When I published my previous Hims & Hers (HIMS) article on October 20, I told you Q3 would come on November 9.

In the three weeks in between, quite a bit happened: Hims launched testosterone injections and a testosterone pill, Eli Lilly started selling its GLP-1 pills and injections through Hims & Hers, the famous health influencer Andrew Huberman, Ph.D., neuroscientist and Stanford professor, joined the board and the first peptides got a launch date. Things are happening fast at this company.

And the stock? It did what this stock does so often: jump around like an ADHD kid on a sugar high. It closed at $26 on Friday, the day of the Eli Lilly announcement, on Monday, it fell 11.8% to $22.94 before the results were out, without a clear reason and on Tuesday, after the results, it fell another 16.6% to $19.13. That’s a 26% difference in two trading days and 34.5% below the $29.21 at which I published my Q2 analysis three weeks ago.

Get a 15% discount for Fiscal AI here

To be honest, I’m very curious as I write this intro. Before this article, I had not listened to the call yet and had not looked any deeper than the results without any context. So, here we go.

The Numbers

Revenue: $890M, up 49% YoY, and exactly at the midpoint of the $880-900M guidance, but $2.9M or 0.3% below the $892.9M consensus. That’s without $18M in weight-loss membership payments Hims & Hers received in Q3 but can only count as revenue in Q4. With it, revenue would have been $908M, a clear beat of the consensus. That shows how arbitrary this sometimes is.

US revenue: $657M, up 21% YoY, after +16% in Q2 and -8% in Q1.

International: $233M, up 321% YoY; $135M of that came from Eucalyptus.

Subscribers: 2.79M, up 13% YoY, but about 100,000 fewer than in Q2. The market didn’t like this.

Monthly ARPU: $105, up from $81 a year ago.

Gross margin: 57%, down from 74% a year ago and 64% in Q2.

Adjusted EBITDA: $82M, a 9% margin, up from $78M (13%) a year ago and $60M (8%) in Q2.

GAAP net income: -$33M, versus +$15.8M last year. EPS of -$0.14 versus +$0.06 to +$0.08 expected, depending on the source. A miss.

Free cash flow: +$26M, after -$68M in Q2.

Guidance Q4: revenue $875-895M, +42% to +45% YoY, but below the consensus of about $953M. Adjusted EBITDA guidance of $70-90M.

Guidance FY2026: $3.13-3.15B, a tighter range than the previous $3.1-3.3B; adjusted EBITDA lowered to $257-277M (from $275-325M), below the consensus of about $294M.

2030 target: confirmation of the goal of $6.5B+ revenue and $1.3B+ adjusted EBITDA

Revenue growth went from 4% in Q1 to 38% in Q2 and 49% now, and the adjusted EBITDA margin improved for the second quarter in a row. But the market looked at the subscriber drop and the weak Q4 guidance. Let’s look at some more details.

At the end of March, Hims & Hers launched a weight-loss membership in the US. You need an active membership to get your medication, and most memberships renew monthly. Part of that money comes in during the quarter but only counts as revenue later, a bit like deferred revenue at a software company. In Q3, that was $18M.

In my previous article, I estimated that the first full quarter of Eucalyptus would add $75-80M on top of the roughly $40M in June. It added $95M, so really strong.

If we strip out Eucalyptus to get a better view of the core business, we see 26% YoY growth and 6% QoQ, still strong. If you add the deferred revenue, that would have been 8% QoQ. If we look at it from the other side, international without Eucalyptus grew nearly 80%, also very strong.

US growth came in at 21%, or about 24% adjusted for the deferred revenue. Hers was about 45% of US revenue, up from about 40% a year ago. Management had a target of $1B for Hers in 2026 and with $296M this quarter, that target looks very reachable.

Even if you wouldn’t think so from the press coverage of Hims & Hers, the majority of US revenue still came from non-GLP-1 offerings. For a long time, I have been saying that this is not a GLP-1 company for me and I’ll keep repeating that until people understand. Which is probably years out, haha.

The 100K Subscriber Drop

Subscribers dropping by about 100,000 doesn’t look good, does it? But there are two reasons that make sense.

Below:

• What really happened to those 100,000 subscribers

• What the Andrew Huberman deal means for Hims & Hers.

• Peptides on Hims& Hers

• The strong AI number that tackles the biggest problem in healthcare

• The updated Quality Score, the valuation and my verdict.

Don’t want to miss this and other valuable information?

Subscriber now!

Read more

Last week, there were 11 Potential Multibaggers reporting and this week four more. That means I have to select the results I prioritize, and your questions are always a guide for me, next to negative stock price movements.

On Tuesday this week, November 10, Sea (SE) announced its Q3 results, and the market didn’t seem to know how to react.

As you can see, first the stock dropped 10%, then it shot up to +9% and then it ended the day slightly down.

If we zoom out, you can see that the stock has dropped by almost half over the last year.

The question I got most after these results was always the same: why are Shopee’s sales and marketing expenses up 60% while GMV grows just 23%? Isn’t Sea simply buying its growth? It’s a fair question and I’ll answer it later in this article.

Let’s first go over the Q3 results.

Subscribe now

Sea Limited Group Results

Revenue: $8.58B, up 43.4% YoY, beating the consensus by about 6%. Growth slowed a bit from 48.1% in Q2, but at this size, that’s still exceptional.

Gross profit: $3.94B, up 51.4% YoY. The gross margin rose from 43.4% to 45.9%. Strong again.

Operating income: $664M, up 39.5% YoY.

Net income: $502.5M, up 34% YoY, or $0.78 per share, a slight miss versus the consensus of $0.79.

Adjusted EBITDA: $931M, up 28.9% YoY.

From the earnings slide deck, through Fiscal AI (get a 15% discount here)

From the earnings slide deck, through Fiscal AI (get a 15% discount here)

Sea bought back $1B of stock and the board approved another $1B buyback program. At these prices, that makes a lot of sense.

Shopee

GMV (gross merchandise volume, the dollar value of everything sold on Shopee): $39.7B, up 23.3%YoY, or close to 26.5% without currency effects, according to management.

Gross orders: 4.5B, up 25.5% YoY.

Revenue: $6.07B, up 41.3% YoY.

Core marketplace revenue: $4.6B, up 50.6% YoY. These are the seller commissions and ads on Shopee.

Value-added services revenue: $724M, flat YoY. VAS are mostly fulfillment.

Take rate: 15.3%, up from 13.3% a year ago. If you wouldn’t remember: take rate is Shopee revenue as a percentage of GMV.

Adjusted EBITDA: $269.2M, up 44.7% YoY.

Sales and marketing: $1.43B, up 60% YoY.

This was a question I got a lot and that’s why I want to answer it right away: “Sales and marketing spending was up 60%, and GMV was only up 23.3%.” Understandably, that worries some people.

Let’s look at the recent history first.

Why does sales and marketing grow 60% when GMV grows only 23.3%? This looks bad, doesn’t it? But there’s context and I would like to point out three reasons.

To start, let’s look at it per $100 of GMV. A year ago, Shopee got $13.30 in revenue on that $100 and then spent about $2.80 on sales and marketing. Now it books $15.30 and spends about $3.60. So marketing per dollar sold rose by about 30%. That’s already another perspective. But because GMV grew as well, you get that 60%: 1.23 x 1.30 = 1.60.

The second reason is that the S&M last year was very low. In Q3 2025, sales and marketing was just 20.8% of Shopee’s revenue, the lowest level in two years. So, the comparison makes S&M jump more. The 23.6% of revenue that’s spent on S&M is about the same as in 2024. 2025 was a year of discipline but in March of this year, founder and CEO Forrest Li announced the switch:

For the near-term, we choose to prioritize growth while upholding financial discipline. For 2026, we aim to grow Shopee’s annual GMV by around 25% year-on-year, with its full year adjusted EBITDA no lower than 2025 in absolute dollars. We believe this is the right strategy to optimize Shopee’s long-term profitability.

And the third and most important reason is that this money is invested in attracting new buyers. Forrest Li this quarter:

We engaged and re-engaged a broad range of users through both online and offline campaigns, brought our content to more social and video platforms, and widened our service options, especially in delivery, to serve different user needs.

The result: the number of new monthly active buyers grew more than 50% YoY, an acceleration from 35% in Q2. If you know the e-commerce business, you already know enough now, but most people don’t, so let me explain it a bit more.

New buyers always start spending slowly. They try the platform with a small first order, maybe a second one a few weeks later. It takes time before they buy higher-priced items like electronics and household goods. But the CAC (customer acquisition cost) comes in at the beginning. Or to put it differently: the LTV (lifetime value) of the new customers can be very high but won’t show up immediately, of course. In other words, if you invest in those customers now, they won’t grow GMV as much as they push up the S&M costs.

The payoff comes over time and it’s expressed in lifetime value. In other words, 60% S&M growth versus 23% GMV growth looks bad if you take these at face value. If you know the industry, listen to management and you see the numbers (especially 50% growth in new monthly active users), management is just investing in its future and that’s exactly what I want my companies to do. If they stop that, growth will drop in a few years, and those criticizing SEA the most now for higher S&M growth will be the first to criticize them when revenue growth drops because they didn’t invest enough now.

To put it differently and summarize it at the same time: the 60% growth in S&M spending should not be seen in direct correlation with GMV growth now, but with (expected) LTV (lifetime value). Just like almost all other e-commerce companies, Shopee doesn’t share LTV numbers. The reason is that if competitors know that, they know how to outbid SEA for media impressions, as LTV is a crucial input in how much you spend on CAC (customer acquisition cost).

There was one more reason why spending more on CAC but I will come back later to that in this article. It’s a good reason.

With that out of the way, let’s look at the other things I saw this quarter.

The business grew 43%, and the stock was cut in half in a year. One of the two is wrong.

Below: the real answer to the question so many of you asked (why does marketing grow 60% when GMV grows 23%?), the IPO logic behind Monee’s spending, what Forrest Li said about Sea Cloud that changes the story from three weeks ago, the Quality Score update and my valuation showing something really surprising.

Don’t want to miss this valuable information?

Subscribe now!

Read more

Last week, on Thursday, November 5, Axon (AXON) reported its Q3 2026 earnings and the market hated them. The stock has dropped more than 30% so far.

If you zoom out, you can see that it’s down 66% from its high, the biggest drop in more than a decade. The last time it dropped so much was actually in 2014.

The stock is also at the lowest level since August 1, 2024.

If you believe in the efficient market theory (which I definitely don’t), then you must think the company is failing outright. But it doesn’t mean the market can’t be right, of course. Separating signal from noise, that’s what I try to do, and that’s what this article is meant for. Here we go!

The Numbers

• Revenue: $909M, up 28% YoY, but 3% below the consensus of $937.59M. It ends the streak of ten consecutive quarters with more than 30% growth and at least 20 quarters of beating the revenue consensus. Ouch.

• Adjusted EPS: $1.72, versus $1.24 a year ago, and 8.5% below the consensus of $1.88. Not good.

• Adjusted EBITDA: $236M, up 33%, a 25.9% margin. That’s better.

• Adjusted gross margin: 59.2%, down from 62.7% a year ago.

• Free cash flow: negative $19M, versus positive $33M a year ago.

• Net revenue retention: 126%, up from 124% a year ago. Strong.

• Future contracted bookings: $15.8B, up 39%. Again: very strong.

• Annual recurring revenue: $1.75B, up 40%.

• Software & Services revenue: $437M, up 43% YoY, an acceleration from 36% in Q2. Strong.

• Connected Devices revenue: $472M, up 16%.
  TASER came in at $233M (down 2%),
  Personal Sensors (mostly body cameras) at $114M (up 6%) and
  Platform Solutions (in-car video, fixed cameras, drones, counter-drone, VR hardware) at $125M (up 106%).

• Axon lowered its full-year revenue guidance from 32-34% growth to about 30%, or $3.61B to $3.63B. The free cash flow target went down from about $450M to $350-400M.

I summarized the results in this visual.

A 30% drop sometimes means a broken company. Is that the case here as well?

Below: the reason behind the miss, my take on the biggest investment in Axon’s history, the customer group that is growing fast.

I also update my Potential Multibaggers Quality Score and look at the valution.

The stock is down 66% in total. Is this an opportunity or a warning?

Find it out below.

Subscribe now!

Read more

What a market...

On Tuesday, Shopify (SHOP) reported its Q3 2026 earnings and the stock dropped like a rocket. It was down 20% shortly after the announcement, but it went back to -6% during the day and ended the day about 10% down, where it still is now, after it went to only -3% at a certain moment.

So, what did the numbers say? Let’s check.

The Numbers

• Revenue: $3.74B, +32% YoY

• Merchant Solutions: $2.91B, +35%.

• Subscription Solutions: $836M, +20%.

• MRR: $230M, +19%, accelerating from 16% in Q1.

• GMV (gross merchandise volume, the dollar sum of everything sold on the platform): $119.1B, +29% (+30% in constant currency).

• GPV penetration (Gross Payment Volume, penetration is how much of GMV is paid through Shopify Payments): more than 68%, up from 65% YoY.

• Gross profit: $1.78B, +28%; gross margin of 47.6%, vs 48.9% YoY

• Operating income: $568M, +66%; operating margin of 15.2%, vs 12.1% in Q3 2025

• Free cash flow: $816M, a 22% FCF margin, vs 18% a year ago.

• Buybacks: $562M, with diluted shares down to 1.291B from 1.313B a year ago.

• Guidance: Q4 revenue growth in the low 30s%, gross profit growth in the high 20s%, operating expenses at 29-30% of revenue and an FCF margin in line with Q4 2025 (19%).

I summarized the most important numbers in this image.

There was a tiny miss on revenue. Well, it depends, some gave 3.74B and then revenue was in line. But this probably caused the sell-off.

Some Facts

Here, I sum up some short facts I liked. They are not worth spending too many words on, but I like mentioning them so you see how the company is doing.

Shopify’s merchants sold almost as much in Q3 as in all of 2020, the pandemic year. The most impressive part is that it’s not even the strongest quarter, since Q4 is always the best for retailers.

Over the last three years, Shopify’s US merchants grew online sales by 23% per year, while US e-commerce as a whole grew by only 8% per year. In Europe, the number was 37% versus 8%. That shows that Shopify merchants are a very young and growing part of the economy.

Three years ago, Shopify merchants were responsible for about a fifth of all additional e-commerce sales in the US and nearly a quarter in Europe. Since the start of 2025, that’s nearly half in the US and more than 40% in Europe. Shopify’s total share of US e-commerce is now a little over 14%.

During the quarter, Shopify paid about $100 million for the domain Shop(.)com. President Harley Finkelstein explained why the company paid such a high price:

Look, in the physical retail world, location, location, location matters, we think the same thing that applies online as well and we think Shop(.)com is an incredible prime address for us.

Shop moved there from Shop(.)app, and according to Harley, traffic to Shop is now growing faster on the web than in the app.

Sometimes, people still see as only for mom-and-pop online stores, but merchants have become bigger. The group selling $2M to $25M per year added the most gross merchandise volume and at the same time, that group is growing. It shows that Shopify is fertile ground for many new brands to grow.

But while that group added most GMV, there are other strong growth factors. Offline GMV grew 30%, and 55% of the merchants that added Shopify’s point-of-sale in Q3 were new to Shopify. That’s impressive. It means that more and more physical retailers come to Shopify directly, not through an online store first. For example, MAC Cosmetics opened 84 locations on Shopify just 120 days after becoming a customer.

The advantage Shopify gives is that you give a physical store the advantage of an online store. If you shop online, the store knows who you are, your preferences, what you bought and so on. In an offline Shopify-powered store, you also get that experience.

And Shopify’s merchants keep getting bigger. Merchants with $2 million to $25 million in sales added the most GMV. Merchants above $25 million grew the fastest, but from a smaller base. That also explains why subscription revenue grew slower than GMV. The bigger merchants get, the more Shopify earns from their payments compared to their subscriptions.

All together, Shopify merchants have had an impact of $1.2T and more than 9.6 million jobs come from Shopify-powered businesses.

This was all interesting, but analysts mainly wanted to talk about one subject: AI agents.

The New Front Door

So, a tiny revenue miss probably caused the drop. If that were all, this would be a short article, haha.

Below: the new front door, how and how much Shopify earns with AI agents shopping, how small merchants (and Shopify) win in this new world, the biggest product update in Shopify’s 20-year history, what Benjamin Graham, the father of value investing, would say about Shopify’s price today, the updated Quality Score and the full valuation. Is it overvalued, fairly valued or undervalued? The answer’s below the line.

I made Shopify a Potential Multibagger at a (split-adjusted) $7.78 in 2017. Don’t want to miss opportunities like that anymore?

Subscribe now!

Read more

Tempus AI (TEM) reported its Q3 2026 results on Monday morning, before the market opened.

If you followed the stock that day, you saw a market that couldn’t make up its mind. TEM opened at $74.43, dropped to $70.30, almost 3% below Friday’s close, and then climbed to $83.42, more than 15% up. It closed at $78.48, up 8.7%. That’s a swing of almost 19% between the low and the high, in a single session. The market is always very consistent, as we know, haha.

The crazy thing is that the stock is down 12% from a year ago, but at the same time, it’s up 88% since the end of July.

I think it’s pretty safe to say that Tempus is a volatile stock, haha.

Let’s look at the numbers.

The Numbers

• Revenue: $411.5M, +23% YoY.

• Diagnostics: $303.5M, +20%, with Oncology +28% on 32% volume growth and Hereditary +3%.

• Data and Applications: $108.1M, +33%, accelerating from 28% in Q2.

• MRD volume: about 12,000 tests, +300% YoY.

• Adjusted EBITDA: $25.6M, versus $1.5M in Q3 2025.

• GAAP net loss: -$14.9M, versus -$80.0M a year ago.

• Cash from operations: +$9.2M, versus -$7.5M in Q2 2026.

• Guidance: full-year revenue of $1.60-1.605B (about 25% growth), up from $1.595-1.605B. Adjusted EBITDA $65M.

This is the overview Tempus gives in its earnings slide deck.

The Turn

I want to look a bit more into the financials and I take this quote from founder and CEO Eric Lefkofsky (my bold):

We generated $25.6 million in Adjusted EBITDA and ended the quarter with $854.4 million in cash and marketable securities. We had long suspected that with gross profit dollar growth outpacing our investments, we would eventually turn the corner and be able to generate meaningful EBITDA. This quarter, we made the turn.

Adjusted EBITDA jumped from $1.5M to $25.6M. At the same time, Tempus still had an operating loss of $56.9M on a GAAP basis. The biggest difference is SBC (stock-based compensation), which accounted for $50.8M. The rest comes from amortization from the acquisitions. Those are both non-cash costs, but of course, as a shareholder, SBC dilutes your position.

