I am Ray, a passionate investor and equity analyst. And today I am covering the Best Company in Latin America.

The Dutch East India Company became exceptionally wealthy trading goods between Europe and Asia.

Cornelius Vanderbilt was one of the US railroad pioneers during the late 1800s. His railroads enabled the transportation of goods from manufacturers to customers at a never-before-seen speed and efficiency!

John Pierpont Morgan empowered money to flow through the booming US banking system.

What if I told you that there is a single company doing all of these 3 things?

Facilitating trade of goods!

Enabling the transportation of goods!

Empowering the flow of money!

If you read the title of this report correctly, you know I am talking about Mercado Libre.

The company has ambitions to become the ultimate commerce and technology champion for the 670+ million people of Latin America.

In my opinion, Mercado Libre is one of the best long-term opportunities, because analysts perpetually underestimate its TAM!

Many investors mistakenly believe that because the economies of Latin America are weaker than those in the US or Europe, e-commerce will have a tougher time there.

Yet, in the last 5 years:

• GMV has grown by 238%

• Payment Volumes by 516%

• Revenues by 700%

• Operating profit by 2,282%

Despite that, the stock has increased only 15%!

In this Deep Dive, I will explain why Mercado Libre is the best company in Latin America and why it is currently significantly undervalued.

Let’s dig in.

1. Facilitating the Trade of Goods

2. Enabling the Transportation of Goods

3. Empowering the Money to Flow

4. Risks

5. Competition

6. Opportunities

7. Financial Analysis

8. Valuation

9. Valuation Model

10. Conclusion

1. Facilitating the trade of goods

Globalization has made us more connected than ever. No longer are we forced to be satisfied with local products. Goods from all over the world are now available with just a click of a button, quickly and cheaply.

That wasn’t always the case, and still isn’t the case in all parts of the world.

Mercado Libre is the company that’s making that a reality in Latin America.

1.1. Mercado Libre Marketplace

Marketplace is their flagship product, an online marketplace that facilitates the trade of products.

Customers can easily browse the marketplace for millions of products in thousands of categories. The platform has become extremely popular with sellers and customers alike.

The platform has all the features of a sophisticated e-commerce platform. Reviews from thousands of other customers, detailed product descriptions, pictures, videos, customer support, a broad set of payment options, delivery information, and more.

Mercado Libre is primarily a technology company that strives to deliver the best experiences for all its stakeholders.

“Our main focus is to serve people in Latin America by enabling wider access to retail, digital payments, and e-commerce services, and by providing compelling technology-based solutions that democratize commerce and money, thus contributing to the development of a large and growing digital economy in a region with a population of over 700 million people and one of the fastest-growing e-commerce and internet penetration rates in the world.” Mercado Libre 2025 10K

The technology-first approach has enabled the company to create one of the fastest-growing e-commerce platforms in the region.

Gross merchandise volume measures all sales on the company’s e-commerce platform.

GMV growth has been parabolic. In Q1 2026, GMV grew by an impressive 42% Y/Y to $18.9B, the fastest growth rate in many years.

Since 2012, GMV has grown by 1,139%, a CAGR of 21%!

The marketplace is essentially the backbone of the company. The more it’s used for commerce, the easier it is for Mercado to upsell various other services, but we will get to that later.

Overall, GMV growth can be a great barometer for the growth of the company.

There are two main direct ways the company monetizes the marketplace:

• First-party sales (1P)

• Third-party sales (3P)

When Mercado sells products that it owns and holds in inventory to customers, it acts as a first-party seller. In this scenario, Mercado sources goods directly from the manufacturer or the distributor and handles the transportation and storage.

Most crucially, the company is in full control of pricing and recognizes the entire sale as revenue.

So, if they sell a blender for $60, they recognize the whole $60 as revenue.

Meanwhile, 3P is when a product purchased by the customer is sold by an independent merchant, not by Mercado Libre itself.

Mercado charges the merchant fees for using the platform and recognizes said fees as revenue.

All the shipping costs, insurance, warranty, returns, and COGS are paid by the seller. For instance, in Mexico, the fees range from 12.5% to 22.5% depending on the product category, and whether sellers use any extra services. In other countries, fees might be different.

The economics of these transactions are quite different.

1P depends on their ability to negotiate a favorable price with a supplier and the customer’s willingness to pay for the product. In 1P, gross margins are low, sometimes as low as 10%-20%.

However, remember that in a 3P sale, the independent merchant is responsible for his own expenses. Mercado incurs the costs of maintaining and managing the platform, but as the platform has a massive scale, Mercado can achieve a very high gross margin on its 3P sales.

In the $60 blender example, Mercado would keep about $9 in 80% gross margin platform fees and recognize those $9 in revenue.

As of LTM of Q1 2026, 2.66B items were sold on the marketplace.

This is a 38x increase in 13 years, a CAGR of 32%!

Subscribe now

1.2. Mercado Shops

Mercado Shops used to empower sellers to quickly, cheaply, and efficiently create an e-commerce website. This offering was similar to Shopify.

This was a platform with hundreds of different features and integrations, not only with all of Mercado’s services but also with various third-party apps. Shops had a website-building portal with lots of templates that enable people without technical skills to design their website.

However, in stark contrast to Shopify, Mercado didn’t charge a fixed monthly subscription. For example, in Mexico, the use of the platform is free, and there was no cost per transaction. Later, however, the company introduced various transaction fees.

Mercado’s strategy was to monetize shops by feeding all their other business lines!

Clients used Mercado’s logistics to store and deliver products. Mercado advertising solutions are quite useful to increase sales. Various financial products can be used to finance inventory and other expenses.

Sellers could build their brand using social media and other channels outside of Mercado’s ecosystem. As they continue building a brand and increasing sales, Mercado benefits as the provider of services. At the same time, sellers are not excluded from Mercado’s massive marketplace.

However, ultimately, Mercado Shops was a leaky bucket for them.

They realized that merchants who grew large enough on Shops eventually migrated to Shopify anyway because Shops lacked all the advanced features that Shopify built over the years to specifically serve very large merchants.

So Mercado Shops was closed in 2026, and replaced with Mi Pagina. (My page in Spanish)

Instead of a separate website, Mi Pagina is a branded storefront hub within the Mercado Libre platform that lets merchants create a personalized business profile complete with their own logo, banners, and curated product collections.

Unlike Mercado Shops, which was a standalone e-commerce website with a custom external domain, Mi Pagina operates entirely within Mercado Libre’s ecosystem, utilizing standard marketplace fees and charging a small monthly maintenance fee after a free trial.

Mercado Libre made this pivot because it decided to stop competing with enterprise software e-commerce players such as Shopify. Instead, they consolidated their technology to focus heavily on their core strengths, locking small-to-medium sellers into their highly profitable platform.

1.3. Mercado Libre Classifieds

Classifieds is a separate service that allows people to list real estate, cars, and various other local goods and services. The platform is comparable to Craigslist and the Facebook marketplace.

The platform is monetized by Mercado, charging real estate agents and car dealerships a placement fee for advertising on the platform.

Additionally, large merchants pay for subscriptions for advanced features.

The company itself doesn’t sell goods. Mercado wants to create an independent ecosystem for real estate agents, car dealerships, and other professionals, so they trust the platform.

Classifieds serve as a great way to further monetize its active client base!

The platform in itself is not a significant revenue driver, its main objective is to generate traffic for Mercado’s other businesses. A customer who purchased a vehicle is more likely to buy accessories such as a phone stand, a foot mat, an air refresher, and so on. Additionally, Mercado can upsell various financial services, such as a car loan and insurance. Similarly, with real estate, new houses require TVs, blenders, coffee machines, paintings, and other products.

Mercado Classifieds is leveraging technology, scale, and customer trust to offer a superior product at a lower cost!

2. Enabling the transportation of goods

Mercado Envios is the company’s logistics arm.

For any E-commerce company to gain scale, a sophisticated logistics network is a must. If one wants to buy sneakers, they are willing to wait if it’s cheaper, however, if the wait time is too long, they will just go to a brick-and-mortar store.

Simply put, people love convenience and are generally not very patient!

However, delivering packages is expensive and takes time.

If E-commerce is to take market share from brick-and-mortar stores, it must offer a superior customer experience.

Customers want quick and cheap delivery, however, there are too many middlemen in the process. Various distributors, warehouses, customs agents, trucking companies, cargo airlines, last-mile delivery providers, and so on. Each involved party delays the process and adds an additional cost.

E-commerce platforms realized that to deliver packages quickly and at a low price, they must take logistics in-house.

That is what Amazon did in the US, Alibaba in China, Coupang in Korea, and that is what Mercado Libre is doing in Latin America!

Read more

I am Ray, a passionate investor and equity analyst.

Today, I am reviewing the May performance of the Global Equity Portfolio.

A reminder that I track all aspects of the Global Equity Portfolio for my Premium subscribers.

• All sales/buys are revealed as they occur in the Substack Chat

• Portfolio composition

• Performance

• Strategy

• Stock watchlist

• Potential catalysts

Since its founding, the portfolio has returned 161.22%, a CAGR of 17.06%, beating the S&P500 return of 126.97%, CAGR of 14.39%!

In the above image, we see a graph from Portseido that compares the performance of the Global Equity Portfolio to the S&P500 if invested in the index at the same time as I bought the stocks.

I have been using Portseido to track the performance of the Global Equity Portfolio for some time now. They are an investment performance tracking and visualizations platform that is extremely easy to use. I especially like their smart benchmarking graphs and tables you can see in the picture above.

This is exactly what I was looking for, and I would recommend them to everyone looking to approach investing seriously and beat the market in the long term.

Use my link below to begin your 2-week trial.

https://www.portseido.com/?fpr=global

If you are new to the Global Equity Portfolio, I recommend reading my full Portfolio reveal and analysis from November 2025.

Let’s take a look at May’s performance.

1. Performance

The Global Equity Portfolio rose 8.16% in May, compared to +6.45% for the S&P 500 and +7.51% for the Nasdaq!

Read more

• AI winner stocks are ripping, reaching new highs almost daily.

• Everything else is either flat or falling.

Mr. Market has placed Meta in the second camp, as Zuck’s social media giant has seen its stock fall 20% from the August 2025 highs.

Investors are concerned that Meta might be overinvesting in AI infrastructure without having an easy and direct way to monetize that infrastructure.

Microsoft, Google, and Amazon have highly successful enterprise cloud businesses that are showing accelerated revenue growth, driven by AI startups and large enterprises implementing AI across their organizations.

Meta doesn’t have such a business, and it doesn’t have highly visible, easily monetizable consumer AI products comparable to ChatGPT, Claude, or Gemini.

Furthermore, Meta just reported its first quarter-over-quarter daily active user decline in years.

These factors have built a strong bear narrative that has dragged down the stock.

But if one looks under the hood, the reality of a high-quality undervalued business smacks you in the face.

Revenue growth is accelerating, margins remain top-class, AI is a clear tailwind, and the valuation is by far the most appealing among the Magnificent 7 companies.

Most importantly, this week, Meta revealed a slew of consumer subscriptions, aiming to further monetize its most active users and create directly monetizable AI products.

In today’s article, I will explain why the market is wrong about Meta!

• It can monetize AI

• It will continue gaining users

• It is undervalued

Let’s begin.

1. Subscriptions

2. AI

3. Growing User Base

4. ARPU Growth

5. WhatsApp Monetization

6. RealityLabs

7. Valuation

8. Valuation Model

1. Subscriptions

On May 27, 2026, Meta announced a set of paid subscription plans for its 3 flagship apps.

