading.

While I do not fully agree with this style of reporting, the reality is that Basic-Fit has always seemed to be a low-quality company with poor returns on the surface. But once you look beyond that, a much higher-quality business emerges.

Once you dig in, you realize that reported profitability has never accurately reflected the economics of the business.

For years, the company opened well over one hundred clubs annually. Every newly opened club weighs on reported earnings during its early years, even though mature clubs generate attractive returns. As a result, the financial statements consistently included hundreds of clubs that were still ramping up.

Even now, as profitability surges, true profitability is still being masked.

The average membership per club increased from 2,770 to 2,999 over the past year, but we are still far from maturity. A mature club averages 3,250 members. That means the average club today can still add another 250 members.

With 1,751 owned clubs (I exclude Clever Fit because mature Clever Fit clubs typically have fewer members, but we do not know the exact number), Basic-Fit can still add nearly 440,000 members without doing anything (250 members times 1,751 clubs). This comes at no cost and almost entirely flows through to cash flow.

It is only a matter of time before we reach full maturity. During the first half of 2026, the Basic-Fit branded network added 269,000 memberships despite opening 18 fewer clubs than in the same period last year. Existing clubs are increasingly driving growth instead of newly opened locations.

Basic-Fit has also recently changed its new club strategy. New clubs are now designed to be better, bigger, and more modern. This does come at a cost. Expansion capex per newly built club increased to €1.47 million from €1.38 million last year.

But this should improve customer satisfaction and retention. While management did not explicitly say so, I can only imagine that bigger clubs can accommodate more members. Over time, this should allow the average membership per club to continue rising even more, unlocking more hidden earnings power.

If that proves true, today's slightly higher capex per newly built club could ultimately produce meaningfully higher returns over a club's lifetime.

Slower Expansion Is Unlocking Free Cash Flow

When I published my original deep dive, it was clear that it was only a matter of time before Basic-Fit stopped reinvesting everything.

I never doubted that Basic-Fit could generate attractive free cash flow. The only uncertainty was simply when management would decide to prioritize cash generation.

That moment arrived when Basic-Fit acquired Clever Fit last year.

During H1 2026, expansion capex declined from €68.1 million in H1 2025 to just €39.6 million. Maintenance capex was €54.4 million, compared with €57.6 million last year, or €31 thousand per club versus €36 thousand last year. Other capex totaled €10.7 million, and management still expects other capex to reach €25 million for the full year.

In my previous update, I estimated total capex for the full year at about €205 million, down from €271 million in 2025 and €335 million in 2024.

This still excludes acquisition capex, such as the wellyou acquisition in Q3 and the acquisition of 17 Clever Fit franchise clubs from one franchisee. But the point about declining core capex still stands. And even if Basic-Fit ends up reinvesting most of its cash flow, so be it.

I would prefer management to continue finding high-return reinvestment opportunities.

For now, H1 2026 capex totaled €104 million. Since management expects significantly higher free cash flow in the second half of the year, and since management already achieved its target of opening 50 clubs, €205 million of total capex now seems too high.

The midpoint of EBITDA less rent guidance is now €445 million, compared with €435 million at my previous update. If we annualize the €33.4 million of interest and tax cash expenses from the first half (for the sake of simplicity), we arrive at €66.8 million for the full year.

€445 million minus €205 million in capex (which I now believe to be a high estimate) minus €66.8 million in taxes and interest results in ~€173 million in free cash flow for the full year.

Considering free cash flow was "just" €24.6 million in H1, Basic-Fit would need to generate almost €150 million of free cash flow in H2 alone to bridge that figure. Free cash flow is almost always higher in the second half, but this is probably a bit too optimistic.

Then again, capex should be quite low in H2. But all this is still speculation.

The point is that €150 million of free cash flow in 2026, as discussed in my previous update, still seems achievable, implying that Basic-Fit trades at a P/FCF of about 17x. More importantly, 2026 is just the first year of a structurally higher cash generation phase. Even if Basic-Fit fails to achieve €150 million in free cash flow, much more will follow in the coming years.

As the title of this section suggests, slower expansion is clearly unlocking free cash flow. I can only imagine what numbers Basic-Fit will generate in 2027 and beyond.

Franchising Is Showing Its Promise

Following the Capital Markets Day, it became clear how management is turning franchising into a core pillar of Basic-Fit’s long-term strategy.

Management introduced franchising as the third leg of Basic-Fit’s growth strategy, alongside organic expansion and bolt-on acquisitions. While the company already operates a franchise network through the Clever Fit acquisition, that was never the end goal.

Clever Fit was, and still is, a means to an end. It was Basic-Fit’s way of entering the fitness franchising industry and acquiring the know-how and processes needed to operate a franchise business at scale.

The real opportunity is franchising the Basic-Fit concept itself.

Management expects its capital-light franchise business, which generated just €10 million of revenue in H1, to gain real momentum during the second half of the year with the launch of the Basic-Fit branded franchise proposition. This is an important milestone because it marks the beginning of using the company’s own operating model instead of simply managing the inherited Clever Fit franchise network.

Management ultimately wants to operate more franchised gyms than owned gyms. That is at least a decade-long ambition, but it makes perfect sense.

Basic-Fit has spent decades building a club network with a strong brand, technology platform, purchasing network, operating processes, and member proposition. Franchising allows the company to essentially capitalize on its own success without having to fund every new club itself.

Because it is so capital light, franchising is another engine that will improve both free cash flow and returns on capital over time, making it a very attractive avenue for long-term value creation.

I look forward to seeing how much franchising revenue Basic-Fit can generate in H2 2026 and throughout 2027.

You can find more information about franchising in my previous update.

Thoughts On Valuation

I already briefly touched on valuation when discussing expected capex for the full year.

My estimate of at least €150 million in free cash flow remains valid, although it now looks like a steeper jump. Basic-Fit would have to generate €125 million of free cash flow in H2 to reach that figure.

The shares traded at about €29 when I published my previous update, where I estimated a P/FCF of 13x based on €150 million of free cash flow.

Now the multiple has increased to 17x, while the uncertainty around achieving €150 million of free cash flow has also increased.

In other words, shares have become more expensive.

That does not mean I think they have become too expensive.

I still think it is a waste of time and energy to project cash flow several years into the future because we simply do not know enough about the potential of franchising. But one thing seems clear. As franchising accelerates and clubs continue to mature, free cash flow will keep expanding, and margins should also improve.

If €150 million this year is achievable, I do not see why €200 million would not be achievable next year, or even €250 million. That is why I think focusing on today's free cash flow multiple (even though it remains on the low side) misses the bigger picture. The more important question is what this business can earn once the current transformation is largely complete.

Management also expects significant free cash flow growth to continue over the medium term (considered 3 to 5 years by management).

I remain extremely bullish on Basic-Fit.

In my opinion, Basic-Fit remains a strong buy below €30. Even at current prices, I still rate the shares a clear buy.

Until next time,
Lucas
Summit Stocks

Disclaimer: the information provided is for informational purposes only and should not be considered as financial advice. I am not a financial advisor, and nothing on this platform should be construed as personalized financial advice. All investment decisions should be made based on your own research.