The GAAP loss of $14.9M is also misleading. It would be a bigger loss on a GAAP basis. But companies have to book stakes they hold in other companies in the GAAP numbers. In this case, $42.8M came from paper gains on Tempus’ shares in Personalis and Recursion.

That’s why I think operating cash flow is best to look at, as it shows the money generated from the company’s core operations. It came in at $9.2M, up from -$7.5M in Q2. That’s why Eric Lefkofsky’s “We made the turn” makes sense.

You can see the same in the operating leverage the company is using. Lefkofsky already referred to that as well. Non-GAAP gross profit (excluding SBC) grew by $58.6M compared to Q3 2025. Non-GAAP operating expenses grew by $35.5M. In other words, of every extra dollar of gross profit, Tempus reinvested about 60 cents and about 40 cents was not spent. That’s close to the formula CFO Jim Rogers explained on the call:

We are still relatively early days in a lot of the offerings, so we have had this perspective of reinvesting two-thirds or roughly two-thirds of any incremental gross profit dollars that we generate back in the business, and then letting a third drop down to the bottom line

Obviously with Personalis coming on board, we are already going to invest a lot in MRD. We will utilize some of that reinvestment dollars to absorb the burn from Personalis.

But then after another couple years, then that flips, right? Then you are generating meaningful gross profit dollars such that you really only need to invest a third, and you can let two-thirds drop down to the bottom line.

I like the clear plan.

So the turn is there. But there’s a lot more.

Below: the change that’s already decided and will make 2027 much better, why Tempus sells much less of its fastest-growing product than it could (for now), the shift in how doctors test that is a strong tailwind for Tempus’ future, what makes the data customers “crazy sticky” in the CEO’s own words, the one part of this company I really don’t like, the updated Quality Score, the valuation and my verdict.

Up 88% since July and still down 12% in a year: buy, hold or sell?

Don’t miss these valuable insights anymore.

Subscribe now!

Read more

Adyen published its Q3 2026 business update this morning (here in Europe), and the stock fell about 13% when I’m writing this. That means the stock is down 45% over the last year.

Very disappointing. Let’s look at what happened. Before we dive in, I have to say that management will talk in about an hour after this article is published, so that could still move the stock. But to me, most of the story is already clear.

If this article's language is somewhat less polished than usually, it's because I wanted to get it out fast.

The Numbers

• Net revenue: €738.1M, +23% YoY, +22% in constant currency

• Talon(.)One and Orb contributed about 2 percentage points to that growth

• Processed volume: €451.1B, +30%

• Digital: €387.4M, +14% in constant currency, volume +28%

• Unified Commerce: €250.5M, +27% in constant currency, volume +22%

• Platforms: €100.2M, +44% in constant currency, volume +55%

• FTEs: 5,596, including 399 from Talon(.)One and Orb

Those are actually strong numbers. Let’s look at a few things in more detail.

Organic growth was about 20%, very reliable again. Full-year guidance remained at 21% to 23%.

Volume grew 30%, very strong. But it also means that the take rate dropped again: from about 0.173% a year ago to 0.164% now. Talon(.)One only brings in subscriptions and no payments, so if you strip that out, it would have been 0.161%.

The reason is the same as always: tiered pricing. The more the volume existing customers bring, the lower the take rate. I emphasized in my recent article that this is just inherent to the business, but some continue to see this as a negative. But it’s just the consequence of big merchants bringing more and more volume to Adyen. In that sense, it’s a positive, not a negative.

Co-CEO Ingo Uytdehaage gave an interesting data point in the pre-recorded prepared remarks: 10% of Adyen’s volume now comes from the hyperscalers and cloud companies. That’s a lot of volume with very few customers. I’m very happy to hear that. We know that OpenAI is an Adyen customer, just like Microsoft, Google, and Meta. The fact that they have 10% of volume is confirming what I have said for a long time already: AI or no AI, Adyen picks up the biggest customers, while Stripe is better for startups.

Platforms keeps performing outstandingly. 46 platforms now process more than €1 billion a year, up from 37 last quarter and 31 a year ago. Active business customers grew 48% to 312,000.

Unified Commerce is also strong. 499 big customers now use Adyen both in their stores and online. Big means at least €10M on both sides and more than €50M in total. That’s up from 486 in Q2.

The Reason for the Drop

With these results, it’s hard to see why the stock dropped. And, as often, the reason isn't in the results but in what management sees for the future.

Strong numbers, and still the stock dropped 13%. The reason is not in the results. It’s in one decision management announced this morning, and I’m not happy with how it was handled.

Below: what management did, the calculation that shows how big the impact really is, what the drop means for Adyen’s valuation, my updated Valuation Score and my verdict.

Don’t want to miss valuable updates like this to know what to do with your positions?

Subscribe now!

Ps: A Dutch version of Potential Multibaggers is in the making. If you subscribe now, you can get that for free extra when it launches!

Read more

Because of my holidays and two new Potential Multibaggers picks, I still haven’t written about some earnings. Now, I want to make this clear, I have looked at all the earnings and I prioritized those I thought were most important or urgent. But writing about them takes more time than just looking at them, of course.

Today, I wanted to write about Chapters Group (MDCKF) (CHG). As a European company, it releases earnings only once a year, but it also releases business updates.

The stock has dropped 8% since the September 8 update.

But overall, it’s still up 67% since the March low.

Year-to-date, it’s up about 7%.

But I wouldn’t read too much into the stock price. This is a small company, which means the stock price will be volatile without any good fundamental reason. That’s already the case with all stocks, but even more so with smaller companies.

If you think the 8% drop in the stock price has to do with fundamentals, I understand, but it’s not true.

This is where the article stops if you are a free reader. Premium readers get much more than you, obviously.

Did you know there’s a VIP subscription too? It costs $1,200 and you get my full portfolio there. I’m extending the services you get in the VIP subscription soon and the price will go to $1,500 (or even more).

But right now, until tomorrow only (!!), you can get a free upgrade to VIP for the very last time ever for just $399.

Only if you use this link.

The offer expires on Wednesday, one minute before midnight, so don’t hesitate!

Read more

Mercado Libre (MELI) reported its Q2 2026 earnings on August 5. In other words, when I was still in Sardinia. The stock first dropped about 5%, then recovered and is now down 7.6%.

Look at the revenue growth in the last 4 quarters: 39%, 45%, 49%, 50%. That’s impressive.

It was the 30th consecutive quarter of more than 30% revenue growth.

But still the stock is down. Let’s break things down to understand.

The Numbers

• Revenue: $10.17B, the first quarter with more than $10B in revenue. It was up 50% YoY in USD, +43% FX-neutral, beating estimates by $378M or 3.9%, which is a lot at this size and with so many analysts following the company.

• EPS: $9.19, beating estimates by $0.20, but down 12% YoY

• Net income: $466M, -11% YoY, a 4.6% margin

• Income from operations: $683M, -17% YoY, a 6.7% margin, down 550 bps YoY but roughly flat versus 6.9% in Q1

• Gross margin: 40.9%, down from 45.6%.

• Adjusted free cash flow: $214M, versus $454M a year ago

Marketplace:

• GMV: $21.9B, +44% YoY, +36% FX-neutral

• Items sold: 795.4M, +45% YoY

• Unique active buyers: 89.3M, +26% YoY, almost 19 million added in a year

• Items sold per unique active buyer: 8.9, +14% YoY. Strong with so many new users.

• Same- and next-day shipments: 225M, +38% YoY

• Live listings: 859M, +60% YoY

Mercado Pago:

• Monthly active users: 88.0M, +30% YoY

• Assets under management: $23.2B, +68% YoY

• Credit portfolio: $16.4B, +75% YoY, with $15.9B originated in the quarter

• Credit card portfolio: $7.7B, +91% YoY, now 47% of the book

• TPV (total payment volume): $101.0B, +56% YoY and +56% FX-neutral

• NIMAL (net interest after losses): 20.7%, better than the 17.8% in Q1

Advertising: +73% YoY in USD, +62% FX-neutral

Revenue by country: Brazil $5.53B (+59%), Mexico $2.34B (+55%), Argentina $1.84B (+20%), Other $463M (+63%)

As always, Mercado Libre gives a clear overview of the key metrics in the slides.

The reason for the price drop is not hard to find. The market saw the same thing: margins down. The operating margin was 6.7%, down from 12.2% last year. It saw the operating income down 17% YoY while revenue grew 50%. If you don’t know the company or what it does, this would be very bad. But it’s not. It’s an intentional investment. That’s also why the margins were more or less the same as in Q1. If this were a breaking company, the margins would keep dropping.

The company gave us a nice waterfall graph again about the margins.

Let me start with the green this time. General and administrative contributed 0.5% to net margins and product development 1.2%. We also already knew this in Q1.

In the COGS, next to the lower free shipping threshold, there were also discounts for buyers who pay with PIX (Brazil’s instant payment system, essentially free for merchants) and lower take rates for sellers in certain categories and price ranges. On top of that, higher shipping costs from higher energy prices and higher POS device costs also played a role. Mercado Libre also decided to spend more to acquire new merchants in Mexico.

Mercado Libre reports under US GAAP, so it must use CECL, which stands for current expected credit losses. This means that from the moment there’s a new loan, Mercado Libre has to write down a provision for all the losses it expects that loan to generate over its entire period. That’s before a single dollar of credit income has come in. If your credit book grows by 75%, this has quite some influence on margins.

Provisions for doubtful accounts came in at $1,276M in Q2, up 85% YoY. That results in 2.4% lower margins. That sounds bad, but let’s do the math.

Credit revenue was $2,278M and provisions for losses $1,276M. The difference is $1,002M, or 9.9% of the company’s total revenue. A year ago, that was $636M, or 9.4% of revenue.

So while the credit book nearly doubled in size, it delivered more profit per dollar of company revenue than it did a year ago, not less. This is very important. The bad debt line is a growth cost that only exists on paper. Sometimes, there’s a gap between GAAP rules and reality.

Brazil, One Year Later

In June 2025, Mercado Libre lowered the free shipping threshold in Brazil from R$79 (about $15) to R$19 (about $3.50). It immediately impacted margin, of course. Many analysts called this a defensive move against Shopee, Temu and AliExpress.

While there may have been some pressure, this is also a typical Mercado Libre move. The company did similar things in the past. Then, the story was that it was a defensive move against Amazon. If it was, it worked like wonders.

And look at what happened since the introduction of the lower free shipping threshold:

• GMV in Brazil: +39%

• Items sold in Brazil: +56%

• Share of users buying 3+ categories per month: +10 percentage points

• Active sellers: +29%

• Ecosystemic user growth in Brazil: from 35% before the change to almost 50% now (ecosystemic users are users active on both the marketplace and Mercado Pago).

On top of that, the unit economics keep improving. A decade ago, exactly the same thing happened after the free shipping threshold was lowered. Half of the sales in the price range between R$19 (about $3.50) and R$79 (about $15), the new free shipping range, already recovered the extra cost of free shipping. Mercado Libre could do this through technological innovation, more scale, and filling up capacity that was not used before. In just twelve months, that’s strong.

And because it works so well, management decided to cut take rates in April. CEO Ariel Szarfsztejn explains why:

So yes, I think what we did in Brazil with take rates goes back to the basics of e-commerce, right? So when consumers are deciding where to shop, they are basically looking for the broadest selection at the best possible price with the fastest shipping and the best financing.

And clearly, getting the right selection is a key part of our strategy. That’s why we’ve been doing 1P (when MELI buys and sells products themselves, Kris). That’s why we’ve been doing CBT. (CBT= cross-border trade, Kris)

And that’s why this quarter, as we did back in 2024 and 2025, we decided to lower seller take rates. We’ve already proven that lever, and we’ve consistently seen that every time we lower take rates, we get an acceleration in effective or successful sellers in our platform.

Active sellers grew 29%. So, just like in 2024 and 2025, it worked again. It’s great to see that margins didn’t suffer extra from this move and that Mercado Libre keeps focusing on what is best for the business, not on what Wall Street wants.

Items sold in Brazil grew 45%, while FX-neutral GMV grew 36%. That means average prices came down, exactly what you expect from lower free shipping thresholds, discounts for PIX payments and much more cross-border articles. These cheaper items are exactly Shopee’s, Temu’s and AliExpress’s strength and this number shows that Mercado Libre is winning those customers as well. It’s also the reason the company can sell more items per unique active customer, despite the significant growth in the number of customers. Normally, you’d expect a slowdown.

New buyers usually buy very little in the beginning. So the users who were already there must be buying much more than the 14% extra we see.

1 + 1 = 5

Last quarter, I called Mercado Libre a STRONG BUY and I added to my position. The stock is up 23.5% since then.

Below: the brand-new number management just shared that proves that 1 + 1 can be 5, an update on the Selling Rules, the PM Quality Score, the Valuation and whether the stock is still a buy right now.

Don’t want to miss such calls anymore? Subscribe now!

Subscribe now

Read more

I hope you can keep up with the pace. After two weeks of holidays, I’m full of energy to go through the earnings I missed while being in Sardinia. The good thing about not reacting directly after the earnings is that I don’t have any emotional reaction to the stock price, so it’s even easier to separate the business and the stock price. As you will probably know, Benjamin Graham famously said:

That means that the short-term price action is sentiment, over the long term, the business will follow the fundamentals. What many seem to forget is that in investing, even the short-term can feel and be very long. Benjamin Graham himself considered anything under three years as the short term and I think that’s still true today.

Duolingo (DUOL) reported Q2 2026 on August 5, after the close.

On Thursday, the stock fell about 15% but it has recovered most of the losses. Yesterday, it was even up, but today it’s down almost 8%.

This price action already shows what I said earlier. Had I written my analysis on the evening of the earnings and the next day, I’d probably be influenced negatively by the price action, or I might have wanted to prove too much that the market was wrong. Now, I don’t have that problem.

Let’s start with the numbers.

The Numbers

• Revenue: $298.5M, +18% YoY (+17% constant currency), beating estimates by $2.9M or 1%.

• Total bookings: $289.1M, +8% YoY (+6% constant currency)

• Subscription bookings: $250.3M, +10% YoY

• GAAP gross margin: 72.6%, quite a bit better than the 71% guidance.

• Adjusted EBITDA: $77.3M, which results in a 25.9% EBITDA margin

• Net income: $33.2M, down 26% but the difference is only taxes.

• GAAP diluted EPS: $0.66, beating the consensus by $0.05

• Free cash flow: $78.6M, a 26.3% margin

And then the most important for me at this point: the users.

• DAUs (daily average users): 58.7M, +23% YoY, accelerating from 21% in Q1

• MAUs (monthly average users): 140.6M, +10% YoY, accelerating from 6% in Q1.

• Paid subscribers: 12.7M, +17% YoY

As always, Duolingo gives us a clear and visually attractive overview.

The adjusted EBITDA margin fell from 31.2% to 25.9%, but that was announced in February. Duolingo had guided for 25% and delivered 25.9%, so not bad at all, especially because management said they wanted to invest in new users and the number of users accelerated.

Important in this context is the DAU/MAU ratio, which went from 37.2% a year ago to 41.7%. Last quarter it was 41%. Some bears say that Duolingo has all the high-quality users already and now focuses on low-quality users. That argument is contradicted by the growing DAU/MAU ratio. Low-quality users would bring that number down.

This is the revenue mix.

Subscription revenue grows 22%, very much in line with the DAUs growth of 23%. For a company that said it would stop squeezing its free users, this is strong.

It also shows that you should own Duolingo only for its subscription revenue, not for the other revenue contributors.

Guidance and the Bonus Plan

These are the numbers for the Q3 guidance Duolingo gave:

• $302M revenue (+11.1%)

• $307M bookings (+8.9%)

• Adjusted EBITDA of $76M at a 25.2% margin.

Revenue guidance was slightly below consensus and may have contributed to the initial drop, but it won’t be the main reason.

For the full year, there were no changes and that’s good news. That means 10-12% bookings growth and 15-18% revenue growth.

Keeping the full-year guidance while Q3 is slightly lower than expected implies the Q4 numbers have to accelerate. CFO Gilian Munson addressed this in the conference call:

We’ve always thought that acceleration is going to come over time. This is not a slow and steady. That’s not the way this place is. But we know that it’s going to take a little time for the work we’re doing to play through, and that’s what you see in terms of us holding which... the guidance, which implies Q4 gets better from a year-over-year rate of growth perspective.

And the incentive is there. CFO Gilian Munson talked about a bonus plan that pays roughly $10M in cash if Q4 DAU growth reaches 25% or higher:

While it is not included in our 2026 guidance, we do want you to know that the company has a bonus plan that will trigger if Q4 DAU growth is 25% or higher and would be paid out during Q1. Since it’s currently uncertain whether that threshold will be met, we have not included it in our 2026 guidance. If it were achieved, we would expect the payout to be roughly $10 million in cash, potentially higher if DAU growth is higher.

My idol Charlie Munger said: “Show me the incentive and I’ll show you the outcome.” Mostly that’s seen as negative, but it’s also very positive and I’m happy with this bonus plan that can act as a carrot for management and employees to reach the goal. Management said that it would focus on DAU growth and this bonus can definitely help.

Duolingo ships a new version of its app every week and each has about 350 changes compared to the last one. Management calls it The Green Machine. The team tests hundreds of things, measures, and keeps what works. It shows up in CURR. From the shareholder letter:

The main metric we use to evaluate the stickiness of our product is Current User Retention Rate (’CURR’), which is the proportion of “current user DAUs” who come back the next day. (...)

Because small improvements in CURR compound into large DAU gains, this is one of our most important leading indicators. Thanks to the Green Machine, our CURR is at an all-time high of 84%, up by about 1% from last year.

1% may not sound like much but at such a high CURR, any improvement is impressive.

Users Came Back

In June, Duolingo ran a one-time campaign that let users who lost their streak restore their longest streak by completing three lessons. From the letter:

The response was incredible: 15.4 million learners revived their streaks, including nearly 8 million users who had no active streak when the event began. This became one of our most successful campaigns. It shows how much the streak matters to our learners, as well as how large the opportunity is to bring former learners back to Duolingo.

Nearly 8 million people who had left Duolingo came back to do three lessons with just one campaign. That’s great. Trust me, the streak is very powerful (mine is 1,661 days now).

So the users came back. But the best news of this quarter is something else.

Below: what that is, the very ambitious target management has, how the CEO admitted he was “massively allergic” to something the company now applies, and whether the stock is a buy at $132, with my Quality Score and the Valuation.

Subscribe now

Read more

Chapters Group (MDCKF) published its 2025 annual report several weeks ago, and this is the first time I get to walk you through its earnings since the stock became a pick.

As a European company, Chapters Group does not report the way American companies do. There is no quarterly call with a Q&A where eight analysts ask about the next twelve months.