• $3.99 a month for Facebook Plus.

• $3.99 a month for Instagram Plus.

• $2.99 a month for WhatsApp Plus.

These tiers operate in parallel to the existing Meta Verified program, which continues to provide verification and impersonation protection for creators and businesses at price points ranging from $15 to $500 per month.

Ultimately, Meta plans to consolidate these social offerings, along with professional and AI packages, under 1 unified brand known as Meta One.

This integrated ecosystem aims to create high-margin recurring revenue products as the company invests aggressively in its Reality Labs VR venture and AI buildout.

The Plus subscriptions provide optional tools for users looking for deeper customization, detailed analytics, and enhanced privacy controls.

Interestingly, these tools do not remove advertising from the feeds.

Instead, they establish a 2-tier user experience where paying subscribers receive functional perks while non-paying users operate with standard features and a less engaging interface.

This makes sense as the average US DAU is estimated to generate over $80 in yearly revenue. This subscription would generate less than that, about $48, and that assumes no churn.

So revenues generated by these new offerings will be incremental, and not reduce advertising revenues.

For Instagram Plus, subscribers paying $3.99 per month receive tools that provide advanced insights and profile design choices.

Key capabilities include:

• More detailed statistics

• How many people rewatched a story

• The option to build unlimited story audience lists

• Secret preview mode to view stories without appearing on the viewer list

• Extend stories beyond the standard 24-hour window

• Spotlight 1 story per week for extra visibility

• Search story viewer lists

Additionally, to enhance privacy, subscribers can post content directly to their profiles and highlights without displaying those posts in the feeds of their followers.

Cosmetic enhancements include animated Super Heart reactions, custom app icons, special fonts for bios, and additional profile pins.

Facebook Plus offers a matching suite of features, including story rewatch statistics, profile customization, and detailed user analytics.

WhatsApp Plus, priced at $2.99 per month, prioritizes messaging utility and visual personalization.

Subscribed users can change the visual theme of the application, apply custom app icons, send premium stickers, set unique ringtones, pin extra chats beyond the standard limit, and access advanced list customization tools.

Overall, these seem like cosmetic features that will appeal to creators, teens, or power users.

I am a user of all 3 of Meta’s main apps, and none of these features raises any interest in me. However, I imagine that millions of people will sign up, especially considering the affordable price tags.

These features remind me a bit of video game skins. For those unfamiliar, many video games offer players the opportunity to purchase in-game customizations that have no impact on the gameplay, they simply change how some weapons or characters look. Despite having no benefit on the game field, gamers spend billions of dollars on these in-game skins.

I expect these subscriptions to have the same effect.

Subscribe now

Meta is also introducing high-capacity plans for AI power users, creators, and businesses, consolidating them under the Meta One offering.

While basic access to Meta AI remains free, high-volume consumers will eventually hit usage caps.

Meta is testing 2 consumer AI subscription levels to resolve this issue:

• $7.99 per month Meta One Plus

• $19.99 per month Meta One Premium

The Meta One Plus for $7.99 a month tier is built for individuals who frequently generate images and videos or require extended reasoning capacity from Meta’s chatbot.

The Meta One Premium for $19.99 a month provides the same features as Meta One Plus but with significantly higher usage ceilings and access to the chatbot’s advanced thinking mode.

According to Helen Ma, the head of subscriptions at Meta, users who purchase these AI packages will also have access to premium features from the individual app subscriptions.

If that was not confusing enough, for business users and professional creators, Meta is launching 2 separate tiers designed to improve visibility, simplify management, and solve common operational pain points.

• $14.99 per month for Meta One Essential

• $49.99 per month for Meta One Advanced

Meta One Essential for 14.99 per month provides a Verified badge, impersonation protection, and enhanced linksheets to promote cross-platform presence.

Meta One Advanced for $49.99 per month gives businesses priority feed visibility on Facebook, higher ranking in Instagram search results, and enhanced Follow buttons on Reels.

It also features optimized scheduling tools, account moderator management without password sharing, and alert notifications when other users copy original content.

A major benefit is direct access to human customer support for Facebook and Instagram pages, resolving a historical pain point for small businesses.

Meta is currently testing all these subscriptions in various geographies before a global rollout.

Yes, it is a bit confusing with Meta now releasing 7 different subscription offerings, each with different features and use cases. This is the first serious foray of Meta into consumer subscriptions, so it is expected that there will be some learning moments on the way. I expect it will figure out which price points and feature sets appeal the most and consolidate and simplify its offering in the future.

Overall, I don’t believe that any of these subscriptions will create game-changing revenues for Meta.

In the last 12 months, Meta generated over $200B from advertising revenues. Let’s assume Meta gets to 100M subscriptions in a few years, which is a generous assumption. If the blended average subscription price is about $5, Meta would make $6B in subscription revenues.

That’s only 3% of its current advertising revenues. By the time Meta reaches that many subscriptions, advertising revenues would have already reached $300B. So, the subscriptions are just a small cake slice that will slightly increase revenue growth.

Nevertheless, these revenues are incremental and likely will have a huge incremental EBIT margin of 90%, as they would only require limited expenses. That could result in a $4-5B bump to EBIT.

Most importantly, it could end up helping Meta beat the narrative that it doesn’t have ways to monetize AI.

2. AI

A lot of investors are worried that Meta is overinvesting in AI development.

Read more

A share cannibal is a company that repurchases huge amounts of its own shares.

When a mature and financially strong company buys back its own stock, it directly reduces the float.

As a result, the remaining shareholders automatically own a larger piece of the business.

This shrinking of shares has a powerful effect on driving EPS growth.

If a company has $100M in net income and 100M shares outstanding, its EPS is $1.

If that company buys back 50% of its shares, leaving only 50M shares outstanding, the EPS jumps to $2.

This is a 2-fold increase in EPS, even though the core business net income did not grow at all.

By investing in growing businesses that regularly repurchase a lot of shares, investors could turbocharge their returns.

Mohnish Pabrai describes this capital return method as Uber Cannibal investing.

He points out that if a business grows its earnings by 15% each year and buys back 84% of its shares over 20 years, the investment turns into a 100-bagger with 0 changes to its price-to-earnings multiple.

If the earnings grow by 10% Y/Y, the company buys back 80% of its shares, and the price-to-earnings multiple expands by 50%, it becomes a 50-bagger in 20 years.

This shows how stock repurchases could compound wealth for patient investors, even in companies that are not seen as “exciting growth stocks”.

For some investors, 10-15% might not seem strong in this AI era, but if combined with consistent buybacks for over a decade, it could lead to quite healthy results.

So, in this report, I decided to take a break from the AI discussion and focus on these share cannibals that Charlie Munger loved so much.

1. Criteria for Investing in a Share Cannibal

Investors often hear quotes from famous investors without really understanding the context in which it was said and apply them the wrong way.

When Mr. Munger said to focus on the cannibals, he didn’t mean to simply look at the buyback yield or the announced buyback plan.

Not all stock buybacks are good. Many companies destroy cash by buying back stock under unfavorable conditions.

Management teams are constantly evaluated based on the performance of their stock. So, in the hope of increasing their share price, some CEO’s announce aggressive buyback plans. While buybacks are great, investors have been conditioned to love them, even when done out of weakness, and that has led to many CEOs focusing too much on buybacks and not enough on their business.

To find the best share cannibals, investors must look for some key criteria:

• Stable business

• BuyBacks funded with cash flow, not debt

• Consistent buybacks

• Net buybacks

• Organic growth

• Valuation discipline

• Reinvestment needs are covered

Let me explain.

A great share cannibal must have a highly stable, mature business.

It should not be in decline. If a company is losing sales or customers, buying back stock is useless because the core business value is eroding. Investors must look for companies with a high level of expected stability over the next 10 to 20 years.

Furthermore, ideally, stock buybacks must be paid for with cash flow, not debt.

They should never be funded by taking on high levels of debt. Some bad management teams borrow huge amounts of money to buy back stock just to artificially boost the share price in the short term. This is highly dangerous and can lead to massive trouble in the long term. Using debt to fund share repurchases is beneficial to long-term investors only if the company has a stable balance sheet and it can borrow at a fair rate to purchase undervalued stock.

A clear bad example of this is Hims and Hers. Last year, they raised $1B in a convertible note offering, diluting shareholders. It then committed to spending $250M of that on share repurchases. They spent $80M of this allocation on buybacks at a high share price in Q4 2025, and when it fell over 50%, they didn’t do any buybacks. Now the company is unprofitable and raising $250M in a convertible note offering. Why dilute shareholders by raising funds that were meant for investments, then not announcing buybacks, then raising money again? Such a mess.

So clearly, for buybacks to have a real effect, the company must be a reliable and consistent repurchaser of its own stock.

Sporadic buybacks often happen only when management has too much cash at the peak of a market cycle, like in the Hims and Hers example. This means that likely they pay very high multiples. Consistent buybacks average out the purchase price over time, showing a disciplined plan to return cash to shareholders.

Crucially, investors must not take the buyback yield and announced repurchase plans at face value.

Many companies issue a massive number of shares to employees every year. If a company repurchases shares just to offset employee dilution, those buybacks do not directly benefit shareholders. A real share cannibal must reduce its net share count over time, even after counting new shares issued for compensation or acquisitions.

Stock buybacks should act like gasoline on a fire, it should not be the only thing growing the EPS.

While a declining share count boosts EPS, the core business must also grow its net income. When organic growth is combined with a shrinking share count, the EPS compounds very quickly. Relying solely on buybacks to grow EPS without any real business growth is a bad long-term strategy.

Also, valuation discipline is a key rule for buyback success.

Repurchasing stock only makes sense if the shares are bought below their intrinsic value. If a company pays a ridiculous price for its own shares (like Hiims and Hers in Q4), it destroys cash that could be better spent elsewhere. Management must act like smart capital allocators and only buy when the stock is cheap, not only when they have cash.

Most importantly, a company must never underinvest in its core operations to pay for buybacks.

All reinvestment needs, like capex and R&D, must be fully covered before considering any buybacks. If a company underspends on R&D to buy back shares, it will lose its long-term competitiveness. Excellent cannibals can easily fund both technological innovation and massive buybacks.

Subscribe now

2. 10 Examples

The following 10 companies show how disciplined stock buybacks can build massive shareholder value over long periods.

2.1. Apple

Apple is the most famous share cannibal.

The company started its buyback program in 2012. At its peak in 2012, its outstanding share count stood at 26.47B. By 2025, Apple reduced its shares outstanding to 15.005B, a massive 43.3% decline.

Apple spends huge amounts of excess cash on repurchases. For example, in 2024, its net share repurchases totaled $94.9B, up 22.4% from 2023.

Apple did this while investing billions in R&D, releasing new highly successful products, and without hurting its product pipeline.

Let’s look at what results these buybacks have had on Apple’s income statement.

Since 2016, Apple has repurchased 31.3% of outstanding shares, reducing the share count by about 3.8% per year.

During this time, net income grew by 168% with a CAGR of 10.8%.

However, EPS grew by 297% with a CAGR of 15.3%.

This is a clear example of how buybacks work over a long period of time.

Total EPS growth rate was 77% higher than the net income growth, while the CAGR was 42% faster.

Essentially, had Apple not done these buybacks, EPS would have been $5.61, not $8.29 it is today.

Buybacks are much better than dividends because share repurchases are taxed more favorably and force the investor to reinvest more of their gains.

In the US, share repurchases are taxed at 1%, while dividends are taxed much higher, depending on the state and situation, at even 40%.

Furthermore, buybacks essentially force an investor to reinvest. If one receives dividends, they can be reinvested to compound for the long term, but most investors don’t do that.

Buybacks can also be viewed as the company essentially reinvesting your dividends in its own stock for you, so you don’t have to.

This is a tax-efficient way to amplify your returns.

2.2. AutoZone

AutoZone is an incredibly aggressive share cannibal.

It launched its repurchase program in 1998. Over 2 decades, it spent more than $36B to buy back almost 90% of its outstanding shares.

AutoZone’s annual shares outstanding dropped from 27.83M in 2017 to 16.52M in Q1 2026, a reduction of 40.7%.

Net income grew with a CAGR of 7.9% compared to 14.9% EPS CAGR.

2.3. NVR

NVR is a US homebuilder that has repurchased stock since 1994.

Its shares outstanding fell from 3.69M in 2017 to 2.63M in Q1 2026, down by 28.9%.

NVR’s buyback yield peaked in March 2026 at 10.5%.

Its average buyback yield from 2021 to 2025 was 7.1%.

Do you like the chart above?

FiscalAI makes managing investments smarter, faster, and stress-free.

• Visualizations

• AI-powered insights

• Financial data

• Earnings transcripts

• Portfolio analytics

All in one place. Instead of wasting hours digging through filings and spreadsheets, Fiscal.ai helps you get to the important information in minutes.

Save time, stay organized, and let Fiscal.ai handle the heavy lifting so you can focus on growth.

Join with my link below.

https://fiscal.ai/?via=global

Net income grew with a CAGR of 10.6% compared to 14.3% EPS CAGR.

The stock hasn’t done well in the last 5 years due to stagnation of the US new build housing market, because of rising inflation, and high interest rates.

However, the company continues buying back shares, spending $1.8B in 2025. Once the cycle turns and housing construction rebounds, investors will benefit from these buybacks.

2.4. O’Reilly Automotive

O’Reilly is an automotive part retailer that began its stock buyback program in 2011.

Since then, the company has spent over to $30B on share repurchases.

It has reduced its share count by 34.2% since 2017.

Net income grew with a CAGR of 10.6% compared to 16.8% EPS CAGR.

As you can see, the long-term EPS CAGR is 58% higher.

The longer the buybacks are done, the more they compound.

2.5. Booking

Booking is an online travel agency that has very low capex requirements.

In 2025, its capex was only $322M, while its OCF was a huge $9.4B.

This leaves almost all of its cash available for buybacks and dividends, a perfect combo to be a share cannibal.

It has reduced its share count by 35.6% since 2017. This is especially impressive considering the disruption the COVID-19 pandemic caused to its business in 2020 and 2021.

Booking’s position is so stable that the company paused share repurchases only for a single year in 2021, returning with a $6.6B buyback in 2022.

Net income grew with a CAGR of 12.4% compared to 18.3% EPS CAGR.

As you can see, the long-term EPS CAGR is 48% higher.

Do you like this report?

Become a Paid Premium member to see dozens of exclusive pay-walled equity research reports such as this.

Curious about which stock I own?

Full portfolio with all holdings, cost basis, and unrealised gains is visible to Premium members. I do weekly updates and detailed monthly reviews. Premium members are also notified of all sales and purchases using the pay-walled Substack Chat.

The annual plan is available for 50% cheaper per month than the monthly plan.

Subscribe now

2.6. Domino’s Pizza

Domino’s is the world’s largest pizza chain that has steadily bought back its stock.

Its shares outstanding fell from 42.9M in 2017 to 33.5M in Q1 2026, a reduction of 21.9%.

Net income grew with a CAGR of 9.6% compared to 13.7% EPS CAGR.

The long-term EPS CAGR is 43% higher.

2.7. Adobe

While historically, Adobe has not been a share cannibal, the company is positioning itself to become one.

Adobe has aggressively repurchased shares to fight off investor fears about AI disrupting its seat-based corporate subscription business.

Its shares outstanding fell from 491.26M in 2017 to 406M in Q1 2026, a reduction of 17.4%.

This might seem modest compared to other cannibals, but this is set to change. In the last 12 months, Adobe has spent $10.5B on share repurchases and only $2B on SBC, for a net buyback of $8B.

A market cap of $100B implies that Adobe is set to reduce its share count by a net 6-8% per year. If the valuation and buyback cadence remain the same, Adobe will repurchase 20-30% of its outstanding shares in the next 3-4 years.

So far, net income has grown with a CAGR of 19.2% compared to 21.6% EPS CAGR.

The long-term EPS CAGR is just 13% higher, but that will change rapidly in the next 3 years.

If AI doesn’t disrupt the business, Adobe will be a massive compounder.

2.8. Oracle

Oracle has historically been cannibal that has reduced its share count significantly over the last 15 years.

However, that trend stopped in 2022 as the company focused on building its AI cloud business.

Despite that, the share count is still 30.5% smaller today than it was in 2017.

As a result, net income has grown with a CAGR of 6.4% compared to 10.9% EPS CAGR.

So, the long-term EPS CAGR is 70% higher.

2.9. Hilton

Hilton has transformed from an asset-heavy hotel builder and developer into a capital-light franchise model that generates huge excess cash flow.

For instance, in 2025, the company returned $3.3B to its shareholders through buybacks and dividends.

Hilton’s shares outstanding dropped from 317.42M in 2017 to 228.33M in Q1 2026, a reduction of 28%.

As a result, net income has grown with a CAGR of 4.4% compared to 8.6% EPS CAGR.

So, the long-term EPS CAGR is an amazing 95% higher. This demonstrates the power of thoughtful buybacks executed at favorable market prices.

Similar to Booking, Hilton used the COVID rebound to repurchase $1.6B and $2.3B in shares in 2022 and 2023.

2.10. American Express

American Express is a prime example of how share buybacks benefit long-term investors like Warren Buffett.

Berkshire Hathaway has held about 151M shares since 1994, which is roughly 21% ownership in the company.

Berkshire has not bought new shares in years, but its ownership share keeps increasing because American Express constantly buys back its own stock.

Amex shares outstanding fell from 857.1M in 2017 to 682MM in Q1 2026, a reduction of 20.4%.

In Q1 2026 alone, the company returned $2.3B to shareholders, including $1.7B in buybacks.

As a result, net income has grown with a CAGR of 18.6% compared to 22.5% EPS CAGR.

So, the long-term EPS CAGR is an 21% higher.

3. Failing Share Cannibal Examples

As I said earlier, buybacks alone do not guarantee compounding for investors.

Many companies buy back stock to hide structural decline.

These businesses show that stock repurchases are useless if the core model is eroding and management underinvests in new revenue sources.

Let’s look at 4 distinct examples:

• PayPal

• Charter

• Western Union

• General Motors

PayPal has spent billions on share repurchases, but this has failed to create long-term value.

Since 2020, the company has spent over $26B on buybacks.

Despite this, its stock has crashed by over 80% from its 2021 peak.

This collapse happened because PayPal’s core business is in decline. Their branded checkout growth has stalled, growing only 2% in Q1 2026 and 1% in Q4 2025. This branded checkout button is PayPal’s highest-margin product.

At the same time, the company is facing intense competition from Apple Pay, Stripe, and Adyen.

PayPal’s active accounts have slowed to 439M, and its ADJ operating margin fell from 20.7% to 18.4%.

Instead of investing in creating high-margin products, PayPal is growing in low-margin products like Venmo and unbranded processing, which cannot replace its lost high-margin revenue.

Buying back shares is useless when the core business model is eroding, and the company underinvests in building new revenue-generating products capable of replacing it.

Charter is another clear example of a failing cannibal.

The company does not pay dividends and focuses entirely on buying back its own stock.

Since 2017, Charter has spent close to $70B on buybacks.

Despite this, its stock has crashed by over 79% in the last 5 years.

This is because Charter is facing a terminal business decline.

Charter lost 120,000 customers in Q1 2026, worse than the 59,000 lost in Q1 2025, showing that its broadband cable subscriber losses are accelerating.

Its total revenue fell 1.0% to $13.6B in Q1 2026. This erosion is caused by fierce competition from fixed wireless providers like T-Mobile and AT&T, as well as new fiber networks.

To fight this competition, Charter is carrying out a massive capex cycle, spending $11.7B in 2025.

This heavy capex drains cash and reduces FCF. At the same time, Charter carries a very heavy debt load with a leverage ratio of over 4x EBITDA.

Spending billions in buybacks while carrying huge debt in a shrinking industry is plain stupid.

If subscriber losses continue, EBITDA will decline, and the leverage ratio will rise. This makes Charter’s buybacks a value trap rather than a wealth creator. It would have been more beneficial for the shareholders if Charter had invested these $70B to create new revenue-generating businesses.

Western Union has a long history of stock buybacks, but that doesn’t change the fact that the company is in structural decline.

Since 2016, the company has spent $3.3B on buybacks.

This has had no beneficial impact on its stock, which is down 66% in the last 5 years.

Its stock sits near multi-year lows, showing that buybacks don’t change business fundamentals.

Western Union’s business is in structural decline as its high-margin, cash-based retail remittance business is shrinking.

In 2025, its North American regional money-transfer revenues fell 11%.

This decline is driven by digital-first competitors like Wise, Remitly, and Revolut. These fintech rivals are growing at 30%+ and stealing market share by offering better service and lower fees. This has lowered Western Union’s average take rate and reduced pricing power.

Western Union is trying to build its own digital platform and stablecoin, but its digital growth is too slow to offset the retail drop.

Its digital platform also operates at much lower profit margins. Western Union’s ADJ operating margin dropped to 13% from 18-19% in Q1 2026. Operating cash flow fell 26% Y/Y in Q1 2026.

Repurchasing stock in a business facing structural decline and profit compression is a waste of capital, they should have invested these $3B in rebuilding its tech stack and releasing new neo-banking fintech products, expanding upmarket from being a simple money transmitter into a more developed fintech platform.

General Motors is a dying automotive dinosaur on its last legs that, instead of innovating, has focused heavily on outsourcing and share buybacks.

In 2025, the company repurchased $6B of stock. In Q1 2026, the board approved another $6B share repurchase authorization.

This aggressive buyback program helped lower its outstanding share count from 1.2B in 2023 to 904M in 2025.

Overall, since 2016, the company has spent close to $34B on share repurchases.

This is madness considering that General Motors is facing massive structural issues.

Its core internal combustion business is eroding under a highly challenging and chaotic EV transition. In late 2025, General Motors took massive impairment charges to shift its Orion assembly plant back from EV to ICE production.

It also discontinued its BrightDrop electric van and wrote off several other EV assets.

Meanwhile, its automotive OCF dropped 21.7% to $18.733B in 2025. Spending billions of dollars on share buybacks while cutting EV platforms and facing declining cash flows is a classic sign of artificial EPS boosting.

Tesla and the Chinese automakers are coming for General Motors.

The company is terribly run with incompetent leadership. If the US automotive market were open to Chinese EVs, GM would be bankrupt already. They are cheaper and better than anything GM can offer.

However, this is only a question of time, as eventually the transition will happen and GM won’t be part of it.

Leave a comment

Become a Premium Member!

Here is what my Premium Members can expect:

Portfolio Review - Each month, I will present the portfolio performance and discuss my stock watchlist and my best ideas.

• Recent developments.

• Unwarranted pullbacks.

• Insider activity.