Chapters Group reports a half-year and a full year, under German HGB accounting. The original report is partly in German and the Shareholder Meeting, which will be held in July, is completely in German, as that’s the rule in Germany.

I will be attending the Capital Markets Day and the Shareholder Meeting again next week, just like I did last year. Oh, and if you wonder, yes, there is simultaneous translation from German to English at the Shareholder Meeting, and the Capital Markets Day is fully in English. My German is also good enough to understand what management says on the Shareholder Meeting.

The preliminary figures were already released in March, so the annual report is the detailed version. The stock is up about 11% since I made it a Potential Multibaggers pick. Not great, but I’m in this for the long term. Maybe even more worrying for some Multis, there is no P/E, because the company reported a loss. How come? Let’s find out.

Subscribe now

Read more

The truth is that you don’t know what is going to happen tomorrow. The market will fluctuate.

Peter Lynch, One Up On Wall Street

Hi Multis

Last quarter I wrote an article called “Tough Love.“ I gave The Trade Desk (TTD) the benefit of the doubt, kept my position, and told you the next few quarters would tell us whether this is a temporary dip or a longer problem.

Well, we have one of those next quarters. And on the things that matter most to the market right now, it’s probably worse than that of Q4 2025.

The stock initially dropped about 14% in after-hours and closed Thursday’s session down roughly 8% at $23.49. We are down around 85% from the all-time high. That’s tough. For you, for me and for everyone who is not a psychopath.

Last quarter, I said “trust takes years to get and seconds to lose.” That sentence is still key. I have not lost my trust completely, but it’s wearing thin. I’m writing this article with a sharper pencil than usual and that’s because these results warrant that. But I might also have found the reason why The Trade underperforms and cannot really communicate it straightforwardly. Let’s dive in.

The Numbers

Revenue came in at $689 million, up 11.7% year over year, beating the consensus of $678.7 million by about 1.5%. A beat is a beat but the growth rate is the slowest since the IPO, with the exception of the first full quarter of the COVID pandemic.

Did you know that Fiscal now has a 25% off promotion? Use it now!

Non-GAAP EPS was $0.28, missing the consensus of $0.32 by about $0.04. GAAP net income was $40 million, or $0.08 per diluted share, down about 20% year over year. Adjusted net income was $134 million.

Adjusted EBITDA was $206 million, a 30% margin, down from the mid-30s a year ago. Q1 is always the seasonally lowest margin quarter (always remember that), so part of this is just the calendar.

Within the costs, operating expenses were up 10.8%.

Source: The Trade Desk’s Q1 2026 10-Q

I saw some pointing out that the costs for platform operations were up “unsustainably.” And yes, they grew 27% YoY (from $143M to $182M). But that’s a long-term investment in self-owned data centers and AI infrastructure. That’s also the main driver of the EBITDA margin compression we see in this quarter. All in all, I think it’s good that the company invests in its own data centers and AI infrastructure, so I don’t have a problem with the margin compression as long as it’s not outrageous, which this was not.

The real question is whether the investment will pay off in reaccelerating revenue. So far, it hasn’t, but often, these things need time.

Free cash flow was excellent: $276 million on $392 million of operating cash flow. The free cash flow margin came in at 40.1%, up 2.4 percentage points YoY. This is the best part of the report. It shows that the underlying earnings power of the business is still very much there.

The balance sheet also remains top-notch: about $1.4 billion in cash and short-term investments, no debt. The Trade Desk repurchased $164 million of its own shares in the quarter and still has $327 million remaining under its buyback authorization.

Customer retention stayed above 95% for the 12th consecutive year. I have already pointed out that it doesn’t mean all that much in that sense; it’s not DBNRR or a dollar-based net retention rate. Years ago, when asked, Jeff Green argued convincingly that DBNRR doesn’t mean that much because it is very campaign-driven. I understand this, but it doesn’t say if the average customer has spent more or not over the last year or years.

SBC remained elevated at roughly 16% of revenue, which means the buybacks are mostly offsetting dilution rather than meaningfully shrinking the share count, although the basic shares were still down about 4% YoY.

So you have a revenue beat, an EPS miss, margin compression, very strong cash generation, and continued buybacks. That’s what I would call a mixed report on the headline metrics. That alone wouldn’t have crushed the stock so much, I would argue. But guidance did.

The Trade Desk guided Q2 revenue of “at least” $750 million, against a consensus of about $770 million. That’s a 2.6% miss on the guide. More importantly, it implies year-over-year growth of just 8.1%.

Single-digit growth. From a company that grew 26% the year before. There is no nice way to put it: this is awfully weak.

The Deceleration: Five Quarters Now

I already showed this before, but let’s zoom in a bit here.

Q1 2025: 25.4% growth, Q2 2025: 18.7% growth, Q3 2025: 17.7% growth, Q4 2025: 14.3% headline (about 19% ex-political, to be fair), Q1 2026: 11.7%, Q2 2026 guidance: 8.1%.

Some decelerating growth is normal from a certain size, but this is much too fast and worse, the deceleration has accelerated. This is no longer just “tough comps” or “political spending normalization” or “CPG and auto headwinds.” Those things are real, mind you, but they are not enough for me anymore to fully explain a company going from the mid-20s growth to 8% guidance in five quarters. This is also the longest deceleration streak in The Trade Desk’s public life, not a record you want to hear about.

The Q2 guide of about 8% growth is roughly in line with global digital ad growth forecasts (Dentsu projects 6.7% for 2026), but it is well below the growth rates expected in the channels where The Trade Desk most heavily plays. The IAB projects US CTV growth of 13.8% and commerce media of 12.1% in 2026.

The Trade Desk has long pitched itself as concentrated in the fastest-growing parts of digital advertising, so the right comparison is not the broad industry average but those specific channels and against those, the gap is meaningful.

When asked about this, Green’s answer was:

We are uniquely one of the few large companies that are focused on being a buying platform for large companies. So most of our revenue comes from Fortune 500 companies and their brands. And of course, they respond differently to macro factors than smaller companies or local businesses, especially when the headwinds are macro and global in nature.

There’s truth in that. Big-brand spend is more macro-sensitive than small business or DTC spend. But it’s also, partly, an excuse. Last quarter, I wrote that excuses are never good. They might be true but they are still excuses. The very best companies prevent having to make excuses. And for a very, very long time, The Trade Desk used to be such a company.

When an analyst said that industry digital and video growth expectations are above 8% this year and asked why The Trade Desk is guiding below the industry, Green didn’t really answer directly. That too is not what I’m used to from Jeff Green. He always had a teflon layer, but he was refreshingly open to give context to criticism. Here, he pivoted to easier comps in the back half, but that doesn’t really answer the question. The question deserved a better answer than it got, and the deflection itself may be a signal that he does not have a clear explanation for why a company that has historically outperformed the market is now guiding to underperformance versus that market. This definitely worries me.

Something else that jumped out to me, but in a positive way this time, were margins. The full-year adjusted EBITDA margin guide is “at least 40%, approximately in line with 2025.” Sounds fine and unremarkable at first. And to be honest, I missed it the first time as well. That’s why if things are rocky and a company goes through a tough phase, I always listen to the conference call first and then the next day (preferably) also read the transcript. It was when I read that this jumped out to me.

Q1 came in at a 30% adjusted EBITDA margin. The Q2 guidance is $260M EBITDA on $750M revenue, so that’s about 34.7%. So, the average for H1 is 32.4%. That means H2 has to deliver an adjusted EBITDA margin above 47% to hit the full-year 40% guidance. That can mean two things: either management put the bar too high or the investments should pay off in H2. If it’s the first one, it’s another red flag and that’s why I add it to the Selling Rules. Yes, we are dealing with an interim-CEO, but I don’t care. She should do her job as well as anyone, as it impacts the company.

I want to stress that, when it comes to the Selling Rules, none are a thesis-breaker on their own; together, they paint a clearer picture of how the company executes. Sometimes, breaking one Selling Rule will be enough to sell, but there’s always context and if that context makes sense, applying the Selling Rules mechanically doesn’t make sense.

The Worrying Departure Pattern

Last quarter, I described the management churn as an “inbetweener” problem, common to companies scaling from $3 billion to $10 billion. I quoted Jeff Green on the need for different leaders at different stages in a company’s life and that definitely makes sense. I gave it the benefit of the doubt.

But the worrying thing is that this pattern continues. This quarter, Adweek broke the news, on the morning of earnings, that Samantha Jacobson, The Trade Desk’s Chief Strategy Officer, is leaving for OpenAI. Here I saw a glimpse of the Jeff Green I know, in his classy answer when a question came:

On a personal level, working with Samantha has been one of the highlights of my career. (...) She’s smart and humble, and she’s just generally amazing.

It’s also a good sign that Jacobson stays on the Board of Directors. That softens the blow. Green’s track record on hiring has historically been strong. But does he still have it?

In the last year or so, The Trade Desk has appointed a new COO, a new CRO (Anders Mortensen, from Google), a new CFO (Alex Kayyal, who left within months and was replaced by interim CFO Tahnil Davis), and now lost its Chief Strategy Officer to OpenAI. And don’t forget that in September 2023, co-founder and former CTO Dave Pickles also left the company. That is a lot of senior-level turnover, and people generally don’t leave a strong and winning company at this rate. The interim CFO label is also still on the books, which is itself a signal that the seat hasn’t been filled with conviction.

It might be deliberate upgrading, as Green phrases it, but it might also be instability. The honest answer is we don’t know yet. But I do know I want this pattern to stop soon, so the full attention can be given to the customers and serving them in the best possible way, instead of internal operations.

Is This The Explanation For The Slowdown?

Another thing that has been a drag for too long is the public dispute with one of the world’s biggest ad agency companies, Publicis. That has been going on for over a quarter now. Green’s response was deliberately short:

Since 2018, we’ve done billions of dollars of business with Publicis through the agreement that we have. And we continue to have a great dialogue with Publicis about the next chapter of our partnership. Our negotiations are ongoing. It’s probably not prudent for me to say more about it in this forum.

That’s corporate-speak for “we don’t have this resolved yet and I’m not going to give you any info.”

Of course, a dispute with one of your biggest customers is never a good situation, but I noticed something in this context worth pointing out. It could explain the situation The Trade Desk is in and it would also explain why Jeff Green is not as open as usual.

But before I explain what might be going on here, let me tell you a bit more about the dispute with Publicis. The company said it did an audit on The Trade Desk’s practices and found issues with fees, opt-ins and other things. It recommended that clients not work with The Trade Desk anymore. Omnicom, another big ad agency, also launched an audit on The Trade Desk. The ironic thing is that ad agencies are notorious for not being transparent, while The Trade Desk is generally seen as the best player in the business when it comes to transparency.

And it’s not just Publicis and Omnicom. Companies like kWPP and Dentsu have pulled back from several The Trade Desk programs, such as OpenPath. What you have to know is that this is how this business used to run.

So, you can imagine how important these ad agencies are for The Trade Desk’s revenue, right?

What blew me away in the results The Trade Desk just published is that the JBPs grew 55% YoY.

Green on the earnings call:

March was our biggest month on record for JBP signings. We signed 45 JBPs in March alone. For Q1, our total JBP count grew 55% year-over-year. And excluding renewals, new JBP deal spend grew 40% year-over-year during the quarter.

That’s incredible growth. JBP stands for Joint Business Plan. It’s a strategic partnership between The Trade Desk and mostly brands (some are with ad agencies as well). These deals include committed ad spend, shared growth targets, priority access to beta features and data, and co-marketing support. But, very important, it’s a direct relation with the brands, not through the agencies.

Cutting out the middleman and going directly to brands is, of course, very positive for The Trade Desk. But the problem is that by circumventing ad agencies, it bypasses its biggest customers. This is how it looks with JBPs.

Both images were made with ChatGPT based on schemes I had drawn on a piece of paper

I can imagine that Publicis and other ad agencies are angry about the JBPs, which largely put them out of the picture. That can make them cut spending on The Trade Desk’s platform and, using their influence with brands, discourage them from using The Trade Desk. They are being circumvented and feel their power slipping through their fingers.

That would be a positive for The Trade Desk’s direct relationship with the world's biggest brands, but it would also explain the weakness in the company’s earnings. The JBP’s 55% YoY growth should fire revenue growth on all cylinders, but if ad agencies cut their spending, that will offset the JBP growth. In the long term, though, this could prove very favorable for The Trade Desk. It could cut out the middleman and have higher margins. Is that also a part of the reason for the high EBITDA guidance for the year? I believe so.

It would also explain many of the other things I see that are not typical of The Trade Desk. I have followed this company since its IPO in 2016 and I have been an investor since May 2019, 7 years ago. If what I think is happening is true, it’s a huge, difficult and probably dangerous positioning shift for The Trade Desk.

It would explain why Jeff Green, who has never been afraid of some bold statements, is so careful right now and can’t say much. After all, if he tells the truth of what’s going on, he confirms the ad agencies’ biggest fear. It’s a process that has been going on for years, probably starting with OpenPath, which SSPs (supply-side platforms) saw as intruding on their business. Green can’t comment too much on what’s going on.

He always downplayed the insinuations of The Trade Desk going into the SSP industry with OpenPath and of course, he should do the same with ad agencies. As the ad agencies seem to choose an open war, fearing for their long-term existence, Jeff Green can’t be as open as he usually is because these are his biggest customers.

It would also at least partly explain the management turnover at The Trade Desk, maybe even starting as early as Dave Pickles in September 2023. If Jeff Green had had this as his secret master plan all the time, it would have felt extremely dangerous from the start. It’s hard to disintermediate your biggest customers by cutting them out and going to their customers directly. That’s what’s happening right now. It’s a very dangerous and risky game, but if Jeff Green can pull it off, it will be a revolution in the business.

If this has been the plan from the start, it also explains why someone like Jeff Green, who has been an incredibly visionary CEO for so long, now suddenly looks like having lost his touch. In reality, he would be pulling off the biggest disruption in the advertising business. A very risky one, I stress that again, but a potentially extremely lucrative one if this succeeds. I can imagine that many of the leaders disagree with this disruption to the business model that has worked for so long. They, too, probably think that Jeff Green lost it all of a sudden.

If this is true, there could be multiple quarters or maybe even years of underperformance during the transition, but in the end, The Trade Desk should come out much stronger, as it has eliminated a powerful middleman.

It would also explain why Jeff Green is still so incredibly insightful to listen to when it comes to the ad industry. If you want a master class on where the industry is going, Jeff Green has always been by far the best source.

The most striking moment for me was a quote he attributed to a partner at an AI-first company:

You cannot overtake 15 cars in sunny weather, but you can when it's raining.

That’s an F1 line, and it’s a great one. The argument is that the brands and platforms that get disciplined and data-driven during macro pressure are the ones that emerge with bigger market share when the weather clears.

It’s been raining for a while already, and right now, it looks like The Trade Desk is the car losing positions, not gaining them. Put differently, there’s a gap between the storytelling and the scoreboard. But if what I said above is true, that’s normal.

Reacceleration isn't really about reinventing ourselves. It's about executing against a larger and expanding opportunity.

In the context of the JBPs and what’s going on, this is interesting.

Amazon competition

Jeff Green has not changed his story on Amazon. He acknowledges it’s a strong competitor and partner but that’s not the reason for the slowdown. Of course, you can only trust his words here, as you can not see the numbers. With what I mentioned above about the JBPs, it’s totally believable that what Green says is true. He tried to convince analysts with an example.

Our pharma team recently went head-to-head against Amazon for one of the largest pharmaceutical advertisers in the world. Lured by seemingly low rates, this brand shifted some investment to PG on Amazon last year. Over the past 9 months, our team delivered consistent partnership and focused on driving real business outcomes for the client. In Q1, our team won back the business and signed a JBP for 2026 that will increase their spend on our platform by 114% year-over-year.

A 114% spend increase from a brand that briefly defected to Amazon and came back. This is a real datapoint against the “Amazon is eating The Trade Desk’s lunch” narrative. The biggest brands are voting with their dollars when given a fair test. Of course, it’s always possible that this is an outlier, but I genuinely don’t believe so. I think the “Amazon eats The Trade Desk” story is just that, a story. The company has competed successfully against Facebook, Google and Amazon, so why would that suddenly have switched so fast? Again, the JBP story explains what could be going on in reality.

So far I've shown you what's happening. The harder question is what to do about it. Below the line, I go through every Selling Rule I have made for The Trade Desk, update my Quality Score, my valuation and I conclude with a Buy - Hold -Sell rating. Want to unlock these? Subscribe here.

Become a paid or free sub

Read more

A few months ago, I wrote about Duolingo (DUOL) after a Q4 release the market really disliked, mostly because the 2026 guidance implied a deliberate slowdown. The stock fell 20% in after-hours and kept sliding.

Source: Duolingo

Since then, the stock has done more or less nothing. It bounced around in the $90-$110 range while the rest of the market was busy discovering new ways to be nervous or greedy (in semis). But in the last days, it saw quite a jump.

After Monday’s Q1 2026 release, Duolingo dropped 15% or so in after-hours but closed -5.6% on Tuesday, the same number as the full period after my previous article. Yesterday, it was suddenly up more than 8%. Long-term moves are often much easier to understand than the short-term silliness.

Of course, I hoped that there would be some relief. Mind you, I’m not getting desperate or need relief myself, but I know many of you are in stress about this position and could use some relief. Unfortunately, that’s not how the stock market works.

After going through the shareholder letter, the call, and Luis von Ahn’s recent communications, I think this quarter actually told us something important. Not the kind of thing that moves the stock the next day, but the kind of thing long-term investors should care about.

Let me walk you through the numbers first.

The Numbers

• Q1 revenue: $292M, +27% YoY (+22% constant currency), beating estimates by about $3M, but it beats the company’s own guidance by 2 full percentage points.

• EPS: $0.89 vs consensus of $0.79.

• Total bookings: $308.5M, +14% YoY (+9% constant currency) vs 11% guidance.

• Subscription bookings: $268.1M, +15% YoY.

• Subscription revenue: +31% YoY. The core engine is still humming.

• GAAP gross margin: 73.0%, up 190 bps YoY.

• Adjusted EBITDA: $83.4M at a 28.6% margin, up 140 bps and better than the 25.5% guidance the company gave.

• Free cash flow: $147.8M, a 50.6% margin. That’s impressive for a quarter the market calls a failure.

• Net income: $43.5M, +24%.

• DAUs: 56.5M, +21% YoY (lapping a 49% comp from the “Dead Duo” campaign).

• MAUs: 137.8M.

• Paid subscribers: 12.5M, +21%.

Here’s an overview from the Shareholder Letter.

Everything in this quarter was stronger than expected. So, where did the negativity come from? Bookings grew 14% reported, but only 9% on a constant-currency basis, and guidance for Q2 is just 5.8% bookings growth. That’s all the market cares about right now, because bookings lead revenue by a couple of quarters. So if you want to know what 2027 revenue growth looks like, you look at 2026 bookings. And those don’t look great.