• Potential catalysts.

Deep Dives – 8,000+ word detailed analysis of a company, delivered in 3 Parts.

• Part 1 – Brief History of the company and its Business Model.

• Part 2 – Management, Moats, Competitors, and Risks.

• Part 3 – Opportunities, Financial Analysis, and a Valuation Model.

You can expect a comprehensive research report that is educational, interesting, and provides actionable insights!

To see what you can expect, read my Palantir Deep Dive!

Members of the Premium service get access to my library of 13 Deep Dives and to all future Deep Dives, which will be released on semi-monthly basis.

Investment Cases – A short, concise report with actionable insights.

This report is about the size of a single part of a Deep Dive.

• Focused Investment Thesis

• Main drivers of the Bull Case

• Valuation Model

To see what you can expect, read my Oscar Health Investment Case!

Earnings Reviews and Updates – For companies that are of great interest to me and my readers, I will provide regular quarterly or semi-annual updates after earnings reports.

• Financial performance

• Business Update

• New developments

• Updated Valuation Model

To see what you can expect, read my Google Q2 2025 Earnings Review!

Equity Research Report List

1. SpaceX

2. Micron

3. Nebius

4. Meridian Holdings

5. Weight Watchers

6. Lumentum

7. Sivers Semiconductors

8. Klarna

9. Ondas

10. Cipher Digital

11. SanDisk

12. Novo Nordisk

13. Sofi Update

14. UiPath

15. Mercado Libre Update

16. Nu Update

17. QXO

18. ASTS

19. Pagaya

20. IREN

21. Rubrik

22. Oscar Update

23. Arteche

24. Nebius Update

25. WhiteFiber

26. Harrow Health

27. Root

28. Zeta Global

29. SentinelOne

30. Golden Matrix Group

31. ASML

32. AMD

33. UnitedHealth Group

34. AppLovin

35. Adyen

36. TransMedics

37. Evolution

38. Nebius

39. Oscar Health

40. SoFi

41. Dlocal

42. Sea Limited

43. Microsoft

44. TSMC

45. Hims and Hers

46. Meta

47. Nu Bank

48. High Tide

49. Tesla

50. Grab

51. Google

52. Amazon

53. Hello Fresh

54. Mercado Libre

55. Celsius

56. Robinhood

57. Palantir

58. Ferrari

You can follow me on Social Media below:

X(Twitter): TheRayMyers

Threads: @global_equity_briefing

LinkedIn: TheRayMyers

Disclaimer: Global Equity Briefing by Ray Myers

The information provided in the “Global Equity Briefing” newsletter is for informational purposes only and does not constitute financial advice, investment recommendations, or an offer or solicitation to buy or sell any securities. Ray Myers, as the author, is not a registered financial advisor, and readers should consult with their own financial advisors before making any investment decisions.

The content presented in this newsletter is based on publicly available information and sources believed to be reliable. However, Ray Myers does not guarantee the accuracy, completeness, or timeliness of the information provided. The author assumes no responsibility or liability for any errors or omissions in the content or for any actions taken in reliance on the information presented.

Readers should be aware that investing involves risks, and past performance is not indicative of future results. The author may or may not hold positions in the companies mentioned in the “Global Equity Briefing” report. Any investment decisions made based on the information in this newsletter are at the sole discretion of the reader, and they assume full responsibility for their own investment activities.

I am Ray, a passionate investor and equity analyst. And today I am covering SpaceX.

Elon Musk is, without a doubt, the most successful and famous internet entrepreneur of the 2000s.

He made hundreds of millions during the internet boom when he sold PayPal. He didn’t retire to living a lavish life, as many dot-com founders who sold did.

Instead, he chose to invest that money into revolutionary and experimental businesses.

Tesla became a highly successful electric vehicle, renewable energy, autonomous driving, and robotics company with revenues nearing $100B and a market cap of $1.3T.

For a long time, Tesla was Elon’s most successful business, but that could change with the upcoming IPO of SpaceX.

SpaceX is the developer of rockets and satellites, responsible for handling the vast majority of payloads going into orbit.

The space economy is booming, as the world looks to use the vast resources of the galaxy to deliver services that previously were only seen in science fiction movies.

For many years, Elon fans were eagerly waiting and hoping for a SpaceX IPO so they could invest in Elon’s dreams of the stars.

That is finally happening, as the company plans to list its shares on the Nasdaq under the ticker symbol SPCX on June 12, 2026.

The offering aims to raise roughly $75B at a $2T valuation.

However, this is not the SpaceX of old that is going public, as during the last few years, Elon has transformed the company from largely a space technology business into a conglomerate consisting of:

• Space

• Satellite Internet

• AI

• Social Media

The company is justifying this valuation by citing a mind-boggling $28.5T global total addressable market! This is looking like the IPO of the century.

However, does it make sense for investors to enter at this valuation?

What is SpaceX’s actual business?

How are individual segments performing?

How much money does SpaceX actually make?

Is the balance sheet strong?

What are the opportunities for growth?

In this SpaceX Deep Dive, I will answer all of these questions.

Let’s begin.

1. Brief History

2. SpaceX

3. Starlink

4. X/Twitter

5. xAI

6. Competitors

7. Risks

8. Finances

9. Valuation

10. Opportunities

11. Conclusion

1. Brief History

Elon Musk has been a famous billionaire for seemingly forever, but that wasn’t always the case.

He made his early fortune through PayPal.

The 2002 $1.5B sale of PayPal generated hundreds of millions of dollars for Elon, providing the seed capital for his subsequent ventures.

He founded SpaceX in 2002 with the goal of building reusable rocket systems to make life multiplanetary.

Simultaneously, he financed Tesla, an electric vehicle pioneer, serving as its chief executive officer and driving its valuation to massive levels.

In October 2022, Musk purchased Twitter for $44B.

Next, seeking to compete with OpenAI and build a “truth-seeking” alternative to existing AI models, he founded xAI in March 2023.

For years, these businesses operated separately until they were merged into a single mega company.

In March, 2025, Elon decided to consolidate its AI and social media businesses with XAI, acquiring X in an all-stock deal valuing X at $33B.

Next, just a few months ago, in February 2026, SpaceX acquired xAI in an all-stock deal valuing SpaceX at $1T and xAI at $250B, for a combined total of $1.25T.

This created a vertically integrated company combining rocketry, satellite internet, social media, and AI.

2. SpaceX

The space and rocket division of SpaceX is reported as the Space segment in the S-1 filing.

It handles the design, manufacturing, and launch of rockets for commercial and government clients.

In 2025, the Space segment generated $4.1B in revenue, a modest 7.6% Y/Y increase.

However, the segment posted an operating loss of $657M for the full year of 2025, compared to the $21M profit generated in 2024.

The business is very R&D intensive, with $3B, an increase of 63.7% Y/Y, about 74% of revenues spent on this category.

This unprofitability continued in Q1 2026, with the Space segment generating $619M in revenue but reporting an operating loss of $662M.

This operating loss is driven by high investment in Starship and launch pads, showing that the launch business is highly capital-intensive and relies on Starlink to provide cash.

Furthermore, this business is not very stable and predictable, as revenues declined 28.4% Y/Y.

Growth is highly dependent on launch cadence and new government and private contracts.

Subscribe now

2.1. Reusability

SpaceX has reshaped the economics of space by demonstrating that orbital boosters can be used multiple times.

The Falcon 9 is a 2-stage rocket where the 1st stage booster can return and land on ground pads or drone ships in the ocean.

This allows SpaceX to fly the booster multiple times, reducing the cost of space launches.

Meanwhile, Starship represents the next-generation, fully reusable heavy-lift platform designed to transport over 100 tons of cargo into orbit

It consists of the Super Heavy booster (powered by 33 engines) and the Starship upper stage (powered by 6 Raptor engines, including 3 standard and 3 vacuum engines). Starship’s 10th flight test took place on August 26, 2025.

The Starship program is highly capital-intensive, costing approximately $4M per day to develop, with total R&D costs expected to reach $10B.

SpaceX spent approximately $3B on Starship development in 2025 alone, being a massive drag on cash flow.

This program is extremely important for the company because full reusability is required to make the economics of lunar and Mars cargo flights commercially viable at scale.

We have yet to see a successful launch.

2.2. Launch Services

SpaceX holds a near-monopoly on Western commercial launches, accounting for over 80% to 90% of global payloads launched annually.

In 2025, the company completed 170 launches, an increase of 23.2% Y/Y and 73.4% over 2 years.

The 2026 launch cadence has been steady, having 40 launches in Q1 2026 compared to 38 last year.

While the number of launches grew by 23.2% in 2025, the mass of those launches grew by 30.3%, to 2,213 tons.

The same trend continued in Q1 2026, with mass growing by 23.6%, despite just 5.3% launch growth,

This indicates that SpaceX is able to launch more satellites into orbit with each launch, thanks to its larger Falcon Heavy rocket.

This is reducing the cost per launch and further solidifying SpaceX’s position as the leading space launch company.

A major customer for these services is the US federal government. SpaceX has secured over $6B in contracts from NASA, the US Department of War, and other agencies over the last 5 years.

In 2025, US government contracts accounted for 20% of SpaceX’s total revenue.

2.3. Moon and Mars

SpaceX has become the most important hardware provider for NASA’s Artemis program.

In April 2021, NASA awarded SpaceX a $2.89B contract to build Starship HLS, a lunar lander variant of Starship designed to land astronauts on the Moon.

To reach the Moon, Starship HLS must refuel in space.

This is an insanely complicated engineering problem that requires a propellant depot in Earth orbit and roughly 10 Starship tanker flights to carry propellant. The Artemis III mission has been delayed to late 2027, and the actual crewed landing has been pushed to Artemis IV in 2028.

For Mars, Musk intends to build a permanent human colony with at least 1M residents.

The S-1 details an incentive package where Musk receives 15 tranches of 66,666,665 shares, amounting to 1B shares, worth roughly $600B at today’s valuation.

These shares only vest if SpaceX reaches a market valuation of $7.5T and establishes a self-sustaining Mars colony of 1M residents while Musk remains employed by the company.

Very ambitious to say the least.

3. Starlink

Starlink is the satellite telecommunications division of SpaceX!

Historically, space-based internet was slow and very expensive, because there were few satellites that were very far from the Earth.

To fix this, the company is utilizing a massive low Earth orbit satellite constellation of thousands of satellites that deliver high-speed, low-latency broadband internet globally.

Read more

While the S&P 500 has largely recovered, many promising businesses are still available at a discount.

And today I am buying 2 such businesses.

1 of the stocks I bought today is an extremely cheap small-cap in a very unique situation.

This is a turnaround play on a company that’s going through a painful transition. If my thesis is correct, this transition will end this year, enabling the company’s stock to rise rapidly. In such a case, the stock could be 2-3x just this year.

If my thesis is correct, and this company successfully turns around its business, this is a 5-10x stock in the next 5 years.

The other stock I am buying is a less risky bet on an undervalued, fast-growing, yet very misunderstood company that is hated by the market.

If my thesis is correct, this stock could 2-3x in the next 2-3 years as the market realizes its mistake.

To raise capital for this trade, I sold a stock that was 4.6% of my portfolio.

Normally, I prefer to hold stocks for longer, but I believe these 2 other stocks have more upside in the next 1-2 years.

Read more

I am Ray, a passionate investor and equity analyst. And today I am covering IREN.

By now, everyone and their grandmother knows that the world is experiencing a once-in-a-century AI-driven technological revolution. However, if we want AI to be able to do more than just create silly cat videos, it needs to be properly trained on lots of high-quality data.

Furthermore, for this data to be processed, the world needs to build more data centers than it has ever built before!

Researchers at McKinsey forecast that by 2030, cumulative data center capex will reach $7T and data center capacity will almost triple to 219 GW!

To meet this insatiable demand for computing power, capacity will come from clever and previously overlooked places. And increasingly, a trend has emerged of bitcoin mining companies turning into AI data center businesses.

For years, aspiring crypto entrepreneurs built ever larger and more sophisticated bitcoin mining operations. These bitcoin farms went from small basement servers to dedicated data centers with thousands of GPUs, custom-designed cooling equipment, and access to cheap energy.

It so happens that an AI data center also needs thousands of GPUs, cooling, and cheap energy!

Thus, many of these clever crypto mining entrepreneurs are taking their crypto mines and retrofitting them into AI data centers.

One of these crypto miners turned AI data center operators is Iren.

Thanks to strong AI-driven demand for their services, Iren’s stock has risen 490% in the last 12 months.

Furthermore, just a few weeks ago, the company signed a 5-year $3.4B AI Cloud customer deal and a potential $2.1B investment deal with Nvidia. This is a transformational agreement that the company aims to use as a stepping stone towards the AI big leagues.

In this report, I will look at their business model, data centers under construction, capex requirements, valuation, and finish off with a 2030 valuation model.

Let’s begin.

1. Business Model

2. Mirantis Acquisition

3. Data Centers

4. Customer Contracts

5. Competitors

6. Risks

7. Opportunities

8. Finances

9. Valuation

10. Valuation Model

11. Conclusion

1. Business Model

The company was founded in 2018 by Australian brothers Daniel and William Roberts with the idea to build efficient, high-performance bitcoin mining data centers that would incur lower operating costs than competitors.

However, after the AI boom began, it became apparent that AI data centers presented a significantly larger opportunity, so the company changed direction.

1.1. Shift from Bitcoin

As of calendar year Q3 2025, Iren had an installed Bitcoin mining capacity of 50EH/s, with plans to increase that to 52EH/s.

However, these plans for further expansion were cancelled to focus on the AI business.

Bitcoin miners use EH/s to measure mining capacity, which stands for exahash per second. The Bitcoin mining process works by performing complicated computations that solve digital cryptographic puzzles. These steps are needed to validate Bitcoin transactions and ensure the security of the network.

As a reward for this task, miners are given newly minted Bitcoins!

With this capacity, Iren was able to mine 2,039 Bitcoins in Q3 2025, generating $233M in revenues!

That was the peak for the bitcoin business.

Since Q3 2025, bitcoin revenues have fallen by 52% as Iren retrofits Bitcoin mining locations in Canada into AI data centers.

In calendar year 2026, Iren stopped disclosing Bitcoin mining hash rate, the number of bitcoins mined, and the cost per bitcoin.

This is a transitioning business, however, it still gives us important insights into the company’s expertise.

In real mineral mining, the type and purity of the end mineral determine the final price. However, in crypto mining, there is no differentiation, as a Bitcoin is a Bitcoin, Ethereum is Ethereum, etc. As there is no differentiation in the final product, all miners sell their bitcoin for the same price.

This means that costs ultimately determine a miner’s profitability!

That is why crypto miners have a lot of experience in not only finding the location with the cheapest electricity but also designing energy-efficient data centers. A key metric used to measure the efficiency of a Bitcoin miner is joules per terahash (J/TH), which essentially shows how much electricity a miner uses to complete Bitcoin mining puzzles.

As we can see in the table above, Iren’s data centers have an efficiency of 15 J/TH, which compares quite favorably to its bitcoin mining peers.

In Bitcoin mining, they use J/TH to measure the energy efficiency, but in data centers, PUE is used.

PUE stands for power usage effectiveness and is calculated by dividing the total capacity of a facility by the power actually used by IT equipment. A low PUE indicates that a smaller share of the total energy capacity is dedicated to overhead tasks, such as lighting, cooling, and others.

Essentially, a data center with low PUE uses a larger share of its electricity on revenue-generating GPUs.

Iren’s data centers are some of the most efficient in the business, with a PUE of 1.1. This means that for each KW used by IT equipment, only 0.1KW is used by the overhead. This compares favorably to other operators whose PUE ranges from 1.2 to 1.8.

This clearly shows that the company has the skills to build some of the most efficient data centers in the world!

1.2. AI Cloud vs Bare Metal vs Colocation

AI requires fundamentally different infrastructure than general-purpose cloud infrastructure.

Simply put, AI data centers use a higher number of powerful GPUs that “eat” a lot of electricity, requiring meaningful grid upgrades. Additionally, these GPUs generate more heat than a conventional data center, demanding improved cooling systems. This means that, unfortunately, using existing general-purpose data centers for AI workloads is just not economically viable.

This is why we are witnessing this extreme data center buildout boom!

However, not all AI data center companies are the same, as there are 3 key ways in which they earn revenue.

• Colocation

• Bare Metal

• AI Cloud

In a colocation model, a company such as Iren builds the data center, secures energy, and equips it with cooling, networking, and other equipment. Then they secure a customer who signs a long-term lease agreement of 5+ years and brings their own GPUs to the facility. This model has the lowest value added and thus has the worst economics, generating the lowest price per MW.

Meanwhile, in a bare metal model, the owner of the data center is responsible for fully equipping the facility with all equipment, including GPUs. It then finds a customer who negotiates access to a pre-agreed number of GPUs for a set per-hour rental price. Bare metal customers use their own software to run AI workloads on these GPUs. As the data center operator provides the GPUs, the bare metal model has a higher value added than colocation, and thus has better economics, generating a higher price per MW.

However, the most profitable AI data center model is the AI cloud provider!

In this model, the data center operator provides advanced AI training and inference software. Purpose-built AI training clusters, model hosting, managed inference services, AI APIs, storage, networking, developer support, and more. An AI cloud doesn’t simply rent out space, like colocation, or rent out GPUs like in bare metal. An AI cloud is essentially an AI development, optimization, inference, and deployment partner for AI start-ups and large enterprises.

As revenues are software-driven, this model has the highest value added and thus generates the highest price per MW of capacity!

Iren operates using the colocation and bare metal model, with some early signs that it is looking to add software offerings to its platform.

Each of the operating models has its own advantages and disadvantages.

In colocation, the data center operator is essentially a landlord, and such a model can be highly profitable if operated well. The key advantage is that it requires less capital to start, is easier to operate, and generates a more stable and predictable rental income from a small number of key tenants.

A bare metal is similar to colocation. But, if we use the analogy of a rental apartment, colocation is comparable to renting an apartment without any furniture, whilst bare metal is a fully furnished apartment. Meanwhile, an AI cloud could be comparable to a fully serviced apartment, with furniture, maids, a personal chef, nanny, tutor, and chauffeur.

An AI cloud might have the highest margins in theory, but it is also the most difficult model to operate. Companies must employ highly skilled software engineers to build the digital environment that AI start-ups need. This requires a lot of upfront capital and years of research and development. Furthermore, AI clouds must have big sales and customer management teams that manage a large number of small customers.

Most importantly, this area of the industry is highly competitive, filled with highly skilled start-ups such as Nebius, and it is led by hyperscaler clouds with hundreds of billions of dollars at their disposal, such as Google, Amazon, and Microsoft.

Historically, Iren stated that they don’t wish to compete with Hyperscalers on software. Their core competency is not AI software, but

1. Finding locations with cheap energy.

2. Building energy-efficient data centers.

However, with its recent acquisition of Mirantis, Iren seems to have changed its mind.

2. Mirantis

On May 5, 2026, Iren announced that it had agreed to buy a company called Mirantis for about $625M in an all-stock deal.

Mirantis is an expert in cloud software and Kubernetes, which is a tool used to manage and organize 1,000s GPUs working together.

This acquisition is designed to give Iren the first software tool it needs to become a full-stack provider of AI services.

This is a divergence from their previous pure bare metal strategy. A reminder that in bare metal, the client comes with their own software stack and only pays for the usage of the equipment.

This has lower margins than AI cloud software services, so with this move, Iren is moving upmarket, with the goal of increasing margins and revenues from its existing capacity.

The main goal of buying Mirantis is to make it easier for customers to use the massive power of Iren’s GPUs.

In the AI computing industry, there is often a gap where companies have the hardware but struggle to set up the software correctly to use all the GPU power.

Mirantis helps bridge this gap with its technical expertise and its AI platform, which manages AI workloads across different types of servers and environments.

Mirantis will operate as a separate subsidiary, but its software will be integrated into Iren’s AI Cloud platform. This will help the company serve a wider range of customers who need more complex AI setups.

By providing a larger share of the software layer, Iren hopes to make its relationships with customers stickier, driving long-term retention.

3. Data Centers

As of calendar year Q1 2026, Iren has 7 data center projects under development with a total contracted capacity of 5GW!

Let’s expand on them:

3.1. British Columbia, Canada

The company is currently working to upgrade the British Columbia sites to support AI workloads instead of bitcoin mining.

As a result of this transition to AI, Iren took a $140.4M impairment charge in Q1 2025.

At the Prince George site, which has 50MW of power, Iren has already delivered the necessary GPUs and is in the process of setting them up. Once this is done, the entire site will be dedicated to AI.

The Mackenzie site has 80MW of capacity that is ready for new chips to be installed in the second half of 2026.

At Canal Flats, the company plans to change all 30 MW of its power over to AI services.

This work is happening while the provincial government is making new rules for how data centers use power.

In 2025 and 2026, the province is setting limits on power for AI and requiring companies to compete for new connections. Iren has said it will keep working and growing in the region by following these new rules.

3.1. Spain

In May 2026, Iren entered the European market by buying data center developer Nostrum Group.

The company paid EUR 165M, with 65% in cash and 35% in stock.

This deal gave Iren 490MW of power capacity in Spain that is already contracted to the grid. This is not a single large data center, but rather a collection of smaller data centers.

Spain is highly attractive for AI data centers due to its abundant and low-cost renewable energy, particularly solar and wind.

Its strategic geographic location offers excellent undersea cable connectivity to Africa, the Americas, and the rest of Europe, turning the country into a vital global data hub. Additionally, Spain boasts a massive pipeline of secured power capacity and favorable government initiatives that streamline the development of large-scale infrastructure.

Depending on how fast the construction goes, these projects likely will come online in 2028.

3.2. Childress, Texas

The Childress site in Texas is the biggest focus for the company right now.

The Horizon 1-4 project is a key part of the current build-out at the site. This project involves 300 MW of data center space that uses liquid cooling.

The company is currently commissioning the GPUs for Horizon 1, which will be handed over to Microsoft for its use in Q3 2026.

The other three phases, Horizons 2 through 4, are on track to be finished by the end of 2026.

This is a major milestone because it shows the company can build and launch a large-scale, high-tech site for a major customer on a very tight schedule.

Construction is also moving forward on two more phases, Horizons 5 and 6, which are part of the plan to reach over 1.2GW of capacity by 2027.

3.3. Oklahoma

Iren is developing a new 1.6GW AI Data Center in Oklahoma!

At $10M ARR, this facility has the potential to generate $12-14B in annual revenues, depending on PUE. And Iren plans to begin energizing it in 2028, with first revenues likely coming in 2029-30.

This certainly means that Iren will issue more convertible notes to finance the development of this data center.