I’ll come back to that low guidance, but first, let’s spend some time on the part of the strategy I find genuinely encouraging.

The DAU/MAU Ratio

Last quarter, I told you that the percentage of monthly users who are also daily users had reached about 40%, an all-time high and very impressive for a freemium consumer app.

This quarter, it went up again, to 41%.

Fiscal now runs a 25% off campaign. Don’t hesitate, it ends soon!

Why does this matter? Because when management deliberately changes strategy to reduce friction in the free experience, the easy bear story is: “they’re attracting lower-quality users now.” I’ve seen that take multiple times on X. But lower-quality users would mean the DAU/MAU ratio drops. The opposite happened, so you can ignore that bearish take, I think.

This was actually one of the items on my Selling Rules from last quarter. I said I expected the ratio to drop a bit (and that would be okay). It didn’t. So far, so good.

Luis von Ahn made an interesting admission on the call when an analyst asked about MAU deceleration. He said:

In terms of growth, the main thing we work on here is daily active user growth. That’s the main thing we have worked on for the longest of time, and monthly active user growth, there’s just no team that’s looking at that, and we do report it, but it’s just not a team that’s looking at that.

That’s refreshingly honest. The company doesn’t even have a team focused on MAUs. They focus on daily engagement and retention, and the MAU number is a consequence of marketing, word-of-mouth and product quality. MAU is an output, not an input, but I think almost every single analyst and investor sees this wrong. I understand that because this is different than other businesses, but it is important to be granular enough if you analyze a company.

Von Ahn went on to say top-of-funnel has been “about flat for this quarter” and that they want to accelerate it, particularly through marketing in underpenetrated regions and product improvements that drive word-of-mouth. More on that in a moment. This shows the plans of management and it should also demonstrate why you shouldn’t panic about the MAUs.

So, let’s focus on the product that should fuel growth over time. In Q1, Duolingo made speaking practice a central feature, not an add-on:

• Spoken tokens: you can now answer almost any exercise out loud instead of tapping.

• Flashcards: for fast verbal recall: saying words aloud rather than reading them.

• Speaking Adventures: real-world conversational scenarios tied directly to the curriculum.

• Video Call continues improving, and it’s expanding to Super subscribers (as promised last quarter).

The most concrete data point Luis von Ahn gave: the average words spoken per user in Video Call has more than doubled year-over-year.

Source: Shareholder letter

This is one of those details that people who don’t follow the company closely will miss. The conventional view of Duolingo is “the app where you tap matching words and the owl judges you.” Speaking has historically been the weak spot, the thing critics pointed to when arguing that AI tutors would eat Duolingo’s lunch. Well, speaking is now everywhere on Duolingo, and it’s getting better.

That brings us to content. In Q1 alone, Duolingo published 20,500 course units. That’s amazing. To put that in context:

• That’s almost 3x the rate of 2025 (~7,100/quarter).

• It’s more than 10x the rate of 2024 (~1,800/quarter).

• It’s about as much in one quarter as the company shipped in the entire year of 2025.

Luis von Ahn’s comment on the call:

And that’s in 1 quarter, and that’s about what we put out the entire year last year. And by the way, last year, we were already using AI. So we’re just getting better and better at using it.

The map of which language pairs are taught up to which proficiency level is now nearly fully covered, all the way to B2 (professional proficiency, “Duolingo Score 129”) across the nine most popular languages. Visually, you can see this in this image.

There’s a quiet point hidden in this. Duolingo has been telling investors for a couple of years that intermediate and advanced English content was a major TAM expansion opportunity but that uptake would take years because word-of-mouth is slow. Now they have the content, so that “couple of years” clock has started. That’s a 2027 and 2028 story, not one for 2026, which is exactly why it doesn’t show up in this year’s bookings. Of course, you have to trust management that this will pay off, but if I look at the past, I wouldn’t know why not.

The other thing that came out of this content build is something I find really interesting. Once you have all the content, you can start using AI to personalize what gets shown to which user. Luis von Ahn:

We may even start generating content just for you. Based on everything that we know about you, we may just be able to generate content just for you, maybe not the immediate next exercise, but like two exercises from now based on everything. We’ve just generated that sentence just for you or that piece of content.

This is the part where Duolingo’s data moat starts compounding in ways that no LLM-built clone can replicate. Anyone can use AI to generate language exercises. Only Duolingo has a billion+ user hours of data telling it which exercises to generate for which user at which moment. That’s the difference between a tool and a learning system.

There’s another quiet but important change this quarter. Duolingo added a session at the end of each unit to assess whether the learner has actually mastered the concepts. If they haven’t, they don’t move forward.

This sounds like a small change, but it shows Duolingo’s intentions. For most of Duolingo’s history, the product was optimized for engagement: keep the streak alive, keep the dopamine flowing, keep coming back. Critics could easily claim that you could maintain a 500-day streak without actually learning much. Mastery checks are the company’s answer to that. You can’t progress without proving you learned the thing. That’s the right decision, in my opinion. Any learning comes with tests.

But there’s more. Better learning outcomes mean stronger word-of-mouth, which is the cheapest and most powerful growth lever this company has ever had. Luis von Ahn said it himself: word-of-mouth is “free, but we don’t have that much control over it.” You control it by making the product genuinely effective, so users tell their friends.

One of the more striking admissions on the call for me was about performance marketing. Luis von Ahn:

We’re pretty excited about that just because we’re finally building the infrastructure to have the right attribution to send users to the right place after you acquire them, et cetera, that a company our size should have probably built years ago, but we kind of just ignored it.

A company with 56 million daily active users is just now building a proper performance marketing infrastructure. Most consumer apps do this in their first year. Duolingo got to a global scale through pure word-of-mouth and “unhinged” social media, never needing to learn how to buy users efficiently. That’s pretty impressive, even to the level of weird, maybe.

Now they’re learning it, and the most interesting place they’re learning it is China, where Luis von Ahn says they can already do profitable performance marketing. That was the first time I’d heard him speak so concretely about that. He also dropped this number that I want to highlight because few people seem to track it:

China monetizes about as well as Western Europe. So about as well as France, which is not as high as the U.S. but pretty high.

That’s impressive and, again, completely different from all other companies that have much lower average selling prices abroad. The standard bear narrative is that international DAU growth is lower-quality because international users don’t pay. China monetizing like France is a strong rebuttal. And on top of that, after the successful campaigns with Luckin Coffee and Meituan, Duolingo will soon launch a partnership with McDonald’s China. The brand is becoming culturally hot in the largest English-learning market on earth. Few people seem to be paying attention to this but it could become very important in the future. At the same time, let’s be honest: it could also mean there’s a higher risk the app will be copied.

Last quarter, I wrote about the overmonetized/undermonetized paradox. Duolingo is overmonetized and undermonetized at the same time. Undermonetized because only about 12% of monthly active users pay, while comparable freemium businesses like Spotify get to roughly 50%. Overmonetized because the way Duolingo had been pushing that conversion rate up was by adding friction to the free experience, and at some point, that friction started killing the DAU growth that makes the whole flywheel spin.

Luis von Ahn talked about this again this quarter, this time with concrete examples of what “monetization without friction” looks like in practice. The most interesting one:

I mentioned one already, which is longer free trials. This is not something we’ve experimented with a lot, the length of our free trial. But if you look at other subscription businesses that are scaled, they have pretty different free trials. Usually much longer than the one we have. You see 1 month. You sometimes even see 3 month free trials. So you’re going to see us experiment with that, and we are definitely seeing certainly the 1-month experiment we already see. We are seeing that, that both increases revenue, which is good but also it’s not at odds with daily active users.

Then he added something that I think is one of the most revealing comments of the entire call:

Experimenting with a 3-month free trial, some of that we said before. That is something that we could have never done if we didn’t have a year like this one because in a 3-month free trial, what happens is that your bookings get delayed by a whole quarter.

In other words, the only reason Duolingo can run a 3-month free trial experiment is precisely that they gave themselves the cover of a “foundational year.” If you’re operating with 25% bookings growth expectations, you can’t take a quarter of bookings and shove it forward by 90 days to see if it improves long-term LTV. You can do that with 10-12% bookings growth, and shareholders are pre-warned that this is what investment looks like.

So that’s why those slow growth projections for bookings in 2026 make sense. Maybe the company will see that it doesn’t work, but at least they will have the data to know. And if it works, it’s the foundation for growing the business for years to come.

There are also the “softer” monetization levers in the pipeline that mobile gaming has used for over two decades. Think of avatars, for example. Luis von Ahn:

Very soon, you’ll see really cool avatar costumes that’s directly coming from mobile games. I think users are going to love that. We’re doing a number of changes in terms of how we show rewards to users. I mean, for example, we’re showing them as cards now, and that feels really collectible. So we’re doing things like that.

He then pointed out that he’s currently dressed as a hot dog on the app, which is the kind of detail that makes me believe Duolingo will not lose its weirdness in pursuit of becoming a more serious product.

Buybacks

Quick capital allocation update. Through May 1, Duolingo has bought back approximately $50.6M, or about 514,000 shares. Gilian Munson framed it nicely:

This represents more than 100% of our 2025 net dilution from equity awards.

That’s worth focusing on, because the SBC argument was a fair criticism in 2024 and 2025 (and I made it myself). The buyback at current prices isn’t enough to fully neutralize 2026 dilution yet. The company is still guiding fully diluted share count to grow 3.5-4% before buyback effects, with SBC at nearly 15% of revenue. So I won’t pretend dilution is solved. But the direction of this is right, and at the current price, stock buybacks are a good choice, as the company has a very strong balance sheet.

The company ended Q1 with $1.25B in cash, $1.36 billion if you add short-term investments and no debt.

That’s a very strong balance sheet. And with the FCF the company generates, it can pay for growth initiatives, fund the $400M buyback authorization, and still end the year with about the same level of cash as they started with. That’s a pretty rare combination, especially for a company with a market cap of just $5B.

What’s Disappointing (And What’s Not)

Q2 bookings growth of 5.8% guidance (4.1% constant currency) is the kind of number that would make me nervous if I didn’t understand why. There are a few reasons. First, the Q2 2025 comp was unusually strong because of the Energy rollout, a price increase, and exceptional ad performance. CFO Gilian Munson said bookings should accelerate by about 3 points in Q3 and rise further in Q4. Fine. But “trust us, it accelerates later” is a story the market usually doesn’t like and doesn’t trust. I have faith in management, but this is a big part of the reason why the stock is down so much. I would not call it a disappointment because of the context I explained but I understand it could be perceived as such.

In-App Purchase revenue and Duolingo English Test revenue both declined YoY (-11% and -6% respectively). These are small contributors to Duolingo’s revenue, but they’re not encouraging. Worth keeping an eye on.

MAU growth is flat. Luis von Ahn admitted it. He’s working on it. But this is the metric that would tell us whether the strategy is actually working at the top of the funnel, and so far it’s not telling us anything yet. Von Ahn said he doesn’t focus on it, but he should (and will, I think) from now on.

The U.S. is growing slowly. Luis von Ahn admitted that DAU growth in the U.S. is “very low” compared to other regions. Asia is carrying the growth right now. International DAU growth is great, but the U.S. is the highest-monetization market, and it would be nice to see it reaccelerate.

What I don’t think is disappointing: gross margin guided down to 69% by Q4. The bear take here is “margins are crashing.” The right take is “they’re putting more AI into the product on purpose because users love it, and the unit economics of AI are actually improving fast enough that they keep choosing to spend on more features rather than just pocket the savings.” CFO Gilian Munson explained this dynamic well:

There are always these waves of efficiency that come with AI. So you might have AI costs come up and then the team optimizes and then you move forward… So if you look, for example, at the Q1 gross margin, it was better than we would have expected and pretty good on a year-over-year basis. And yet there’s still a lot of new AI content in our product. And that’s because on a per unit basis, the costs have come down a lot.

Q1 gross margin actually came in better than where Q4 margin will land. So the trajectory toward 69% isn’t a margin collapse — it’s a deliberate reinvestment of AI cost savings into more AI features.

The Bottom Line

Q1 2026 didn’t move the stock much and I think that may go on for a while. 2026 is a transition year for Duolingo.

But what Q1 told us, in the data, is that Duolingo is still a very strong company. Let me summarize the strengths from this quarter:

1. Subscription revenue grew 31%.

2. Gross margin expanded despite AI investment.

3. Free cash flow margin went over 50%.

4. The DAU/MAU ratio went up again.

5. Content velocity went up roughly 3x.

6. Speaking is now core, mastery checks are live, personalization is starting.

7. Buybacks have already neutralized 2025 dilution.

8. China is accelerating and monetizing like France.

9. Performance marketing infrastructure is finally being built.

10. The free trial experiments are showing they can lift revenue without hurting DAU growth.

None of this guarantees that the strategy will work, of course. There are few certainties in the stock market and especially not about the future. Two years from now we’ll know whether Duolingo took the right steps. Right now, we can only guess and that’s what makes the market so uncertain about this company. As I said in my Selling Rules last quarter, the proof is in Q3 and Q4 of this year and beyond.

Many investors say they can hold through uncertainty. But if the stock is down 80%+ and there is uncertainty (the chicken and egg problem), they usually fall into pessimism. And I understand that.

But for me, I see a management with a great track record and a credible plan for the future. There’s still a lot of potential left and the general story up to (AI will kill it) doesn’t show up up to now.

What’s for sure is that the stock is much cheaper right now. So, does that make it a buy? We will look at the Quality Score and Valuation to answer that question.

But first: the Selling Rules!

Selling Rules

It’s now more important than ever to follow up Duolingo, to know whether it’s just talking or executing. Focusing on the long term is what we want as long-term investors, but sometimes, companies use it to hide weaknesses. I don’t have that feeling for Duolingo but a feeling is not enough. Hard data matter and that’s why I made the selling rules.

Do you want to see if Duolingo is a BUY? Become a paying member of Potential Multibaggers!

(If you already are, just scroll past this)

What do you get as a member?

• ✍️ Multiple high-quality articles per week

• 📚 Full access to our entire library of deep-insight articles

• 🔎 Full investment cases

• 📊 Access to the Private Community

• 🎥 Regular Webinars

• 📈 The Best Buys Now every month

• ✍️ The Overview Of The Week each Sunday

• 🔎 Deep earnings analysis

Many Multis say that the community alone is already worth the price, but you get so much more.

Go to this page.

Subscribe to the annual plan.

Read more

Last week, Kinsale (KNSL) reported its Q4 2025 earnings.

The stock dropped as much as 15% but that loss is now reduced to 7.7%. Still quite a bit for a stock that hasn’t done great overall.

So, why did the stock drop? Gross written premiums grew only 1.8%. That scared people. But I’ll explain why that number is probably one of the strengths of this company.

First, the results.

The Numbers

Revenue: $483.27 million, up 17.3% year-over-year, beating estimates by $15.89 million.

Net Income: $138.6 million ($5.81 per share), up 27% year-over-year from $109.1 million. That’s a beat of $0.50 on the consensus estimate. For the full year, Kinsale posted record net income of $503.6 million, up 21%. You don’t need me to tell you those are phenomenal numbers, but I will tell you anyway.

Operating ROE: 26% for the full year. Best in class, as usual. For context, the average US insurer operates somewhere in the low teens or even single digits. Kinsale continues to impress here.

Book Value Per Share: Grew 33% to $84.66. Awesome result!

Float: Grew 23% to $3.1 billion. This is the money Kinsale holds between collecting premiums and paying claims. Think of it as an interest-free loan from policyholders. If your float is bigger, you can grow investment income faster. Warren Buffett built an empire on this concept at Berkshire, and Kinsale is running the same playbook, albeit at a much smaller scale and more conservatively than Buffett (more bonds and such).

Gross Written Premiums: Up just 1.8%. Ouch. Right? Uhm, right? Let’s zoom in here.

One Division Is Distorting Everything

Kinsale has about 25 underwriting divisions. One of them, Commercial Property, which handles large, catastrophe-exposed commercial property, declined 28.3% in the quarter. For the full year, it was down 17.9%.

If you exclude that division, gross written premiums grew 10.2% in Q4 and 13.3% for the full year. But Commercial Property is the largest division by premium collected.

Commercial Property is the largest division by premium collected.

Why did Commercial Property drop 28.3%? It had an extraordinary run and grew roughly 20x over the prior five years. During that period, many insurers ran from catastrophe-prone risks, and Kinsale stepped into the void, charging crisis-level premiums. Those days are over now. Competitors have noticed the margins and are coming back in. Those extremely high premiums are gone.

I have shared the insurance cycle before.

Right now, we are in a soft market because of the increased competition. This will last until a major catastrophe happens and then some undercapitalized competitors will either withdraw from that market or go out of business, which will be the reason why there will be a hard market again, which is good for strong companies like Kinsale.

Stuart Winston, the Chief Underwriting Officer, said that in September and October, it looked like Commercial Property was stabilizing. Then, in November and December, a fresh influx of competition brought premiums down further.

Is this a problem? Yes and no. Yes, because it’s Kinsale’s largest division and the premium contraction is pulling down growth. No, because this is exactly how insurance cycles work. And more importantly, Kinsale is not chasing this business at bad prices. CEO Michael Kehoe was clear. Kinsale manages every product line to a low 20s ROE or greater, and it’s not going to compromise to make the growth look better. Its discipline is its moat, and therefore, it’s good to see the company doesn’t suddenly change course.

The companies that now chase revenue growth at all costs in this competitive pricing environment are the ones that will end up with ugly surprises in two or three years.

Kehoe expects the division to stabilize “after the next couple of quarters,” but the competitive environment in large property will continue into 2026. The company’s other property divisions all grew at double digits. Think of divisions like small business property, high-value homeowners, inland marine, personal insurance, agribusiness property, and so on.

The Real Story: 71.7%

There’s something else that really popped out but on the positive side. The combined ratio of 71.7% is the best Kinsale has ever posted since going public. To quickly recap for newer readers: the combined ratio adds up an insurer’s loss ratio (how much it pays out in claims) and expense ratio (how much it costs to run the business). Below 100% means profit. The lower, the better.

71.7% means that for every $100 in premiums, Kinsale kept about $28.3 as underwriting profit. That’s before a single dollar of investment income. That’s an almost absurd level of profitability for the insurance industry.

I have already shared this earlier, but I want to show you how exceptional this is. This table was shown at the Markel Brunch in Omaha in May 2025. See how much better Kinsale already does here, with 78.7%. Now it was, and I want to repeat this, 71.7%. That’s just mind-blowing.

This is how that 71.7% combined ratio looks like.

You can see that there’s a 50.1% loss ratio and a 21.6% expense ratio.

The loss ratio improved by 220 basis points YoY. There were simply fewer catastrophes to pay.