3.4. Sweetwater, Texas

On May 1, 2026, Iren reached a crucial goal at its Sweetwater site in Texas by energizing the first substation.

This substation has a massive capacity of 1.4GW and is now connected to the Texas ERCOT power grid.

This connection is a huge advantage because getting power from the grid is often the biggest delay in building a data center.

The whole Sweetwater campus is designed to eventually support 2GW of total power.

The first phase of building the actual data halls is underway, and this site will be a major part of the company’s growth through 2027 and 2028.

By finishing the substation first, the company can now build and turn on new data halls one after another as they are finished.

4. Customer Contracts

Read more

I am Ray, a passionate investor and equity analyst. And today I am covering Dlocal.

On Thursday, Dlocal released its Q1 2026 results, which showed strong payment volume and top-line growth but weaker profitability.

The market continues to pay too much attention to the short term, missing the forest for the trees, as the stock fell 8% and is now down 20% in the past month.

The market had such a negative reaction largely because of two reasons:

1. Missing earnings estimate.

2. Falling take rate.

Analysts forecast that ADJ EBITDA will come at $72M, but came significantly lower at $60.5M.

At the same time, DLocal continues to decrease as the company attracts payment volumes with structurally lower take rates.

Revenue take-rate decreased from 2.67% in Q1 2025 to 2.39% this quarter, while gross profit take-rate decreased from 1.05% in to 0.84%. While certainly a negative development, this is a clear overreaction to a metric that is a bit volatile and ignores the fact that these new volumes are profitable, despite the lower take rate.

In my opinion, the same as with Nu, Dlocal’s stock price is suffering from the FOMO trade.

Sell quality to buy AI stocks!

As a result, Dlocal trades for a FWD P/E of 11, despite strong fundamentals and fast profitable growth.

In this report, I will look at the results of the quarter and build a new valuation model that reflects the upside potential I believe Dlocal has.

Let’s begin.

1. Financial Summary

2. Payment Volumes

3. Take-Rate

4. Retention

5. Valuation

6. Valuation Model

7. Conclusion

1. Financial Summary

• Total Payment Volumes: $14.1B +73% Y/Y, +7% Q/Q.

• Revenue: $336M +55% Y/Y, -1% Q/Q. (Estimate of $331M)

• Gross profit: $119M +40% Y/Y, +2% Q/Q.

• Operating profit: $53M +15% Y/Y, -16% Q/Q. (Estimate of $57M)

• Net Income: $42M -10% Y/Y, -25% Q/Q.

• EPS: $0.14, -10% Y/Y, -23% Q/Q, (Estimate of $0.17)

• ADJ FCF: $15M, -63% Y/Y, -77% Q/Q.

Dlocal delivered very strong top-line results, beating analyst expectations by $5M, around 1.5%.

And while analysts focus on the falling take rate, they completely ignore the operating leverage that is being built in the business!

While operating profit declined 16% Q/Q, it was because of a one-time $9.7M tax adjustments that relate to last year, if not for this tax impact, operating profit would have grown by 25% Y/Y, not 15%.

Furthermore, the profitability was also weaker due to Dlocal continuing to invest in the business.

Employee count increased by 15% Y/Y, and this had an impact on opex. The company expects to achieve higher operating leverage in the second half of the year.

Gross profit per employee now sits at $91K, flat compared to Q4 2025, but up 21% from $75K in Q1 2025.

Meanwhile, net income grew 11% Y/Y if ignoring these one-time tax adjustments.

But even without this adjustment, 11% net income growth is quite weak.

This is due to Dlocal being in a strong investment period, with operating expenses growing by 58% Y/Y to $66M. You would expect a scalable, high-quality tech business to grow operating expenses slower than revenues and gross profit.

That is not the case now, as Dlocal invests to build new products for its customers and grow in Asia and Africa.

This is also what contributed to the sell-off, as investors do not seem to believe the management when they promise lower operating expense growth and higher profits in the second half of the year.

1.1. Guidance

Dlocal left its 2026 guidance unchanged!

The company expects 50-60% TPV growth, 22.5-27.5% gross profit growth, and 27.5-32.5%

In this market, not increasing guidance is seen as a miss, so this likely contributed to the sell-off.

Subscribe now

2. Payment Volumes

For a payments fintech, payment volumes are the most important metric that investors must track. Simply put, the more money that goes through Dlocal’s rails, the more money the company keeps for itself.

In Q1 2026, Dlocal grew payment volumes by an exceptionally strong 73.4% Y/Y, reaching $14.1B!

As we can see in the graph above, it was the best growth rate in a while and the 8^th quarter in a row of accelerating payment volume growth.

Furthermore, as we can see in the picture below, payment volumes grew strongly across most of its core verticals!

E-commerce, the largest and most important vertical, decreased by 6% Q/Q due to seasonality, as Q4 is always the strongest due to holiday shopping. Comparing Q1 2026 to Q1 2024, we see that this segment has grown by 186% in 2 years.

This is an exceptionally strong growth that demonstrates the potential of this segment. Let’s remember that e-commerce penetration is still lower in emerging markets compared to developed ones, providing Dlocal with a strong opportunity to scale.

Another outstanding performance came from the travel vertical, which grew 38% Q/Q, driven by Dlocal’s many customers, including Booking.com, Trip.com, and Klook.

In the above picture, we see 2-year TPV volume growth for key verticals.

All are demonstrating strong growth, except for Advertising.

Growth would have been even better if not for slower growth in Africa and Asia.

Dlocal didn’t disclose the TPV mix per region, but judging by the revenue growth of 36% (vs 55% for total revenue growth), these regions are ramping slower than I would have liked. These regions should grow faster than Latam, as they have a much lower base.

However, despite this slower growth, Pedro Arnt believes there is a huge opportunity in Asia and that he believes Dlocal is still early.

“We think that our strategy and the key success factors that have led us to be very successful in Africa and Latin America actually are applicable to Asia as we cross-sell our growing Asian capabilities to our global merchants.

The second, I think, myth we’ve dispelled is that Asia necessarily was a lower take rate market for the enterprise segment. From what we’re seeing in our current Asian operations as it scales out, that doesn’t seem to be the case either.

I think you will see us increasingly build out capabilities in Asia, and ideally, Asia over a multi-year period becomes quite relevant to our overall P&L.

That really could be trajectory-changing, just because when you look at the size of the total market, Asia still dwarfs LATAM and Africa.

Encouraging initial steps.” Pedro Arnt, Dlocal CEO

This quarter’s pay-ins grew faster than pay-outs, changing the previous trend of pay-outs having a larger share in the mix.

Pay-outs are now 28% of TPV, a decrease from 30% in Q4 2025 and 33% in Q1 2025. This could be due to the weakness in the remittances vertical, which this quarter only grew 3% Q/Q. Similar to e-commerce, Dlocal explained this by seasonality.

As Dlocal has barely scratched the surface in the remittances vertical, and it is growing strongly in ride-hailing and delivery, the share of pay-outs will likely continue increasing in the future. This will have a negative effect on the take rate. These activities have lower fees, but we will discuss that in the take-rate section.

Share

3. Take-Rate

Dlocal defines its take rate as gross profit divided by total payment volumes.

In Q1 2026, Dlocal reported a take rate of 0.84%, a significant decrease from 0.88% in the prior quarter and 1.05% in Q1 2025!

As we can see in the graph above, there is a clear downtrend of falling take rates, with the take rate more than halving in 4 years.

Every quarter, the street panics over the take rate, and this quarter is a perfect example.

The thought process is very simple, if Dlocal keeps less money for itself from every dollar it processes, its growth suffers.

However, this is a very simplistic thinking that ignores all other factors that also drive earnings growth.

Essentially, 3 factors largely determine a take rate, and each of them might have a larger effect in a particular quarter.

First, the financial terms under which the company has agreed to process payments for a merchant. A large customer renegotiating financial terms could have an outsized impact on the take rate.

Second, transaction processing costs. Dlocal operates a payment aggregator business where it handles the front-facing relationship with a merchant, whilst it integrates services from other payment acquirers, merchant banks, FX services, and more. Sometimes they earn higher FX fees, allowing the take rate to increase, whilst at other times the company incurs higher processing costs, reducing the take rate.

Third, payment volume mix shift. Fewer credit cards, more local payment methods, fewer pay-ins, more pay-outs.

Dlocal simplifies the payout processes for e-commerce, ride-hailing, and travel companies. However, these transactions are less complex and usually happen in batches, thus Dlocal charges lower fees to merchants for handling them.

This quarter, we saw a mix of all 3.

I really love the above chart from Dlocal, where the company easily bridges the difference in take rate from Q4 2025 to Q1 2026, explaining how Dlocal lost 4 basis points in take rate. I would love all payment companies to disclose so much.

As we can see, a 1.39 basis point gain in take rate was due to a larger share of pay-ins. Pay-outs are 28% of TPV, and as their share increases, they will continue putting downward pressure on the take-rate. This is largely due to seasonality, and I expect the share of pay-outs to put more downward pressure on the take-rate in the next few quarters.

Another 1.22 basis point gain came from a higher share of SaaS and other higher take-rate verticals in the TPV mix. 1.49 basis points were gained from processing more transactions with higher take rates, and 2.87 basis points came from other reasons.

This is where take rate gains end, as 0.9 were lost due to a higher share of local vs cross-border payments. Cross-border payments have higher take-rates.

1.58 basis points were lost due to lower volumes in smaller frontier markets that have higher take rates, such as Bolivia, Pakistan, and South Africa. Another 2.45 basis points were lost due to lower FX spreads in Egypt and elsewhere.

However, most of the 4-basis point take rate decline was caused by seasonality and “volume tier upgrades”.

Simply put, Dlocal is likely giving meaningful discounts to its largest merchants, such as Uber, Netflix, and Didi, to retain them and keep driving volume growth. This is a standard strategy in payments, because without such discounts, some merchants would move volumes. It is cheaper to give a large merchant a discount than to lose it and then try to acquire payment volumes from another merchant to replace the lost volumes.

The bigger picture is that payment volumes are growing incredibly quickly across verticals and geographies!

Dlocal might keep a smaller share of each dollar processed because of the nature of and economics of these money flows changes, but that doesn’t matter, as each dollar they process is still quite profitable.

As we can see in the graph above, despite the take rate falling by 66% from 2019 to 2026 (blue line), Dlocal’s EPS grew by 1,000% (purple line).

I know which metric I care about!

4. Retention

Dlocal retention is simply incredible!

Revenue retention was 152%.

This means that had Dlocal not found new customers, revenue would have grown by 52%. So only 3% of the 55% revenue growth came from new merchants. That is certainly low, but this is because of Dlocal’s land and expansion strategy. In payments, customers always try out processing providers in a few use cases before giving the processor more volume.

The revenue share of the top 10 merchants has remained at around 62%. This means that despite strong revenue growth, other non-top 10 merchants grow fast enough for the concentration not to increase.

So, despite strong growth by the top merchants, concentration is not increasing!

This quarter, they didn’t disclose TPV retention, but it must be even higher, likely at 170%. This is due to falling take-rates.

Here are a few examples from Dlocal’s presentation that demonstrate why the retention is so high.

A ride-hailing merchant (likely Uber or Didi) that was acquired in 2016, went from a single country in 2016 to 14 in 2023 and 18 in 2025. TPV is up 5x in 3 years, and Q1 2026 volumes grew 250% Y/Y.

A SaaS merchant went from a single country in 2019 to 19 in 2030 to 40 in 2025, with TPV growing by 6x in 3 years and 110% in Q1 2026.

An e-commerce merchant went from a single country in 2023 to 16 in 2024 and 21 in 2025, with TPV growing by 94x in 3 years and 70% in Q1 2026.

Overall, there is a clear trend of accelerated growth, with newer cohorts expanding faster than earlier ones.

5. Valuation

As a result of Dlocal falling 20% in the past month, it now has a market cap of $3.3B.

I see a clear disconnect between the strong execution and the stock price!

The company now trades for a TTM P/E of 17, which is just an unbelievable entry point for such a strong growth company.

Analysts expect Dlocal to grow revenues by 40% and 27% in 2026 and 2027, leading to revenues being 89% higher in 2028 than in the last 12 months.

But most importantly, they expect Dlocal to meaningfully improve profitability, with EBIT growing 96% by 2028, and EPS 82%.

Furthermore, the company has a shareholder return plan, with about $300M to be spent on share repurchases, reducing share count by potentially 10%. Additionally, the company pays a dividend, which could end up being about 1.8% dividend yield.

Taking analyst growth expectations into account, the 2028 P/E is just 9.

Simply put, even if the multiple remains the same as it is today, Dlocal stock should increase by 82% in the next 2 and a half years!

However, I find analyst estimates to be on the low side, and there is a potential for the multiple to increase.

Let’s model that.

6. Valuation Model

Since I made the previous Dlocal valuation model back in November, I have increased my confidence in the business, thus, I will upgrade my assumptions.

First, with the pace of innovation continuing and Dlocal signing new partnerships continuously, I have increased my 2030 TPV estimate from $131B to $178B!

Due to the take rate performance of this quarter, I have reduced my revenue take rate to 1.9%. That would be a decrease of 30% from the 2.7% revenue take rate of Q3 2025.

That would lead to revenues of $3.3B in 2030, an increase from $2.6B in the previous model!

Next, I model the operating margin increasing from around 21% the company has today to 27% in 2030.

Tax and other expenses of 21%.

The result of these assumptions is a net income $697M, an increase from $561M in the previous model.

Buyback of 15% over the next 5 years.

With an exit multiple of 17, we get a $47.48 stock, an upside of 325% from today!

Discounting that back at 18% per year, we get a fair value per share of $20.75, implying that today’s price of $11.16 is 46% below fair value.

However, if the market assigns the company a better multiple of 25-30, the upside increases to 525-651%!

Dlocal is a clear opportunity that is underappreciated by the Mr. Market, but with time, strong execution will force Mr. Market to appreciate it.

7. Conclusion

The long-term thesis for DLocal is still extremely favorable.

The company is uniquely positioned to benefit from the digitization of emerging markets, driving massive adoption of digital services.

E-commerce, streaming services, ride-hailing, food delivery, SaaS, travel, and financial services are undergoing a once-in-a-generation shift!

With its comprehensive and improving platform, DLocal is taking an increasing slice of the emerging markets digital payments pie.

Thanks to the quality of their services, DLocal has built strong relationships with some of the world’s largest and fastest-growing digital companies. Just by servicing the likes of Amazon, Meta, and Google well and upselling new products, DLocal can grow immensely.

The current short-term take-rate concerns are overblown, and Dlocal is well-positioned to have a strong, thriving, and highly profitable business.

Dlocal has built a capital-light, high-margin business that trades for a reasonable valuation.

As the valuation model shows, with reasonable assumptions, investors could see 325% upside by 2030!

With more aggressive assumptions, investors could even experience 7x returns by 2030.

Furthermore, Dlocal’s share price has suffered from the FOMO trade, with investors selling high-quality assets to buy speculative AI stocks.

I don’t believe that this is a profitable trade. Selling quality companies at such depressed levels to buy hype stocks is a recipe for disaster.

I believe that this trade will end soon, as fintechs such as Nu, DLocal, Klarna, and Sofi are trading at unsustainably low valuations.

Companies with strong balance sheets, increasing TAM, and 30-40% topline growth don’t deserve to trade for sub-20 FWD P/E. Dlocal trading for a FWD P/E of 11, despite such strong growth and quality characteristics, is objectively unbelievable.

Eventually, the market will return to wanting profits, as it did in 2022 after the 2021 EV/SaaS bubble burst.

Leave a comment

Become a Premium Member!

Here is what my Premium Members can expect:

Portfolio Review - Each month, I will present the portfolio performance and discuss my stock watchlist and my best ideas.

• Recent developments.

• Unwarranted pullbacks.

• Insider activity.

• Potential catalysts.

Deep Dives – 8,000+ word detailed analysis of a company, delivered in 3 Parts.

• Part 1 – Brief History of the company and its Business Model.

• Part 2 – Management, Moats, Competitors, and Risks.

• Part 3 – Opportunities, Financial Analysis, and a Valuation Model.

You can expect a comprehensive research report that is educational, interesting, and provides actionable insights!

To see what you can expect, read my Palantir Deep Dive!

Members of the Premium service get access to my library of 12 Deep Dives and to all future Deep Dives, which will be released on semi-monthly basis.

Investment Cases – A short, concise report with actionable insights.

This report is about the size of a single part of a Deep Dive.

• Focused Investment Thesis

• Main drivers of the Bull Case

• Valuation Model

To see what you can expect, read my Oscar Health Investment Case!

Earnings Reviews and Updates – For companies that are of great interest to me and my readers, I will provide regular quarterly or semi-annual updates after earnings reports.

• Financial performance

• Business Update

• New developments

• Updated Valuation Model

To see what you can expect, read my Google Q2 2025 Earnings Review!

Equity Research Report List

1. Micron

2. Nebius

3. Meridian Holdings

4. Weight Watchers

5. Lumentum

6. Sivers Semiconductors

7. Klarna

8. Ondas

9. Cipher Digital

10. SanDisk

11. Novo Nordisk

12. Sofi Update

13. UiPath

14. Mercado Libre Update

15. Nu Update

16. QXO

17. ASTS

18. Pagaya

19. IREN

20. Rubrik

21. Oscar Update

22. Arteche

23. Nebius Update

24. WhiteFiber

25. Harrow Health

26. Root

27. Zeta Global

28. SentinelOne

29. Golden Matrix Group

30. ASML

31. AMD

32. UnitedHealth Group

33. AppLovin

34. Adyen

35. TransMedics

36. Evolution

37. Nebius

38. Oscar Health

39. SoFi

40. Dlocal

41. Sea Limited

42. Microsoft

43. TSMC

44. Hims and Hers

45. Meta

46. Nu Bank

47. High Tide

48. Tesla

49. Grab

50. Google

51. Amazon

52. Hello Fresh

53. Mercado Libre

54. Celsius

55. Robinhood

56. Palantir

57. Ferrari

You can follow me on Social Media below:

X(Twitter): TheRayMyers

Threads: @global_equity_briefing

LinkedIn: TheRayMyers

Disclaimer: Global Equity Briefing by Ray Myers

The information provided in the “Global Equity Briefing” newsletter is for informational purposes only and does not constitute financial advice, investment recommendations, or an offer or solicitation to buy or sell any securities. Ray Myers, as the author, is not a registered financial advisor, and readers should consult with their own financial advisors before making any investment decisions.

The content presented in this newsletter is based on publicly available information and sources believed to be reliable. However, Ray Myers does not guarantee the accuracy, completeness, or timeliness of the information provided. The author assumes no responsibility or liability for any errors or omissions in the content or for any actions taken in reliance on the information presented.

Readers should be aware that investing involves risks, and past performance is not indicative of future results. The author may or may not hold positions in the companies mentioned in the “Global Equity Briefing” report. Any investment decisions made based on the information in this newsletter are at the sole discretion of the reader, and they assume full responsibility for their own investment activities.

I am Ray, a passionate investor and equity analyst. And today I am covering Nu.

On Thursday, Nu Holdings reported earnings, and they were again excellent!

Yet the market continues to focus on short-term narratives instead of fundamentals and long-term execution.

Nu’s stock is now down 21% over the past month, despite accelerating revenue growth, strong profitability, and the prospects of US and global expansion.

It is clear that we are witnessing a FOMO trade in real time.

Sell quality to buy AI stocks!

As a result, Nu trades at just 14x FWD P/E.

I find this valuation to be extremely attractive for long-term-minded investors, especially considering Nu posted such strong financial results:

• Total revenues $5B +53%

• Net income $872M +56.5%

• EPS $0.18+ 63.6%, slightly missing analyst estimate of $0.19

• ARPU $15.9 +42%

• Cost per ARPU increased from $0.7 to $1 Y/Y.

• The total number of customers increased by 4.2M to 135.2M, +14%

Let’s look at the quarter in more detail to see what Wall Street liked and what they didn’t.

1. Customer Level Results

2. Brazil

3. Mexico

4. Colombia

5. Deposits and Loan Portfolio

6. Revenue

7. Profitability

8. Conclusion

1. Customer Level Results

Nu continues to gain customers at an incredible pace.

The number of customers grew by 14% Y/Y and 3.2% Q/Q to 136.2M, an increase of 4.2M over Q4 2025!

In the above graph you can see how the growth has decreased significantly in the 2 years, this is understandable. You can’t grow customers at 30% per year once you already have over 130M customers.

However, interestingly, the growth rate has plateaued in the last few quarters and appears not to be decreasing as fast as some expected. Simply put, Nu continues to acquire new customers, as the total addressable market is vast.

Moreover, 83% of these customers are monthly active customers, the same as last year. In fact, the activity rate has remained unchanged for 4 years, which is remarkable given the rapid growth.

One could expect a slight decrease in activity rate after expanding into two new countries and adding tens of millions of new customers.

Above we see Nu’s share of adult populations in its core markets as of Q4 2025. They didn’t include such a slide this quarter, but it is clear that these ratios have only increased.

• 62% of Adults in Brazil.

• 15% in Mexico.

• 11% in Colombia.

Overall, each active customer generated $15.9 monthly ARPU in Q1, an increase of 42% Y/Y, and 6% Q/Q!

This is incredible growth.

In the graph above, we can see a strong acceleration from Q1 2025, when the company posted a decrease in ARPU by 1.8%. This fall in ARPU caused quite a lot of panic in the Nu investor base, as ARPU growth is a key bull thesis for the company.

This was largely driven by Nu gaining a lot of new, and temporarily under-monetized customers in Mexico and Colombia!

This is what I said last year during the Q1 2025 review:

“As Nu develops Mexico and Colombia, ARPU is likely to increase later in the year, especially as Nu releases new products in Mexico after receiving a full banking license. “ Global Equity Briefing

This quarter, we saw the continuation of that trend as significant growth in both countries directly contributed to the recovery of the ARPU.

On an FX-neutral basis, ARPU actually grew 24% Y/Y, so the company benefited from a weaker USD.

Per the above graph, it takes Nu between 60-72 months to get a new customer to the $15.9 ARPU.

Understanding customer cohorts is incredibly important for Nu investors.

A customer who’s been with Nu for 8 years generates $30.8 in monthly revenues, which is double the average Nu customer, 6x the 1-year-old customer, and 4x the 2-year-old customer.

This is because Nu is very good at upselling customers new products.

A customer goes from using a bank account and a debit card in the first year, to a credit card, savings account, children’s account, investment account, insurance, and a personal loan in year 6.

In the last 3 years, Nu acquired 60M customers that are on average using a few products and generating ARPU of below $10, likely the average is around $6. In the next 5 years, this cohort will get millions of credit cards, put billions in the investment accounts, and borrow billions.