4 basis points come from favorable prior year reserve development, up from 2.6 points last year. Now, what does that mean?

When an insurer sets aside reserves for future claims, it’s making an educated guess. If the actual claims turn out to be lower than the reserves, the difference flows back as profit.

Kinsale has overestimated its claims every single year since its founding. That tells you two things: the company’s genuinely conservative in its reserving, and it’s good at what it does.

If you look at the expense ratio, you see the company in full strength. At 20.8% for the full year, Kinsale spends about 15 percentage points less than its competitors, who operate in the mid-30s or above 40%. Just look at Markel, for example, a highly-regarded quality company.

Kinsale’s operational efficiency piece of the expense ratio (stripping out commissions) was 10.5% for the year, half a point better than 2024. So, Kinsale even gets more efficient.

Kehoe said something on the call that I think perfectly shows Kinsale’s strength:

Kinsale last year had an expense ratio under 21%, and many of our competitors tend to run in the mid-30s or higher, some even above 40%. Given the customers’ focus on low cost, it’s hard to overstate the significance and the durability of this advantage.

Because the expense ratio is about 10 or 15 full points lower than its competitors, Kinsale keeps stacking advantages. Customers care mostly about price and Kinsale has a natural advantage there because of the low expense ratio. It means Kinsale can offer lower prices than the competition and still make significantly more money. Or it can match the competition’s pricing and make a much bigger profit than the competition.

In a soft market, this advantage becomes even more powerful, because competitors with 35%+ expense ratios start getting squeezed while Kinsale is still comfortably profitable. Even more profitable than ever, as the lowest combined ratio in years proves.

The Float Machine

Net investment income grew 24.9% in Q4 to $52.3 million. New money yields are averaging around 5%, with a 4-year average duration for the fixed-income portfolio. The gross return on the investment portfolio was 4.4% for the year, consistent with 2024.

This is the part of insurance that Buffett has been talking about for decades. Kinsale’s float is now $3.1 billion. The float keeps growing and growing, and that means Kinsale has more and more revenue (and net income) that can be used to invest.

Investment income is therefore increasingly a meaningful contributor. This quarter, $52.3 million in investment income on a base of $120.6 million in underwriting income adds almost 43% to that base. And unlike underwriting income, which fluctuates with catastrophes and pricing cycles, investment income is much more stable and predictable and flows directly to the bottom line.

With premium growth moderating for now, Kinsale gives capital back to shareholders. In December, the board authorized a $250 million share buyback program. In Q4, the company bought back $50 million of stock, using the last money of the previous $100 million authorization. And Kinsale also raised the quarterly dividend by 47% to $0.25 per share. Kehoe also said it’s likely the company will deploy the $250 million authorization over the next year or so.

Kinsale’s AI

For Kinsale, AI is genuinely relevant because its entire competitive advantage is built on technological efficiency.

Over a year ago, Kinsale launched a company-wide AI initiative. Every employee now has an enterprise AI license and they have built dozens of bots and agents used daily in their business processes. Not entirely surprisingly, Kehoe pointed to IT and analytics as the areas where AI is delivering the most today. AI writes and tests code, converts unstructured data into structured data, and improves risk segmentation and pricing.

Kinsale processed about 1 million submissions last year. They quote more than 70% of all new business submissions, much more than most competitors. When you’re handling that volume, even small automation gains compound. And remember, 27% of Kinsale’s employees work in the technology department. It’s Kinsale’s differentiator and the reason I picked the company as a Potential Multibagger. The company owns its own modern custom-built core operating system, with no legacy software. Kinsale’s competitors use systems built 20, 30 or 40 years ago.

Kehoe summarized the AI opportunity in two buckets. One is even more automation, which lowers the expense ratio even further. The other is even better at segmenting and pricing risk, which further improves the loss ratio. So, it’s not a coincidence that the combined ratio dropped. Kehou called the gains “material,” even at this early stage. In other words, Kinsale’s technological advantage only expands.

Submission Trends and Pricing

New business submission growth was up 6% for the quarter (excluding unsolicited submissions). Without Commercial Property, new business submission growth was 9%. About half of Kinsale’s divisions are seeing double-digit submission growth. Renewal retention sits in the low 70% range, steady, with no growth in business moving away from the E&S market.

The combined pricing trend, per the Amwins Index (a large wholesale broker that sees massive volumes of E&S transactions), showed a rate decrease of 2.7% in Q4, compared to just 0.4% in Q3. So the pricing environment softened, again pointing out the soft market.

Within the whole, the several lines vary significantly. Large commercial property is under heavy pressure, but commercial auto, excess casualty, and general casualty still see meaningful rate increases.

Kehoe pointed out something I think many analysts missed:

The 13% growth overall ex-commercial property is pretty strong. If you look at how all the public brokers reported, growth rates tend to be kind of low to mid-single digits.

He’s right. In this competitive environment, 13% growth in the overall premiums with a 71.7% combined ratio shows how strong Kinsale is.

The Call

Unless you get to listen to The Trade Desk’s Jeff Green or Cloudflare’s Matthew Prince, the Q&A section is almost always the best part of the conference call, and this one was no exception.

About 80% of the analyst questions boiled down to: “Isn’t competition going to eat your lunch?” Analysts asked about commercial property declines, casualty pricing trends, submission flows, standard carriers moving into the E&S space, and whether the growth headwinds would persist.

There’s a clear pattern you can see if you have followed so many Kinsale calls as I have. Analysts obsess over pricing cycles and competitive threats, while management keeps calmly pointing to the same things: we have lower costs, we underwrite with discipline, and we reserve conservatively.

The analysts seem to want a more complicated answer, some sort of magic formula no one else sees. But sometimes the business really is that simple and a few fundamental differences make a huge difference in results.

An analyst from Bank of America asked whether social inflation and increased litigation were migrating into the small-account space where Kinsale operates.

Kehoe, in his typically understated tone:

Small accounts definitely aren’t immune from that.

He acknowledged that the U.S. litigation industry is “large and growing” and that plaintiff attorneys are “entrepreneurial as hell.” CFO Petrucelli’s addition, “We’re vigilant for sure,” was another classic Kinsale understatement. This is a management team that doesn’t hype and doesn’t panic. They take on problems the way they run their business: calmly but without hesitation.

A good question came on how Kinsale exercises its expense advantage: does it write at lower ROEs to grab market share, or does it keep printing 26% ROEs?

Kehoe’s answer was clear (and no suprise for me): Kinsale always targets a low 20s ROE or better on each product line, and the assumptions lean conservative. Historically, Kinsale has outperformed its own targets. Kinsale’s not going to sacrifice margin for volume. To me, that’s the only right choice. If you erode on quality, you know where you start, but not where you end.

So, you can say that the slowdown is actually demonstrating Kinsale’s moat. It won’t compromise on quality. I think you and I know that analysts would have been much more excited if premiums had grown much faster, but as a long-term investor, I couldn’t be happier with Kinsale’s decision to prioritize quality over dumb growth. And that comes from a growth investor, haha.

There were also questions about growth areas. Homeowners insurance is a long-term project. Kinsale is already in about 4-5 states for high-value homes and 15 states for manufactured homes. The expansion is slow on purpose. Kinsale looks where the opportunities are. Major insurers like State Farm and Allstate have retreated from catastrophe-prone states and therefore, insurance there is moving from standard to nonstandard (E&S) markets. Kinsale is filling the gap, but very carefully, which is in its nature.

On the question of whether AI could disrupt the broker model, Kehoe quoted Pat Ryan (founder of Aon and Ryan Specialty): “The customer needs an adviser and an advocate. I don’t think that changes with AI.”

But of course, we already saw earlier that Kinsale benefits from AI.

Conclusion

I couldn’t be happier with the earnings. I understand the initial drop, for sure. 1.8% gross premiums growth looks subpar. But insurances are a cyclical business in that respect. It’s only when the times are tougher that the diamonds shine. What I mean is that Kinsale can be above the street fighting competitors, nipping a good glass of quality wine from a distance, as it were. They know the street fighters will eventually harm each other enough to be both out of the game and then Kinsale can come in again, unharmed and still going for that high quality. At that moment, growth will return and all analysts will be fanboys again.

Kehoe has been running this company for 17 years with the same philosophy: build for cycles, not quarters and aim for a low 20s ROE or better. Reserve conservatively and let the expense advantage compound year after year. Simple but not easy under shareholder and analyst pressure. But Kehoe has been there, done that, bought the t-shirt, so to speak.

The competition has intensified but printing a 71.7% combined ratio in the most competitive quarter of the most competitive year in recent memory is the best proof that Kinsale deserves a place in my portfolio.

As Kehoe put it at the end of the call:

I think our investors should expect us to generate very high and attractive returns for the foreseeable future.

After seventeen years of backing that up with results, I see no reason to have any doubt about these words.

If you are not on the paid plan yet, this is your chance. (If you are a premium member already, just scroll past this message). What do you get as a Premium member?

• ✍️ Multiple high-quality articles per week

• 📚 Full access to our entire library of deep-insight articles

• 🔎 Full investment cases

• 📊 Access to the Private Community

• 🎥 Regular Webinars

• 📈 The Best Buys Now every month

• ✍️ The Overview Of The Week each Sunday

• 🔎 Deep earnings analysis

Many Multis say that the community alone is already worth the price, but you get so much more.

Go to this page.

Subscribe to the annual plan.

Read more

Cloudflare (NET) reported its Q4 2025 results yesterday evening (February 10, 2026), and this was an absolute monster quarter.

The numbers were terrific but what’s happening underneath the surface is at least as important. If you follow this company closely, and we have since I picked it at $39 in 2020, you can see that the company has been ready for the future at any moment. Now that AI is here, the company benefits.

But at least as important is that management delivers. In 2025, CEO Matthew Prince repeatedly promised revenue reacceleration. Well, he and his team delivered again. And the setup for 2026 looks exciting as well.

Let me walk you through the numbers first, and then we’ll get into the things that really matter.

The Numbers

The numbers were very strong, with acceleration across the board:

• Revenue: $614.5M, +33.6% YoY growth, beating the consensus by $23.1M or almost 4%, very impressive at this size.

• Non-GAAP EPS: $0.28 per diluted share, beating the consensus by $0.01.

• Non-GAAP operating income: $89.6M, which means an operating margin of 14.6%.

• Free cash flow: $99.4M, or 16% of revenue, up from $47.8M (10%) in Q4 2024. That’s more than doubling the free cash flow year-over-year.

• RPO*: $2.496B, up 48% YoY.

• Current RPO**: +34% YoY.

• DBNRR: 120%, up 1% sequentially and 9% year-over-year.

• Customers: ~332,000 paying customers, up 40% YoY. 4,298 customers paying >$100K/year, up 23% YoY.

(* RPO: Remaining Performance Obligation. The money that is already under contract but can’t be recognized as revenue yet because the services still have to be delivered

** Current RPO: the RPO of the services that will be delivered (and so recognized as revenue) in the next 12 months.)

That DBNRR bounce-back is worth looking at. If you’ve followed Cloudflare over the last two years, you know this metric dropped to 110% at its trough. Management kept repeating that it was not the best metric to look at, as there was downward pressure from the pool of funds system. Now those pools are being consumed, and DBNRR has returned to 120%. That’s a very strong for a company at this scale.

From the company’s slide deck

Cloudflare guided for revenue of $620M to $621M for Q1 2026, or about 29-30% growth. That’s above the consensus of $614M. Knowing them, they will beat this.

For the full year, management sees revenue of $2.785B to $2.795B, or 28-29% growth, versus the consensus of $2.74B. For non-GAAP EPS, guidance comes in at $1.11 to $1.12. The consensus was at $1.18, so optically, that’s a miss. But there’s context here. The company expects a 20% effective tax rate and is assuming a cash settlement of its 2026 convertible notes, which affects the share count.

Network CapEx is guided at 12-15% of revenue for 2026. That’s remarkably efficient compared to hyperscalers that invest tens of billions of dollars on Capex.

The Go-to-Market Strategy Clearly Works

Matthew Prince got quite some flak a few years ago when he said that his go-to-market team was underperforming and that they fired some people and did a reorganization. I actually like the open communication. With the reorganisation, things turned around.

New ACV (annual contract value) booked grew nearly 50% year-over-year. That’s not only a record in absolute dollars but also the fastest growth percentage rate since 2021.

Global sales productivity increased for the eighth consecutive quarter, surpassing the all-time high set back in Q4 2021. Quota attainment was the highest in four years. For the fifth straight quarter, Cloudflare added a record number of $1M+ customers.

And the deal sizes keep getting bigger. After signing a $130 million total contract value deal earlier in 2025 (the biggest in Cloudflare’s history), Cloudflare closed a $42.5 million annual contract value deal this quarter, also a new record.

Here’s a stat that really tells the story of how fast this company is scaling upmarket. CFO Thomas Seifert on the call:

For some perspective, we only had 96 $1 million-plus customers in total, just 2.5 years ago. Now that is what Cloudflare is adding in a single year and is still just getting started moving upmarket into the enterprise.

Read that again. They’re adding in one year what used to be their entire $1M+ customer base. An English word expresses this kind of phenomenon perfectly. It’s the word “compounding.”

Just look at how the different customer cohorts are scaling:

The acceleration at the top end is key here. The bigger the customer, the faster the growth. $1M+ customers grew 55% year-over-year, from 173 to 269. That’s a 47% CAGR since 2022. This is the proof that Cloudflare is successfully moving upmarket while not ignoring its self-serve public that has brought the company so much data and feedback over the years.

The pool-of-funds model is clearly working for this group. These deals essentially mean the customer is betting on Cloudflare broadly. They don’t know exactly which products they’ll use or how much, but they’re committing because they trust the platform. Matthew Prince put it well:

Every time you hear that somebody is signing up for a pool of funds deal, they’re betting on Cloudflare, they’re betting on the broad product suite that we have, and they’re betting on the ability for us as a team to continue to execute.

Pool of funds was about 20% of Q4 ACV and mid-teens for full-year 2025, said CFO Seifert. But it’s still very early and there’s a lot of runway left.

The Agentic Internet: Cloudflare’s Biggest Tailwind

This is where it gets really interesting, and where I think most analysts are still underappreciating the opportunity.

Prince laid out a vision on this call that was one of the most compelling things I’ve heard from any CEO in a while. Again, I should say. If you haven’t listened to a conference call with Matthew Prince yet, you should. While most CEOs talk in dense business language, Matthew Prince is freshly direct, open and willing to share. But again to what he said this time. In essence, it comes down to this: the Internet was built for humans. Agents are about to change everything. And Cloudflare sits right in the middle of that shift.

Think about it this way. If a human looks at 5 websites before making a purchase decision, an agent might look at 5,000. If humans follow a night-and-day rhythm, agents never sleep. The volume of internet traffic that agents generate is an order of magnitude higher than human-driven traffic. In January alone, the number of weekly requests generated by AI agents more than doubled on Cloudflare’s network.

With more than 20% of the web sitting behind Cloudflare, it is in the front row to see the agentic Internet. Prince on the call:

If AI agents are the new users of the Internet, Cloudflare is the platform they run on and the network they pass through. This creates a virtuous flywheel — more agents drive more code execution on our Workers development platform, which in turn drives more demand for Cloudflare’s performance, security and networking services. At scale, this compounds powerfully.

And it’s not just the developer platform benefiting. The traditional application services (’Act 1’) are seeing a surge because companies need to handle the massive increase in agentic traffic.

Zero Trust (Act 2) is getting more adoption because, surprise!, people are realizing that giving an AI agent access to everything in your life might not be the best idea. Prince was funny about it: “We’re sort of speedrunning all of the security challenges.”

Unlike the hyperscalers, Cloudflare didn’t blow its CapEx budget to chase this new opportunity. The company spends about 12-15% of revenue on network CapEx, which is nothing compared to the hyperscalers. Why? Because Cloudflare charges for work done, not machines rented. Prince claimed they get up to 10x the work out of the same GPU as a hyperscaler.

The hyperscalers have no incentive to make AI workloads more efficient, as that would just mean fewer leased machines and less revenue. Cloudflare does.

I love this line from Prince:

I do feel at times like we’re kind of goldilocks where we have a business model which I think is going to sustain whatever AI brings. AI agents aren’t going to replace the massive global network that we have. But we’ve also done it in a way which is much more capital efficient than some of the hyperscalers.

Don’t forget that Cloudflare is the company that was visionary about GPUs. I quote Matthew Prince from the Q2 2024 earnings call:

Starting six years ago, we intentionally left one or more PCI slots in every server we built empty.

When the moment and the technology made sense, we started deploying. That means we can use our existing server infrastructure and just add GPU cards, allowing us to add this capability while still staying within our forecasts capex envelope, and customers are excited.

The company continues to benefit from this visionary foresight.

Workers, Vibe Coding, and the Developer Flywheel

The Workers developer platform is becoming a serious growth engine. Most vibe coding platforms are either built on Cloudflare Workers or have it as their preferred deployment target. Cloudflare exited 2025 with more than 4.5 million human developers active on the platform and a lot more if you count their agents.

The examples from the call were telling.

• A Fortune 500 pharmaceutical company built an internal vibe coding platform on Cloudflare using Workers AI and Durable Objects.

• A publicly traded tech company is migrating its entire plug-in sandbox infrastructure to Cloudflare Containers.

• A leading financial services company launched an MCP server on Workers that allows AI agents to interact directly with payment services using natural language.

The acquisition strategy fits perfectly here. Cloudflare acquired Human Native (out of Google) to help figure out what the next business model of the Internet looks like. And they brought in Astro, one of the leading developer frameworks that powers a lot of the vibe coding ecosystem. Smart, targeted acquisitions that strengthen the platform.

Act 4: The Business Model of the Internet

This is the long-term bet that I think many investors are enthusiastic about, and that many others don’t understand. Cloudflare is trying to define the future business model of the Internet. What comes after ads and subscriptions?

Up to now, the internet works for free (you see ads to pay for the content) or with subscriptions (like Potential Multibaggers). If you read this for free, thank those great people who pay for Potential Multibaggers, or even better, take responsibility and take a subscription now!

Go to this page.

And subscribe to the annual plan.

But with AI, there’s a problem. AI answer engines don’t drive traffic to content creators the way search engines did. If there’s no traffic, there’s no monetization. Content creators (journalists, analysts, researchers, blog writers, photographers) need to get paid, or the content that makes AI useful will dry up.

Prince made a fascinating point about this beyond the traditional media:

We’ve actually been getting called not just from like the Associated Press and BBC and New York Times, but we’ve been getting calls increasingly from banks where their research teams are saying, we’re actually seeing fewer people subscribe to and read our research because the AI companies, the people are just turning the AI companies, they’re slurping all the data down and taking that intellectual property.

Cloudflare’s position as a neutral broker here is key. Cloudflare is not an AI company building an LLM (or other foundational model). It is not a content company. But 80% of AI companies use them, and a big percentage of the Internet sits behind them. Both sides, AI companies and content creators, prefer Cloudflare to figure this out rather than a hyperscaler with huge conflicts of interest.

Microsoft and Amazon have announced content marketplaces of their own, but Prince noted that both sides prefer a neutral third party. Media companies are already reporting that deals they’re securing with AI companies have improved since Cloudflare got involved.

Prince said 2026 is when they’ll start talking about how this impacts them financially. If Cloudflare successfully establishes itself as the infrastructure for agentic commerce, it could be as transformative as any of the previous “Acts.”

A media company already signed a $3.1 million AI Crawl Control deal and migrated massive internet properties in just two weeks using Workers.

The Margin Story: Still Early Innings

One thing I want to flag because I think it’s underappreciated: the long-term margin trajectory. Cloudflare’s (non-GAAP) operating margin went from 4% in 2022 to 14% in 2025. Free cash flow margin went from negative 4% to 12% over the same period. Those are big moves. But the long-term targets management laid out show there’s still a long way to go: 20%+ operating margins and roughly 25%+ free cash flow margins.

Sales & marketing is the biggest lever. It fell from 42% of revenue in 2022 to 36% in 2025, with a long-term target of 27-29%. You don’t get to 20%+ operating margins just by coincidence. You only get there by growing revenue faster than expenses, which is exactly what we see happening at Cloudflare.

Conclusion

This was an outstanding quarter, probably one of the best in Cloudflare’s history. 34% revenue growth at this scale, 50% new ACV growth, record deal sizes, 48% RPO growth... Cloudflare clearly is a top-of-mind platform for companies.

But what excites me most isn’t the backward-looking numbers. It’s the positioning. Cloudflare is uniquely positioned at the crossroad of the major trends reshaping the Internet: agentic AI, developer platforms, security, content monetization, and multi-cloud infrastructure. If you had to dream up a company for this time, you’d come pretty close to Cloudflare. The company doesn’t need AI to succeed, but AI is making everything they do more valuable.

Prince closed the call with something similar:

Given the opportunity we have ahead and how effectively and efficiently our team is executing on it, I wouldn’t trade places with any CEO of any other company, public or private.

When a founder-CEO with this track record says that, you should pay attention.

So, is Cloudflare a buy now or not?

Let’s find out in the rest of this article, through the Potential Multibaggers Quality Score, the Selling Rules and, of course, the valuation.

Do you want to read the rest of this article and know if Cloudflare is a BUY here or not?

Become a paying member of Potential Multibaggers!

(If you already are, just scroll past this)

What do you get as a member?

• ✍️ Multiple high-quality articles per week

• 📚 Full access to our entire library of deep-insight articles

• 🔎 Full investment cases

• 📊 Access to the Private Community

• 🎥 Regular Webinars

• 📈 The Best Buys Now every month (coming the next few days!)

• ✍️ The Overview Of The Week each Sunday

• 🔎 Deep earnings analysis article like this one

• 📊 The 10 Best Stocks for 2026 (the selection for 2025 beat the S&P by 54%!)

Many Multis say that the community alone is already worth the price, but you get so much more.

Go to this page.

Subscribe to the annual plan.

Read more

Datadog (DDOG) reported its Q4 2025 results this morning (February 10, 2026, to avoid any confusion), and this was a strong close to what was already a strong year.

Datadog earnings have become synonymous with beat-beat-raise. As a shareholder, that’s very comforting and exciting.

Again, there was acceleration in revenue growth. But while Datadog benefits from AI customers, the broad base of non-AI customers performed even stronger. That last part is really important, and that’s why we will get back to it later in this article.

Let me walk you through the numbers, the key developments, and what I think matters most here.

The Numbers

Datadog’s Q4 revenue came in at $953 million, up 29% year-over-year and above the high end of guidance. The consensus was $918 million, so this was a beat of almost $35 million or a strong 3.8%.

Non-GAAP EPS was $0.59, beating the consensus of $0.55 by $0.04. Non-GAAP operating income was $230 million, good for a 24% operating margin. Free cash flow was $291 million with a 31% margin. Healthy numbers.

For the full year 2025, Datadog delivered $3.47 billion in revenue, making the full-year guidance raised from the previous quarters look conservative in hindsight. But that’s what Datadog does. They guide conservatively, then they beat. Every single quarter this year. And the previous years. Just look at this immaculate track record. It’s a beauty.

Source: Seeking Alpha

Datadog ended Q4 with about 32,700 total customers, up from about 30,000 a year ago. There were 4,310 customers with an ARR of $100,000 or more, a 19% increase from 3,610 a year ago. These large customers generated about 90% of ARR.

Datadog now has 603 customers with ARR of $1 million or more, up 31% from 462 a year ago.

If you want these incredibly insightful charts from Fiscal, use my link to get 15% off.

That’s a significant acceleration from the last two years, as you can see. When your biggest customers are growing fastest, that tells you something about the stickiness and the expanding value of the platform.

The dollar-based net retention rate held steady at about 120%, and gross revenue retention remained in the mid-to-high 90s. Churn is low and that indicates that Datadog’s platform is mission-critical.

The land-and-expand model continues to work exactly as you want to see it as an investor. By Q4-end, 84% of customers used 2 or more products (up from 83% a year ago), 55% used 4 or more (up from 50%), 33% used 6 or more (up from 26%), 18% used 8 or more (up from 12%), and 9% used 10 or more products (up from 6%). Look at these beautiful climbing lines.

The last metric is new: 9% of customers use 10+ products. When you think about the switching costs that come with that level of integration, you start to understand why the retention rate is so high. It’s not practical to rip all of that out and manage those integrations separately. Or, to put it differently, you can’t go back to point solutions when you have a unified platform.

48% of the Fortune 500 are now Datadog customers, but the median ARR for those Fortune 500 customers is still less than $0.5 million. There is a massive runway just from growing within the existing base.

The Core Business Is Accelerating

Here’s what I think many investors might be underappreciating. CEO Olivier Pomel on the call:

We saw a continued acceleration of our revenue growth. This acceleration was driven in large part by the inflection of our broad-based business outside of the AI-native group of customers we discussed in the past.

So, revenue growth from the non-AI-native customer base accelerated to 23% year-over-year in Q4, up from 20% in Q3 and 18% in Q2. That’s three consecutive quarters of acceleration in the core, broad-based business. This is cloud migration. This is digital transformation. This is the secular trend that was always the backbone of the Datadog thesis, and it’s getting stronger, not weaker.

It’s funny, in these times of what I call the Anti-Non-AI Bubble, to see a company benefiting from AI actually grow most from what’s now seen as ‘the old’: cloud migration and a bigger digital stack. It shows that what the market emphasizes and reality are not the same.

CFO David Obstler confirmed that this acceleration trend has continued into January. Noice!

Meanwhile, Datadog also fully profits from the new paradigm. The AI-native cohort continues to grow rapidly. There are now about 650 customers in the AI-native group (up from about 500 last quarter), with 19 customers spending $1 million or more annually with Datadog. And 14 of the top 20 AI-native companies are Datadog customers.

Datadog shows single-handedly what I’ve been shouting in the wind: Yes, AI is great, but don’t exaggerate the speed of innovation. That always goes slower than you think. Even now, many companies are still switching from on-prem to the cloud. That’s the previous paradigm, yes, started about 15 years ago. That’s how it goes in the real world.

Record Bookings And The Anthropic Deal

And more positive news. Datadog posted a record $1.63 billion in bookings, up 37% year-over-year. Another acceleration.

Datadog signed 18 deals over $10 million in TCV in Q4, of which 2 were over $100 million, and 1 was an 8-figure land with “a leading AI foundational model company.” This is Anthropic. That AI model deal is worth highlighting: it was Datadog’s biggest new logo deal ever.

Anthropic was known for using a mix of open-source tools (like Prometheus/Grafana), in-house solutions, and cloud-native tools across AWS and Google Cloud. The move to consolidate “more than 5 tools” into one platform is a classic Datadog “consolidation” play used to win over Anthropic’s complex, multi-cloud environment. Anthropic consolidated everything onto Datadog.

When an analyst asked whether the “you can do it on the cheap with open source“ narrative holds up, founder and CEO Olivier Pomel was direct with his answer:

Your engineers typically are very well compensated and are the biggest part of the spend in this company. Their velocity is what gates just about everything else in the business. And so usually, when we come in, when customers start engaging with us, we can very quickly show value that way.

In other words, the “build it yourself” argument collapses when you factor in the real cost of engineering time. Datadog has made this case for years, and these massive deals prove it.

The ARR Milestones Tell A Powerful Story

Several major product milestones were disclosed this quarter and are worth paying attention to.

• Infrastructure monitoring now contributes over $1.6 billion in ARR.

• Log Management crossed $1 billion in ARR, with Flex Logs (only launched in August 2023) nearing $100 million in ARR on its own.

• APM plus Digital Experience Monitoring also crossed the $1 billion ARR mark.

That’s three Datadog pillars at or above a billion, and Datadog noted that about half of its customers still don’t purchase all three pillars. So the cross-sell opportunity remains enormous.

APM specifically is now growing in the mid-30% range year over year, which makes it Datadog’s fastest-growing core pillar. That’s a reacceleration, driven by easier onboarding, the strength of digital experience monitoring, and better go-to-market coverage.

Security ARR growth continues to accelerate too. Datadog is actively displacing established market leaders with Cloud SIEM in large enterprises. The tight integration with Bits AI Security Agent is becoming a strong differentiator.

AI: Both Sides Of The Coin

Datadog frames its AI opportunity in two buckets, and I think it’s a useful way to think about it.

First, AI for Datadog: making the platform smarter.

Bits AI SRE Agent launched for general availability in December, and over 2,000 trial and paying customers have already run investigations in the past month. The Datadog MCP server is in preview for thousands of customers, with tool calls growing 11-fold in Q4 compared to Q3. That’s explosive growth, exactly what we want to see as Multis.

Second, Datadog for AI: observability and security for AI workloads.

More than 1,000 customers use LLM Observability, and the number of spans has increased 10x over the past 6 months. Datadog is also building GPU monitoring, an AI agent console, and security products to protect against prompt injection attacks, model hijacking and data poisoning. With such an abundant use of LLMs, this is crucial and much appreciated.

On the topic of whether general-purpose LLMs could threaten Datadog’s moat, Pomel gave a thoughtful answer.

First, he acknowledged that LLMs will get much better at analyzing broad datasets. But Datadog’s advantage is twofold: the ability to assemble and structure the context (data aggregation, parsing, understanding how systems fit together), and the ability to run analysis in-stream, in real time, at massive scale.

Pomel:

Where the world is going is you’re going to have many more changes, many more things. You cannot actually afford to have incidents to look at for everything that’s happening in your system.

So you need to be proactive. You need to run analysis in stream as all the data flows through. And for that, you need to be embedded into the data plane, which is what we run.

That’s the vision. Not post-hoc analysis fed into an LLM, but real-time, in-stream detection and resolution. That’s a fundamentally different problem, and it’s one Datadog is uniquely positioned to solve. LLMs are phenomenal at solving problems of the past, Datadog will do it in real-time and preventive.

Guidance: The Serial Sandbagger Strikes Again

For Q1 2026, Datadog guided revenue between $951 million and $961 million. At the midpoint, that’s 25-26% year-over-year growth. The consensus was $934 million, so even the low end of guidance beats the estimates. Another raise, in other words, like clockwork.

For the full year 2026, the guide is $4.06 billion to $4.10 billion, or 18-20% year-over-year growth. The consensus was at $4.10 billion. Full-year non-GAAP EPS guidance of $2.08 to $2.16 even came in below the consensus of $2.34.

So, why did the stock jump so much today?

If you are not familiar with Datadog, you will look at that 18-20% full-year guidance and think it signals a deceleration.

The market starts to get it, though. At this time last year, Datadog guided to 19% year-over-year revenue growth for 2025. How did that end up? 28% year-over-year after Q4 results. Now they’re guiding for 19% growth again for 2026. Two years ago, it guided for 19-20% growth. Reality 26% growth.

So... Do you really believe Datadog will really only grow 18% to 20% in 2026?

Datadog is a serial sandbagger when it comes to guidance. They’ve beaten and raised every single quarter in 2025. And I like that: underpromising and overdelivering is a real Potential Multibaggers criterion.

The guidance explicitly models that the broad-based business (excluding the largest customer) grows at least 20%, with conservative assumptions on Anthropic. CFO David Obstler noted on the call that this doesn’t mean the large customer is declining; it’s just that in a consumption model, they can’t control it and so they apply extra conservatism there. You bet, David!

Full-year operating margin guidance is 21%, down from 24% this quarter, reflecting continued heavy investment in R&D and go-to-market. But as we’ve seen in previous years, the beat-and-raise pattern tends to be in the margins as well. So, again, no worries. And this time, it seems like the market finally catches up with this year-long pattern.

Conclusion Earnings Insights

This was one of Datadog’s best quarters ever. Let me recapulate why:

• Accelerating revenue growth in both the core business and the AI cohort.

• Record bookings.

• The biggest new logo deal in company history.

• Three product pillars at or above $1 billion in ARR.

• A 31% free cash flow margin.

Datadog shows more and more that it’s a platform that is fully embedded in companies’ tech stacks and difficult to replace. To the contrary, companies consolidate point solutions on Datadog’s platform, as Anthropic showcased very clearly.

The guidance looks ultraconservative, but that’s nothing new. Important: the product innovation shows no signs of slowing. More than 400 new features were shipped in 2025, and Pomel said he’s even more excited about 2026.

If you zoom out, the thesis is simple: the world is moving to the cloud, building more software, deploying more AI, and all of that complexity needs to be observed, secured and managed. Datadog is the perfect platform for all of these developments.

Now, does this make Datadog a buy? Let’s find out with the Potential Multibaggers Quality Price, the Selling Rules (new for Datadog!) and the valuation.

Do you want to read the rest of this article and know if Datadog is a BUY here or not?

Become a paying member of Potential Multibaggers!

(If you already are, just scroll past this)

What do you get as a member?

• ✍️ Multiple high-quality articles per week

• 📚 Full access to our entire library of deep-insight articles

• 🔎 Full investment cases

• 📊 Access to the Private Community

• 🎥 Regular Webinars

• 📈 The Best Buys Now every month (coming this week!)

• ✍️ The Overview Of The Week each Sunday

• 🔎 Deep earnings analysis article like this one

• 📊 The 10 Best Stocks for 2026 (the selection for 2025 beat the S&P by 54%!)

Many Multis say that the community alone is already worth the price, but you get so much more.

Go to this page.

Subscribe to the annual plan.

Read more

I asked David to write an article in which he looks back on the Q3 results of the Potential Multibaggers stocks and he peers into the future with a Q4 preview. In the upcoming week, the first Potential Multibaggers stocks start reporting and this article looks at those reporting companies. What can we expect and what will we focus on?

I ended up contributing more than I expected to this article, adding (and deleting) quite some content and making significant edits. So, you can see this as a collaborative effort and that’s why I added who did what. I give the word to David.

Hi Multis

David here.

In this preview, I’ll provide a brief overview of our Q3 analysis, review stock performance from Q3 to today, and share our preliminary expectations for Q4.

Kris has already written extensively about Potential Multibagger picks over the past few weeks as markets have remained highly volatile and many stocks have pulled back meaningfully.

From my perspective, there is no reason to panic. Markets will always fluctuate; it’s one of those sayings where history doesn’t repeat itself, but it often rhymes. The same can be said about the market.

Software-related concerns on Wall Street have pushed markets broadly lower, particularly across the technology sector. The concern is that AI’s impact may be more widespread than initially expected. There is little doubt that AI will have a profound impact on both our daily lives and the companies around us.

I view this impact largely as a positive, and as Kris has noted in several recent articles, some of these concerns appear overstated. Fundamentals across the AI landscape remain strong, but we are still early in the cycle and continue to need substantial investment in hardware, tools and agents to support growth.

From experience, it takes time before tools become truly reliable and even then, issues often persist. While modern tools make software development easier, they primarily help startups reach an initial Minimum Viable Product (’MVP’) more quickly. However, MVPs are exactly that, minimal. They still require extensive refinement, security assessments, and debugging.

I spoke with several specialists, who warn about this trend: when non-technical professionals build tools, the results are often faster to market but riddled with bugs and security vulnerabilities. This can be useful for demonstrating an idea’s potential, but little more than that.

Beyond that, these tools can certainly assist experienced developers but the broader implications matter more. Most Potential Multibaggers are IT-heavy companies with deep expertise and strong engineering teams. I prefer to see these companies leverage new tools to enhance their product offerings, which they are already doing, and generate incremental revenue as beneficiaries of the AI wave.

When I began writing this article a few days ago, several stocks were already down meaningfully, including names with strong buy ratings. Recent pullbacks have only made these opportunities more attractive, in my view.

Do you want to see my stocks?

Do you want to know what I buy for my real-money portfolio?

Become a paying member of Potential Multibaggers!

(If you already are, just scroll past this)

What do you get as a member?

• ✍️ Multiple high-quality articles per week

• 📚 Full access to our entire library of deep-insight articles

• 🔎 Full investment cases

• 📊 Access to the Private Community

• 🎥 Regular Webinars

• 📈 The Best Buys Now every month

• ✍️ The Overview Of The Week each Sunday

• 🔎 Deep earnings analysis

• 📊 The 10 Best Stocks for 2026 (the selection for 2025 beats the S&P by more than 40%!)

Many Multis say that the community alone is already worth the price, but you get so much more.

Go to this page.

Subscribe to the annual plan.

Read more

You probably remember I picked CrowdStrike (CRWD) as a Potential Multibagger at $98.10. It's done well for us, but this isn’t a set-it-and-forget-it game. It’s buy-and-verify, always. So let’s dig into the latest results with Anand and see if the thesis still holds. Take it away, Anand!

The company reported the result on Wednesday, August 27, and the stock was down around 6% after the release. The next day, the stock was up around 5%.

The Numbers

CrowdStrike's management team informed us about the reacceleration in the last quarter, and the numbers for this quarter confirmed the reacceleration.

CEO George Kurtz:

I'm proud of CrowdStrike's ability to deliver reacceleration, our return to year-over-year net new ARR growth a quarter early.

Revenue for Q2 came in at $1.17 billion, up 21% year-over-year, slightly ahead of expectations. More importantly, Annual Recurring Revenue or ARR grew 20% to $4.66 billion, with $221.1 million in net new ARR added during the quarter.

Source: Fiscal.ai (get 15% off with this link)

Subscribe now

To put that in perspective, CrowdStrike's ARR has skyrocketed from $791 million to $4.66 billion in just four years, resulting in a 42.53% compound annual growth rate. That level of sustained growth at scale is truly remarkable.

Source: Fiscal.ai

In Q1, CrowdStrike customers who adopted 6 or more modules grew to 48% from 45% a year ago, 7 or more modules grew to 33% from 29% a year ago, and 8 or more modules grew to 23%. Notably, among customers with over $100,000 in ending ARR, 60% have adopted eight or more modules, highlighting the effectiveness of the CrowdStrike platform consolidation strategy.

Source: Fiscal.ai

The total GAAP gross margin was 77%, and the non-GAAP subscription gross margin remained at 80% of revenue.

On a non-GAAP basis, CrowdStrike reported net income of $237.4 million, compared to $221.6 million in the same period last year. Non-GAAP EPS came in at $0.93, compared to $0.88 in Q2 last year. Both figures came in ahead of internal estimates for the quarter.

GAAP results, however, were more heavily impacted by the compensation from the outage last year. The company reported a GAAP net loss of $77.2 million, which included $35.7 million in expenses for outage and related matters and $38.4 million in strategic plan-related charges, compared to a net income of $47 million in the same quarter last year. GAAP net loss per share was $0.31, versus a profit of $0.19 per share in Q2 last year.

Free cash flow was $283.6 million or 24% of revenue, compared to $272.2 million. Expenses for outage-related and strategic plan costs impacted Q2 free cash flow by approximately $29 million. Free cash flow began to tick up as the outage-related expenses started to fade away slowly. The Free cash flow margin was 25.9, and the management team guided to exit the year with 27% of free cash flow margin.

Source: Fiscal.ai

CrowdStrike reported that cash and cash equivalents grew to a record $4.97 billion. Debt remains more or less the same, at $810.5 million.

Source: Fiscal.ai

CEO George Kurtz on the call:

Quarters like this one highlight our momentum and progress on the path to $10 billion in ending ARR, setting new records, achieving net new ARR reacceleration sooner than anticipated, and rising competitive win rates highlight CrowdStrike leading the way in cybersecurity.

Guidance:

Q2 2025 Outlook

Management is known for providing conservative guidance. They guided for revenue between $1.208 to $1.218 billion, a 20% growth compared to a year ago, which is in line with expectations. Here is an impressive number that the company expects at least 40% year-over-year net new ARR growth for the second half of 2025. Non-GAAP net income is in the $256 to $262 million range, and non-GAAP net income per share will be in the $0.93 to $0.95 range, also in line with the consensus estimate.

Full-year 2025 outlook

Management guided for revenue between $4.749 billion and $4.805 billion. Non-GAAP net income is expected to be between $1000.1 $1040.1 million, and non-GAAP net income per share will be in the $3.60 to $3.72 range. The revenue expectations for the year remain the same, while net income and net income per share full-year guidance increased from the previous quarter.

CrowdStrike annual conference Fal.Con 2025 begins on Monday, September 15. With over 100 sponsors and over 8,000 attendees, Fal.Con is going to be the largest customer event yet. That will help improve the second half of the year and the following year's pipeline.

CFO Burt Podbere on the call:

We have a clear line of sight to well exceed the $5 billion ending ARR milestone by fiscal year-end, achieving the ambitious goal we set in 2022 as we execute on our path to $10 billion in ending ARR by FY 31.

In less than 3 years, the company already has confidence of $5 billion in ARR, which will mark the halfway point from the long-term goal of $10 billion ARR. That is impressive.

Subscribe now

Highlights from the conference call

CrowdStrike Platform Updates

CrowdStrike's cloud, Next-Gen Identity, and Next-Gen SIEM platform solutions are now more than $1.56 billion in ending ARR, growing more than 40% year-over-year. That's remarkable progress and strengthening CrowdStrike's lead in the cybersecurity space.

There are also more specialized solutions. The Exposure Management business, which includes capabilities such as vulnerability management and attack surface management, has surpassed $300 million in ending ARR and was named a leader in the 2025 IDC Worldwide Exposure Management MarketScape. CrowdStrike's market leadership was further reflected in Gartner's latest Magic Quadrant for Endpoint Protection Platforms, where CrowdStrike was placed in the Leader box for the sixth consecutive year.

Source: CrowdStrike investor presentation

Falcon Next-Gen SIEM had a stellar Q2 with year-over-year growth of more than 95% and an ending ARR of more than $430 million. Next-Gen SIEM is becoming synonymous with AI SOC transformation.

CrowdStrike recently announced the launch of Next-Gen Identity Protection, which extends the best-in-class identity protection to nonhuman identities or NHI, SaaS applications, and most importantly, AI agents. As agentic identities proliferate, customers require an identity security solution to safely leverage agentic AI, preventing exploitation, misuse, and breaches. Including Falcon Shield and Next-Gen Identity Protection business exceeded $435 million of ending ARR in Q2, growing more than 21% year-over-year.

Based on customer demand and another opportunity to innovate, CrowdStrike launched its own Privileged Access Management or PAM offering in Q1. CrowdStrike PAM solution delivers privileged account password rotation, privileged user identification and risk assessment, privileged escalation detection, user risk profile insights, and flexible MFA controls for different departments within the organization, helping its customers consolidate with confidence. Driven by the excitement for Next-Gen Identity Protection and Next-Gen Privilege Access, these new solutions significantly expand their identity solution opportunity.

You probably heard the news that Palo Alto Networks (PANW) announced the acquisition of CyberArk for $25 billion. That's a huge acquisition and clearly shows Palo Alto Networks' formal entry into Identity Security, establishing it as part of the company's multi-platform strategy. It combines CyberArk's long-standing leadership in Identity Security and Privileged Access Management (PAM).

This is where two schools of thought emerge: small acquisitions with unified platform offerings vs. big, bold acquisitions with a multi-platform strategy. There’s no right or wrong, but I like the former more.

CrowdStrike is a leader in cloud runtime protection, with the largest and most sophisticated enterprises trusting it to protect their most critical production environments. With the need to secure AI as a backdrop, CrowdStrike delivered impressive net new ARR in the cloud this past quarter. Total cloud ending ARR exceeded $700 million, growing more than 35% year-over-year.

Subscribe now

Falcon Flex for the win

Falcon Flex is significantly evolving the go-to-market and customer experience. CrowdStrike surpassed the 1,000 Falcon Flex customer milestone, with the average Flex customer representing more than $1 million of ending ARR. CrowdStrike added more than 220 new Flex customers.

CrowdStrike and its partner channel see strong adoption of the unified Falcon platform. Utilization of Flex contracts is more than 75% across the Flex customer base. Building on last quarter's re-Flex momentum, now more than 100 customers have already re-Flexed. CrowdStrike had more than doubled the number of re-Flexed accounts to nearly 10% of all Flex customers.

Cybersecurity is more non-negotiable than ever

As organizations of all sizes embrace AI transformation, there are some worries and thematic concerns from executives and Boards. Worries like where shadow AI is emerging in businesses, how to control what data enters AI systems, and how to control what AI systems can do in the enterprise? That leads to the overall basic question: how to secure AI agents?

AI has made the role of Chief Information and Security Officer, or CISOs, and Chief Information Officer, or CIOs, more complicated than ever. Answering the above questions is difficult, expensive, nuanced, conditional, and incomplete. It is essential to adopt AI in a safe, secure, and meaningful way.

CrowdStrike's threat intelligence research uncovered Chollima, a North Korean nexus group, using GenAI to infiltrate more than 320 enterprises by automating fabricated resumes and conducting deepfake interviews. The threat is real. CrowdStrike's role in the agentic era is staying ahead of AI-armed threat actors to secure AI at every layer. If we believe in AI, and we think there's going to be meaningful AI in the future, in the next year, 3 or 5 years, then security is non-negotiable.

CrowdStrike's superhuman Charlotte

CrowdStrike is rapidly accelerating the pace of its AI innovation. Charlotte is its SOC analyst, automating actions and now end-to-end autonomous workflows across the SOC. Charlotte had a record quarter, growing more than 85% over Q1. The company is integrating Charlotte across the entirety of the Falcon platform, empowering customers to achieve their agentic security goals out of the box with immediate Return on Investment or ROI.

CrowdStrike has a unique competitive advantage of vast, clean, reliable cyber data, coupled with the data science expertise, which creates a reinforcement learning flywheel, continuously adapting and improving autonomous detection and response.

Charlotte is constantly learning and improving as they train it on its market-leading threat intelligence, battleground incident response and scaled Falcon Complete MDR analyst behavior. The outcome is Charlotte turning its data moat into a fortified and dynamic AI wall. The time and effort of training the model will be hard to compete in the AI era. CrowdStrike made an early investment and is creating meaningful solutions for its customers.

There’s a lot more under the hood this quarter including a stealth acquisition that could reshape the platform, a quiet turf war brewing with Datadog, CrowdStrike’s partnership with Nvidia and a GTM shift that could boost margins. Let’s break it down.

If you're a paying subscriber, keep reading. If not, now’s a great time to upgrade and get the full story.

Upgrade now

Small Tuck-in Acquisition

CrowdStrike announced its acquisition of Onum, a data pipeline platform company. Built on a proprietary stateless in-memory architecture, the Management team believes Onum is the perfect complement to Next-Gen SIEM. It offers unparalleled speed, scale, and efficiency in onboarding to Next-Gen SIEM while giving customers control of their data. Onum delivers 5x more events per second than its nearest competitor and processes data in real-time versus legacy batch and store methods.

Onum will bring Falcon's AI-powered detections closer to third-party data sources in the pipeline, starting analysis before data even enters the Falcon platform. Onum is delivering up to 70% faster incident response with 40% less ingestion overhead. That means less cost for the customers since all these solutions are cost-based on the amount of data ingestion.

I love small tuck-in acquisitions. The financials about the acquisition are not disclosed publicly, but Onum was just a series A company with $40 million in total funding, so it won't be crazy high numbers. The Management team is thoughtful, has clear intentions, the ability to integrate into the existing platform and expand the product offering, and has no overhead for the go-to-market team. CrowdStrike has a proven track record of making tuck-in acquisitions and expanding the product offerings. I want them to continue doing it. I do not expect any major acquisitions.

With this acquisition, there is some overlap between CrowdStrike's offerings and Datadog's offerings. CrowdStrike is expanding in observability (especially data pipeline observability), and Datadog is expanding in security with Cloud Security, Cloud SIEM, Sensitive Data Scanner, and Workload Protection offerings. It's nothing to worry about; it's a big enough space to have more than one player, and luckily, we have both companies in our portfolio.

If you are thinking about merging both, then my immediate answer is NO, and I am not expecting it in the near term. I could be wrong, and if it becomes one, then it will be a powerhouse that no one will be close to. Indeed, I would like both companies to excel individually.

Update on Key Partnership with Nvidia

CrowdStrike announced a partnership with industry stalwart NVIDIA last quarter. Nvidia software stack announced integration of Falcon Cloud Security, with NVIDIA universal LLM NIM microservices and NeMo Safety. NVIDIA customers now benefit from full AI life cycle protection for over 100,000 LLMs through Falcon.

Final thoughts

CrowdStrike reported better-than-expected numbers. They raised guidance, too. The company faces a short-term headwind due to outage-related expenses until Q4 2025, but the management team handled the situation extremely well. The Falcon platform is expanding and innovating, and shows great strength combined with the Falcon Flex model.

Charlotte and the importance of cybersecurity in the AI era make CrowdStrike even more important than ever.

Disclosure: I personally own the shares of CrowdStrike.

Kris will update the Potential Multibaggers Quality Score and say whether CrowdStrike is a buy later this week (probably tomorrow).

In the meantime, keep growing!

Karan here.

Confession time: I’ve been working on this article for 3 weeks. Why did it take so long?

The first version was a comprehensive review of the company’s earnings and how well they did, considering their challenging environment. I probably would have published it, but when I fall asleep proofreading my own article, it’s probably not fun for the rest of you to read it.

The second version was focused on the state of the insurance industry, I found great company with the analysts on the earnings call - but it turned out that the company was navigating most of those pressures just fine. Also, it was still pretty boring.

To a certain extent, this is the nature of the company’s business – insurance is hard to understand and boring most of the time – but writing about it for a generalist audience can be even worse. Where do we draw the line between detail and noise?

So, this is the third attempt, written more directly and simply, and I hope, gets the job done. It’s also the shortest, because I’m trying not to bore our loyal readers.

A quick recap, Kinsale Capital Group (KNSL), Inc. operates as a specialty insurance company focusing on the excess and surplus lines ('E&S') market in the United States. E&S is a catchall term for “hard-to-price risks”.

And so, in other words, the company provides customized insurance solutions for complex and hard-to-place risks. This includes stuff like property, casualty, and specialty risk insurance products, but targeted mostly at small to mid-sized businesses (which they defined as having gross revenues between $1 million to $50 million annually).

These businesses are generally high-risk commercial enterprises and emerging market segments, including transportation, construction, environmental services, and hospitality.

This is how the customers are split up between different industries.

Source: Kinsale's Q4 2024 Earnings Call Slides

This works out to about 15,000 commercial insurance clients, sold mainly through a network of independent brokers.

You and I likely can’t walk up to Kinsale to buy life insurance (yet, see further), which is why most never have, and never will, hear about the company. That's also why it is niche and kind of boring to read about. It’s still barely followed despite crushing the market since its IPO.

Subscribe now

Now what makes Kinsale special? After all, there are thousands of insurance companies in the world. To recap the thesis, Kinsale is good at 4 things:

1. Underwriting Expertise: They are willing to take risks that are complex and too difficult for most.

2. Technological Innovation: The company invests in machine learning risk assessment, real-time risk monitoring, and cybersecurity risk evaluation to enhance its risk management capabilities. (Obligatory AI reference of course). Other insurance companies don't have 27% of their employees in the tech department, as Kinsale has.

3. Efficient Claims Processing: Average processing time of 7.2 business days and a 94.6% claims resolution rate. These are some of the best in the insurance industry.

4. Customer Relationships: The company maintains a retention-focused approach, resulting in an 85% customer retention rate and an average client relationship duration of 7.2 years.

These are all pretty good benchmarks for the US insurance industry, which is widely hated for its byzantine rules and slow claims processing. But Kinsale stands apart, as do its results.

Financial Performance

1. EPS: Q4 2024 EPS of $4.62 vs the analyst forecast of $4.30. (Beat)

2. Revenue: Reached $412.12 million, $45 million better than the expectations, a 29.7% year-over-year increase. (Beat)

3. Net Income: $109.1 million, up from $103.4 million in Q4 2023, while full-year net income surged 34.5% to $414.8 million

4. Gross Written Premiums: Grew by 12.2% to $443.3 million compared to Q4 2023. Underwriting income was $97.9 million for Q4 2024. For the full year this figure was $1.9 Billion, a 19.2% increase YoY. Net Earned Premiums also grew 21.2% to $359.74 million. They also got a boost of $9.6 million due to releasing reserves from prior years that weren’t realized, showing their risk management.

5. Net Investment Income: Increased by 37.8% to $41.9 million compared to the same quarter last year.

6. Combined Ratio: Maintained a strong combined ratio of 73.4% for Q4 2024. (Analysts had estimated 76.3% so this was much better) with a loss ratio of 52.3% and an expense ratio of 21.1%. For the full year, combined ratio was 76.2%, which was steady, but remarkably, this means they performed better in Q4, which bore the brunt of the wildfires, than they did in the year as a whole. Return on Equity clocked in an excellent 29%.

A quick recap of what some of these mean for newer readers or people looking into Kinsale for the first time.

Gross Written Premiums: Total amount of premiums collected from policies issued by Kinsale in the period.

Net Earned Premiums: Portion of premiums actually earned by the insurer after deducting reinsurance and other adjustments.

Expense Ratio: Percentage of the premiums earned goes towards operating costs, such as salaries, marketing, and administrative expenses

Loss Ratio: What percentage of the premiums earned is paid out in claims. If the loss ratio is 52%, it means that for every dollar earned in premiums, 52 cents is paid out in claims.

Add these 2 and we get the Combined Ratio: To over-simplify, it’s a measure of profitability, where a ratio below 100% indicates profit, and above 100% indicates loss. And Kinsale is the best of its peers in E&S.

Source: Kinsale's Q4 2024 Earnings Call Slides

In other words, in Q4 2024, Kinsale collected a $100 from selling policies, at a cost of $21 and ended up paying out $52 in claims on those policies, thereby leaving $27 for the company as profit. Simple right? That’s pretty much the business model – keep most of what you make.

Now, of course, every company says it has competitive advantages over its peers. Fortunately, we can check using standard metrics to see how meaningful those claims are.

For context, there is only one company with a market cap larger than Kinsale and a return on equity higher than Kinsale – Progressive, which dominates auto insurance all across America. PGR’s combined ratio? 89%.

If we use Finchat’s industry filter, specifically for E&S insurers, we see Kinsale is top of the pile when it comes to profitability:

I did a quick scan of these names and not a single one has a combined ratio less than 85%. I believe the technical term for Kinsale is “best of breed”.

Hence, understanding how such a company grows is also a simple matter, broadly it can be done through the following ways:

1. Raise premium prices – This provides natural inflation protection but can’t be done too aggressively lest they get undercut

2. Sell more policies to more people – But this must be done at the right price with the right benefits to entice people to take them up

3. Increase efficiency by reducing expenses and keep more of the premiums (Kinsale is world class at this as we’ve seen)

4. Invest the float earned wisely and prudently, managing both return and liquidity (Perfected by the GOAT Buffet himself). CFO Bryan Petrucelli expects continued benefits from higher new money yields, averaging in the low 5% range, which should further enhance net investment income moving forward.

5. Expand into other verticals.

On that note, the company specifically announced expansion plans:

1. Kinsale plans to expand into personal insurance and adjacent markets, including high-value homeowners' insurance in catastrophe-prone areas.

2. New Business Unit: Introduced a new agribusiness underwriting unit, focusing on farm, ranch, and related spaces

We see CEO Michael Kehoe’s philosophy in full play here – which is to expand deliberately both vertically (through market share gains) and horizontally (into newer insurance lines)

Subscribe now

Now, some may question the wisdom of going into catastrophe-prone areas given how many established insurers are retreating and the frequency of extreme weather events keeps increasing.

Kinsale itself got burned (pun intended) from the Southern California wildfires with $6.2 million in after-tax catastrophe losses for Q4, primarily from the Southern California wildfires. In January 2025, preliminary estimates pegged additional wildfire-related losses at $ 25 million pre-tax.

While this doesn’t sound great, it’s the point of insurance – you pay out when bad things happen, but collect a fortune to protect against them at all other times. Ironically, with more established insurers retreating, it would likely make it more lucrative for those willing to step into the void and offer hard-to-price insurance in disaster zones, as long as they can make the math work. For example, the initial loss estimates were actually $45 million, but KNSL had re-insured about half that amount and hence ended up with only $25 million in losses. That’s excellent risk management, without a doubt.

The other risk to worry about is as Kisnale moves away from niche areas into more “mainstream” categories like personal insurance, the field gets quite crowded quite quickly. Expenses go up because you have to market aggressively, pricing goes down because there are plenty of cheaper alternatives and hence margins get hit. So, it will be interesting to see how the company can navigate these challenges.

Despite strong results, the stock sold off 6% on the day. My best guess here is because Gross written premium growth decelerated to 12.2% in Q4 from 19% in Q3, this is clearly a result of increased competition, particularly in larger property accounts, which impacts premiums, and was clarified by the company in the call.

Will this continue? We will have to see, that said, the company maintained its outlook of 10-20% growth in gross written premiums over the long term. Until this is disproven, we have no reason to believe otherwise.

As CEO Kehoe asserted: “Our model is built for cycles, not quarters”— that’s a great line and one that I hope will bear out for years to come.

Conclusion

Not every business is like the Trade Desk, where the CEO gives masters level theses on the earnings call or innovates at the cutting edge of ad tech. And because so many people follow it, we see the stock routinely having a 50% move at least once a year. In contrast, Kinsale is a very simple, but equally well-run business. A boring one, and that’s a good thing in this case.

Most of the earnings call (and I encourage everyone to listen) consisted of analysts worried about, if not outright sceptical of competition and pricing pressures. I totally understand where they’re coming from, but this line of questioning suffers from the same problem as my first 2 attempts at this article – the tendency to overcomplicate something simple but complex drivers. It’s really exciting to model in a wildfire or hurricane impact or re-insurance rates or distribution of small casualty accounts. I’m kidding, but professionals get paid to do this for a living and so have no choice but to ask questions about them.

Fortunately, we have a founder CEO that has spent a long time making things simple and continues to show results through consistent execution. Kinsale remains confident in their growth projections and their expense outlook, and if those 2 hold up, the rest is just simple math.

That’s what we should all be focused on and hopefully this is a lesson I will remember next time.

Is The Stock Attractive Right Now?

Read more

No long introduction, let's just jump into Celsius Holdings' (CELH) Q4 2024, look at the Alani acquisition it announced and see if the stock is interesting right now or should be avoided. Let's dive in!

The Numbers

To put the numbers into context, this is Celsius revenue growth, as seen through Finchat.

I find this overview extremely insightful for at least two reasons. The first one is that you can see that Celsius has grown its revenue by 52.8% per year on average. But this big drop in revenue growth, just 2.9% in 2024, is not the first drop. 2011 and 2012 were particularly weak years as well. So, people declaring Celsius dead here are premature, to say the least.

In 2021, Celsius signed a distribution agreement with Pepsi. That had an immediate effect. Growth more than doubled for three years in a row. And from 2020 to 2023, revenue went up 10x. In that context, it's not that strange there is a pullback now.

In the meantime, Celsius became the third biggest player in energy drinks, after Red Bull and Monster.

(All images in this article are from the Celsius Q4 2024 presentation, unless mentioned otherwise)

While this looks great, since the second quarter of 2024, the company had been losing some market share, from 12.3% to 10.9%.

In Q4 2024, net sales were $332.2 million, down 4.3% from last year. The main reasons were promotional discounts and order timing. But still, this beat the estimate by $5.18 million or 1.6%.

It's also important to look at the comps. The company's quarterly revenue was down 4% year-over-year, but that was compared to 95% growth in the same quarter a year ago.

Full-year 2024 revenue reached $1,360.3 million, up 2.9%.

Profitability was better than last year's Q4, with gross margin at 50.2% in Q4, up from 47.8%, thanks to lower costs. But gross margins didn't reach the Q1 & Q2 2024 numbers.

On a GAAP basis, Celsius reported a loss of $18.9 million, mostly because of one-time costs like legal expenses for the Alani acquisition. In this case, I think adjusted earnings show you reality better. On an adjusted basis, net income was $23.7 million, or $0.14 EPS, beating the estimate of $0.10.

The Alani Acquisition

The big news from the earnings was the Alani acquisition. Celsius bought its competitor for $1.8 billion.

One of the lessons I learned from my losers in the past is this:

Logically, I considered selling Celsius. But there are at least three differences with Teladoc and Block.

1. The size of the acquisition

The acquisitions for Teladoc and Block were much bigger. Teladoc paid a whopping $18.5 billion for Livongo. That was about the same size as Teladoc itself. Block issued 20% new shares to buy Afterpay, valuing Afterpay at a price of $39 billion at the time the acquisition was announced. The Celsius acquisition is much smaller.

You could say that Alani is about 30% of Celsius' market cap of about $6B. That's close to Block's initial AfterPay acquisition price, which was about 30% of its market cap too. However, the difference is that the AfterPay acquisition happened at the top of Block's market cap.

Here you can see when the acquisitions happened.

For Celsius, which is down a lot from its highs, a 30% acquisition at the top of its market cap would have meant an acquisition of about $7 billion, almost 4 times more than the $1.8 billion price.

2. The Valuation

When Block bought AfterPay, that company had about $650 million in yearly revenue. That means the initial price valued the acquisition at a whopping sixty times sales. When the deal closed, the valuation of Block had collapsed, and the deal was only worth about $12 billion. But even then, this is a sales multiple of 18.5 times.

The year before the acquisition by Teladoc, Livongo had $170 million in sales and in 2020, the year of the acquisition, it was projected to have $287 million in sales. At the price of $18.5 billion, that means Teladoc paid 109 times revenue or 64.5 times forward sales.

Looking back, these valuations were extreme red flags.

Celsius buys Alani for $1.275 billion in cash and $500 million in stock for revenue of $595 million. That's a bit less than three times Alani's revenue.

3. The Type

Block bought AfterPay to get access to a market it was not present in before, BNPL (Buy Now, Pay Later). For Teladoc, the reasons were not so clear, although the company cited preventive healthcare, more data insights and (don't laugh) shareholder value.

For Celsius' Almani acquisition, the reasoning looks more straightforward and more focused. It's taking out a competitor that was growing fast and instead of competing, the companies can work together now to get bigger.

It’s this together now.

So, how does this affect the balance sheet?

Celsius has about $900 million in cash and no debt. So, that means the company will have to issue debt for the acquisition, as there's $1.275 billion in cash in the deal. Celsius said it would borrow $900 million, which was already committed. That will leave the company with a net debt of $375 million. On top of that, there's an immediate tax benefit of $150 million for Celsius.

In 2024, Celsius generated $170 million in free cash flow. With Alani, that's potentially more, although there may also be costs involved. We don't know that. Of course, companies will always pound their chest about synergies. I see them as bonuses. If they really come to fruition, that would be great, but I don't believe that strongly in synergies as a given.

But I do believe in the brands working well together. This is the product line of Alani NU.

So, there's energy drinks, but there's also more, with preworkout powders, sticks, packs, shakes, and snacks.

And while you often hear management talk nonsense about why they did the acquisition, the overview Celsius gave actually makes sense.

Yes, it's still business buzz, but all in all, the company is just throwing out some PowerPoint bullet points.

If you talk about number three, for example, the complementarity of the brands, this is shown convincingly.

This is not just promo talk of a management team that has a new toy. This looks like good business strategy to me.

Alani Nu is targeting more women and you can already see that jump off the page.

And the fifth point looks at the numbers. These too look very convincing.

So, all in all, I really like this acquisition. It's a sizeable one but not so big it chokes the company chewing it up. It will add more revenue and profitability (which was definitely not the case for the previous acquisitions we discussed, those of Teladoc and Block). The brands are close enough to understand each other but different enough not to really cannibalize each other. The balance sheet remains strong enough.

Of course, I know the numbers. 72% to 84% (depending on the source you consult) of the acquisitions are not good for shareholders and actually destroy shareholder value.

Could this be the case here? For sure. There can be integration problems, cultures that don't mix well, dead bodies in the Alani Nu closet, etc.

I've become much more skeptical about acquisitions, except for tuck-in acquisitions and acquihires (where you buy the company to have great people on your team).

Of the many deals I hear about, this is in the top range when it comes to business logic. This makes intuitively sense. But execution will be needed, of course.

More Insights From The Call

For 2024, Celsius posted meagre revenue growth, at just 3%. But the real story is in the retail sales numbers, which were up 22% year over year. That number excludes the Pepsi inventory problems. So, it seems that in stores, Celsius is still selling well. Once the inventory problems at Pepsi are (finally!) over, it also shows that Celsius can go back to growth.

The 22% retail sales growth was also boosted by Celsius being in more stores. CEO John Fieldly on the earnings call:

Consumers can purchase cold, refreshing Celsius in more than 241,000 outlets in the United States.

In an industry like the energy drink space, being everywhere matters. Celsius will also be available in Subway in March. That's great, as Subway has 18,000 potential locations. It will also be sold in Home Depot.

That 22% retail jump also outpaced the energy drink market as a whole, which grew at 5%. So, even with this tough year, overall, Celsius is still grabbing market share.

Important for the investment thesis: Celsius drove 30% of the category’s growth in 2024. The investment thesis has always been that the company would

CEO John Fieldly acknowledged the increased competition.

We expected our competitors to respond eventually, and it came with a flurry of new sugar-free products last year.

Monster launched its sugar-free product in 2023 and Red Bull doubled down on zero sugar with its very first totally sugar-free flavour, the Pink Red Bull, with a combination of raspberry, other forest fruits and verbena.

In this context, Celsius’s 22% retail sales growth in Q4 shows considerable strength.

International growth remains a highlight, although still from a small basis.

The expansion to Canada, the UK, Ireland, France, Australia and New Zealand had effect, but there's still has a ton of upside left.

The Potential Multibaggers Quality Score

So, how does this all influence the PM Quality Score? Let's find out.

Read more

Karan here with the Upstart analysis and Kris will chime in to update the Quality Score and valuation.

The last time I wrote about Upstart (UPST), I ended with this paragraph, covering the Q2’24 earnings

To be honest, I have nothing to dispute the above characterisation. The stock ripped due to an increase in guidance, rightfully so. I don’t think this is the quarter that changes the story after years of disappointment, but that quarter is one step closer.

The nice part about being the patron saint of lost causes, is the satisfaction you get if they eventually find their way.

I’m not claiming victory on a stock that’s still down 90% from its all-time high (even if that was a bubble so be it) but the company hasn’t gone out of business yet, indeed they may finally be on the cusp of becoming a sustainably profitable company with a business model even CNBC traders can understand. I look forward to Q3’24 with the highest confidence I’ve had in a while.

6 months have passed since – let’s find out if I was right.

Subscribe now

The Numbers

First, the highlights.

• Q4 2024 results, adjusted EPS of $0.26, blowing past the $0.04 consensus, jumped from -$0.06 in the prior quarter and -$0.11 a year before.
  

• Q4 revenue of $219M, also surpassing the $181.9M consensus, climbed from $162.1M in Q3 and +56% from the year-ago period.
  

• Revenue growth accelerated from 28% sequentially in Q3 to 35% in Q4.
  

• “In Q4 of 2024, our business grew dramatically across all product categories, delivered Adjusted EBITDA at levels not seen since the first quarter of 2022, and came within a whisker of returning to GAAP profitability,” said Co-Founder and CEO Dave Girouard.
  

• Adjusted EBITDA of $38.8M, handily beating the Visible Alpha estimate of $6.1M, compared with $1.4M in Q3 and $0.6M in Q4 2023.
  

• Adjusted EBITDA Margin was 18% of total revenue, up from 0% in Q4 2023.
  

• During the quarter, 245,663 loans were originated, totalling $2.1B, up 68% Y/Y and up 33% Q/Q. The conversion rate was 19.3%, up from 11.6% in Q4 2023.
  

• Contribution Profit was $122 million in the fourth quarter of 2024, up 28% YoY, with a Contribution Margin of 61% compared to 63% in Q4 2023.

FY 2024:

(*Checks date to see if I’ve somehow time travelled back to 2022 when Upstart was beating and TTD getting smashed*)

No? It’s February 2025 and UPST has a double beat?? UPSTART?

The former pandemic darling turned zombie company turned left-for-dead object of CNBC jokes? Every single metric is trending in the right direction. What a long way we’ve come since 2023.

And not just a beat, but a blown-out-of-the-water beat.

Wait, what if it was just seasonality? Let’s look at the guidance.

• For Q1 2025, Upstart expects revenue of ~$200M (vs. the $184.6M Wall Street consensus); adjusted EBITDA of ~$27M (vs. $10.7M Visible Alpha estimate).
  

• “The company sees full-year 2025 revenue of about $1B, surpassing the $823.1M average analyst estimate, with revenue from fees of ~$920M and net interest income of ~$80M. It also expects adjusted EBITDA margin of ~18% (vs. 13% estimate).

For the first quarter of 2025, Upstart expects:

• Revenue of approximately $200 million

  • Revenue From Fees of approximately $185 million

  • Net Interest Income (Loss) of approximately $15 million

  • Contribution Margin of approximately 57%

  • GAAP Net Income (Loss) of approximately ($20) million. Adjusted Net Income (Loss) of approximately $16 million

  • Adjusted EBITDA of approximately $27 million

  • Basic Weighted-Average Share Count of approximately 95 million shares. Diluted Weighted-Average Share Count of approximately 105 million shares

For the full year 2025, Upstart expects:

• Revenue of approximately $1 billion

  • Revenue From Fees of approximately $920 million

  • Net Interest Income (Loss) of approximately $80 million

  • Adjusted EBITDA Margin of approximately 18%

  • GAAP Net Income to be at least breakeven.

1 billion dollars of revenue and break-even on a GAAP basis? You can read the rest of this article with a mental image of me as follows:

Now what the heck is driving such an optimistic forecast?

Origination volume and adjusted EBITDA margins snapped back to their highest levels since 2022. Improving conversion and funding availability were named, because of a better underwriting model and macro improvements.

Subscribe now

All product categories contributed, with auto and HELOC originations both up 60% quarter-over-quarter, and small-dollar relief loans showing 115% growth. (HELOC = home equity line of credit, so a loan secured by your house).

The launch of Upstart's Model 19, which introduced the Payment Transition Model ('PTM'), resulted in more accurate underwriting.

In the HELOC business, this gave Upstart more than 1,000 loans and no defaults. The HELOC business operates in 36 states covering 60% of the U.S. population.

CEO Dave Girouard highlighted that these products now represent over 35% of total volume, diversifying revenue streams beyond personal loans.

Don’t be too scared by the 115% of small dollar loans surging, it represents only 3% of origination dollars since these loans are capped at $5k anyway.

With this kind of origination growth, Upstart's superior credit model must still be holding up to attract more volumes, let’s check in:

Yep, the best risk grades are still much better than FICO equivalents and Upstart can still find good borrowers amongst the untouchables from FICO’s data sets.

How about the capital situation?

CFO Datta noted an upsized $1.3 billion commitment from capital partners and a $150 million personal loan warehouse facility, which underscored growing confidence in Upstart’s platform.

How about those pesky balance sheet loans?

Down across the board. One word: excellent!

Where does this leave UPST as a company?

Operating expenses grew just 7% YoY despite 56% revenue growth, driving the adjusted EBITDA margin expansion to 18%. Upstart ended 2024 with $788 million in unrestricted cash and reduced balance sheet loans by 28% to $806 million, lowering capital intensity.

Contribution margin, similar to gross margins, remained robust at 61%, supported by pricing power in fee-based revenue streams. Both came in at multi-year highs. This might be the furthest Upstart has been from bankruptcy since it came public. What a turnaround!

Highlights from the call

I’m also sharing a summary of some of my key takeaways from the conference call.

AI Model Enhancements

The launch of Model 19 introduced three critical upgrades:

1. Payment Transition Model (PTM): Improved risk separation by 14% and model accuracy by 8%, enabling better borrower outcomes.
   
   

2. APR Integration: Enhanced ability to match borrowers with optimal rates, reducing defaults.
   
   

3. Economic Resilience: Better performance across interest rate environments, as evidenced by 25% improvement in delinquent borrower repayments within 14 days of contact.

Upstart AI Day and Product Roadmap

The upcoming May 14, 2025 AI Day will showcase new developments in embedded finance and cross-platform lending capabilities. Management telegraphed plans to:

1. Expand Auto Retail Lending: Partner with dealership networks to capture point-of-sale financing
   

2. Launch Credit Card Solutions: Leverage existing bank partnerships for revolving credit products
   

3. International Pilot Programs: Initial focus on Canadian and UK markets

So, we have finally arrived at the dawn of the two narrative drivers – credit cards and international expansion. How these proceed will determine if Upstart sees a triple digit stock price again (without short interest).

However you look at it, this is not bad news by any means. If Upstart can crack the US credit card industry, we have a whole new ballgame on our hands, even if it takes 5 years.

Macro Tailwinds

The company estimates every 25 bps rate decline drives 5-7% incremental originations. We can clearly see this in the boost to volumes in Q4 alone as the 2024 rate cuts passed through. If we get more in 2025, it would be relatively straightforward for Upstart to trounce guidance again. Upstart’s refinance-focused products are currently 40% of volume and could explode higher as rates come down, fully reversing the headwinds from 2022.

Conclusion For The Quarter

Is it too early to take a victory lap? Probably.

An errant tweet from a certain person could dissolve the Fed tomorrow and Upstart might once again collapse 40% overnight, who knows, really.

But I’ve spent 2 years looking at this company’s business drivers and I’ll be damned if I let go of the opportunity to celebrate hanging in there.

I said in Q4 2023, that if UPST were to survive this period, it would print money on the other side. We’re on the cusp of the other side now; if you’ve been waiting to right size your position or take profits, the moment shall soon be upon us, but likely so will be the upswing back to triple digits, invest wisely.

Upstart’s Q4 results to me were a watershed moment, demonstrating what the company CAN achieve with the right tailwinds. Its AI-driven underwriting models can drive profitable growth at scale. The combination of 60%+ product growth in auto/HELOC, expanding EBITDA margins, and $1 billion revenue guidance suggests the company is transitioning from recovery phase to sustainable expansion. With the possibility of credit cards and international expansion, we see something we haven’t seen in a long time for Upstart: Optimism. Finally.

Maybe I should only write about Upstart once every 6 months from now on?

I’ll leave it to the community to decide.

But let's go to Kris now for the rest of this article.

Subscribe now

PM Quality Score

Did you hear that?

You thought it was the wind, right?

Or a man with respiratory problems somewhere around you?

No, it was Karan's sigh of relief. As the Patron Saint Of Lost Causes, it's good to have a miracle every now and then to prove you are what you claim.

As you know, I always look at the quality, through the PMQS (Potential Multibaggers Quality Score) and the valuation to get to the QPI (Quality-Price Index). That system allows us to balance out quality and the valuation and not look at valuation alone, as so many do.

Read more