So, doing some back-of-the-napkin math, we get 60M people going from $6 ARPU to $30.8 in 5 years, that’s $17.9B in additional revenue.

That’s about what Nu made in the last 4 quarters. So just by upselling this cohort of users’ existing products, Nu can double its revenues in the next 5 years. We are not even talking about increasing revenues of year 5-8 cohorts, about releasing new products, entering new countries, or acquiring users in existing countries.

This company has incredible opportunities for growth!

Meanwhile, the cost to serve an active customer increased by 42.8% Y/Y, 19% on FXN basis. I will expand in the profitability section on the causes, but in short, Nu is investing and preparing for the next leg of growth.

Despite this increase, if we compare the cost to serve an active customer to ARPU, we see that after about 2 months, a customer is already at break-even in servicing costs!

If we assume a linear growth in ARPU from month 1 to 12, in the first year, an active customer will generate Nu $35.40 in revenue, but it would cost only $12 to service.

This leaves a lot of room for Nu to utilize marketing as organic customer growth slows down when they begin reaching maturity in certain markets.

There is significant potential to increase ARPU.

“I think a couple of our opportunities really on the revenue side, and as a reminder, our ARPAC is $15 today and our incumbent competitors are something like 40, so we have a significant opportunity to increase ARPAC, is around new cross-sale and new products that we can be delivering to the very significant consumer base that we have. I think everything around cross-sale, everything about using the data that we already have to offer new products and services, it’s a big opportunity, and AI is an enabler.” David Velez Nu Q4 2025 Earnings Call

Subscribe now

2. Brazil

Brazil, as Nu’s home and Latin America’s largest country by population and GDP, remains the most important market today.

As of Q1 2026, Nu has around 115.5M customers in Brazil, an increase of 10.4% Y/Y.

This quarter, Nu added 2.8M customers, an increase of 200K from the 2.6M it added in Q4 2025.

It is quite remarkable that they can add so many customers in a quarter, despite already having 62% of the adult population as customers.

At such a scale, it is becoming harder and harder to acquire new customers, as realistically, it would be crazy if the share of the adult population surpassed 70%. For a single bank to have 70% of the country as customers would be insane.

Thus, it is quite clear that the focus in the future will shift from user growth to user monetization.

Total revenue in Brazil grew 53.3% Y/Y, to $3.6B!

Please note that total revenue doesn’t match the sum of the geographic revenue, as some of the corporate HQ revenue is not allocated to a specific geography.

Unfortunately, as in the last quarter, Nu did not disclose detailed loan origination statistics for Brazil. My guess is that since expansion in Mexico and Colombia is going steadily, the company wants to emphasize its business as a portfolio of loans, rather than just Brazil.

However, they did disclose Brazil’s lending unit economics!

Brazil’s credit card portfolio achieved a return on assets of 8.9%. For context, global bank credit card portfolios generally have an ROA of 3-5% and Brazil’s 8-14%, making Nu’s 8.9% a really strong result.

Because the average duration is so low, just 3 months, Nu can turn over capital 4 times in a year. This enables Nu to generate 28.3% in net interest revenue, and 8.5% in fee and transaction revenue, for a total credit card margin of 36.8%. Accounting for funding costs and losses, we get to a net credit card margin of 19.9%.

Meanwhile, the unsecured lending portfolio generated a return on assets of 15.8%.

Nu was able to generate a 62.2% interest revenues, and after taking into account 12.2% in funding costs and 22% in losses, it had a net lending margin of 28%.

For anyone familiar with US or European interest rates, such returns seem absurd, but that is normal in Brazil, as the country deals with chronically high interest rates.

The Banco Central do Brasil’s interest rate is now at 14.5% compared to 2.15% in the EU and 3.75% in the US.

As you can see in the chart above, Brazil’s central bank rate has been above 8% for more than 4 years now.

When your asset yields are so high that you can absorb 16.6% credit card losses and 22.0% lending losses and still walk away with an 8.9% and 15.8% ROA, you are operating a highly resilient lending machine. Competitors globally simply do not have the spread environment, nor do domestic incumbents have the low cost-to-serve, to match these specific cohort returns.

This is also a key bear case for Nu: “What if interest rates go down, reducing lending spreads?”

Look at the chart above, from 2019 to 2022, interest rates were low, and Nu thrived, and they will thrive if that happens again.

Here is what will happen if interest rates go down:

1. Loan volumes will increase as cheaper loans increase demand. More loans mean more interest income, compensating somewhat for the lower spreads.

2. Lower defaults will reduce losses as cheaper loans lead to fewer customers unable to repay them.

3. Nu will increase fees, driving fee income. Nu fee income has been growing more slowly than net interest income for years. I will expand on this in the revenue section of this report.

4. More transaction income, as low interest rates lead to more transactions.

So lower interest rates would certainly be a headwind for net interest income, but Nu would find other means of filling in the gap, reducing the net loss from lower spreads.

Moving on to deposits, Brazil showed weak growth, attracting $500M in new deposits to reach a balance of $33.7B!

This was an increase of 38% Y/Y, but just 1.5% Q/Q, with net deposits decreasing by 62% compared to Q1 2025.

The management explained that slower deposit growth is due to seasonality, as Q1 is historically weaker. Another explanation was that Nu has a large enough deposit base to fund its loans. So they don’t believe there is a need to be more aggressive with their deposit interest rates to attract more deposits.

They can always increase deposit yields and run promotions if they see that additional capital is required to fund lending.

This quarter, the company revealed that they only have a 7% gross profit market share in Brazil, despite having over 62% of the adult population as customers.

A large share of this industry profit pool is generated by high-income customers, so Nu is working to attract these customers.

“Out of the high income, we now have about two out of every five high-income Brazilians are customers of Nubank. The customer base has grown by about 24% year-over-year ….. With Nu monthly credit card, kind of, volumes up 42% year-over-year. Assets are under custody by like 36%. We are seeing lots of traction there.” Guilherme Lago, Nu CFO

The Nu CFO said that the company is focused on attracting and serving what they describe as “Super Core” segment, which consists of people earning BRL 5-12K per month, about $1-2K. The number of customers in this category is up 24% Y/Y, compared to 10% overall growth in Brazil.

Share

3. Mexico

As of Q1 2026, Nu Mexico has 15M customers, around 15% of the adult population.

This is an increase of 36.5% Y/Y and 127% in just 2 years!

While I doubt that Nu can reach as high a penetration as Brazil’s 60%, I am confident there is still a lot of growth left in Mexico.

A reminder that Nu received a full Mexican banking license just last year. Previously, they operated under a non-bank financial institution license (SOFIPO), which is a special Mexican regulatory framework that allows FinTechs to provide some banking services, but with a lot of limitations.

A full banking license enables Nu to offer more services to its clients. This helps it to attract more customers and increase wallet share with existing customers.

Deposit balance fell to $5.9B, down 6.3% Q/Q, but still up 9% Y/Y.

Similar to Brazil, the leading explanations for slow deposit growth are seasonality and a focus on lower-cost capital funding means.

Next, revenue in Mexico grew by 96.5% to $289M.

The same as with Brazil revenue, this is an accounting and classification issue. In Q4 2024, a lot of revenues were classified as originating in Mexico and fewer from the HQ. Ignoring that spike, we see a clear trend line of revenue going up at a steady pace.

This quarter, Nu disclosed some new and interesting KPI’s for the Mexican business.

Nu revealed that Mexico is now at net income break-even, compared to a $37M loss in Q1 2025!

This is why Nu didn’t want to overpay for deposits, as then they would not have been profitable. Now that they are, deposits will likely grow in the second half of the year.

Most importantly, we see that the efficiency ratio fell to 42% from 120% in 2022.

Simply put, in 2022, it cost Nu $1.2 to generate $1 of revenue in Mexico, but today it costs $0.42.

ARPU is up to $12.4, double the $6.5 the company was making in 2022.

Compared to Brazil, Nu’s runaway in Mexico is even longer.

The company only has a 1% market share, despite having about 15% of the adult population as customers.

4. Colombia

In Q1 2026, Nu Colombia added 500K customers, reaching 4.7M.

This compares favorably to 400K added in Q4 2025.

This is an increase of 62.1% Y/Y and 422% in just 2 years!

Nu Colombia has already onboarded 11% of the adult population, quite impressive considering they only began operations 5 years ago.

This quarter, Nu Colombia reached $2.8B in deposits, an increase of 56% Y/Y and 12% Q/Q!

As we can see in the graph, this is a 10x growth in the deposit base in just a year and a half!

Nu didn’t disclose anything else about Colombia, as the business is still small.

I will be paying close attention to see how Colombia develops. So far, user and deposit growth have been stronger than in Mexico and Brazil at this age in the market.

5. Deposits and Loan Portfolio

Total Nu deposit balance grew 34% Y/Y to $42.4B.

On an FX-neutral basis, deposits grew 22% Y/Y.

Net deposits were $500M, which was an 81% Y/Y decrease and 84% Q/Q decrease.

As I already explained in the Brazil and Mexico section, deposit growth was low this quarter due to seasonality and focus on lower-cost sources of capital. If Nu needs more capital, they can increase deposit rates.

Nu cost of deposits is now at 88% of the interbank exchange rate, down from 90% in Q1 2025. Essentially, Nu is paying its depositors smaller yields than last year, helping the company lower interest costs, increasing net interest income.

Banks grow their deposit base for essentially only one reason, to issue more loans to generate interest income.

Total loan portfolio grew 54% Y/Y and 40% FXN to $37.2B.

53% FXN growth in unsecured loans was the main contributor to the loan portfolio growth. Meanwhile, the credit card loan share of the portfolio decreased from 68% in Q1 2025 to 65% this quarter.

While the credit card share is still huge and significant, Nu’s change in the portfolio mix displays Nu’s strategy to originate more diversified, higher-quality, longer-term loans.

This means that Nu is less reliant on revolving credit card balances that are paid off monthly and don’t generate interest.

If we look at Q/Q growth in secured loans, we see that it was 11%, from $2.7B to $3B. This is an improvement, as last quarter there was no growth due to the regulatory changes in Brazil. Last quarter, there was a 50% decline in new volumes for the secured unemployment fund loans, as the government made changes to regulations.

In Brazil, there is a fund that is designed to help workers in case of unemployment, but there is a way to use that fund as collateral and borrow against it. That was not the intention of the regulator, so they introduced new limitations, making such loans less attractive to borrowers.

Nevertheless, the quality of their loan portfolio remains stable.

The share of 15-to-90-day nonperforming loans increased from 4.1% to 5% Q/Q, and from 4.8% in Q4 2024.

90+ day NPL also increased from 6.4% in Q1 2025 to 6.5%.

The company explained that this increase in delinquency is largely seasonal, as Q1 always has more defaults, with people who overspent during the Q4 holidays not paying their loans.

Nu released this bridge to compare delinquency to Q4 2025.

From the 99 basis point increase, 65bps are seasonal, 4bps relate to different product mix, and 17bps are due to intentional risk expansion. Essentially, took more risk this quarter, but the company believes that additional risk will result in higher profits as additional interest will compensate for higher delinquencies.

6. Revenue

Total revenue grew 53% to $5B, driven by a 56.5% increase in interest income and a 34.4% increase in fee income!

However, net revenue (which I defined as Fees + Interest income – interest cost (NII) – provision for loan losses) grew slightly slower, 43.7% Y/Y to $2B.

This was a slight Q/Q decrease, but as you can see in the chart above, that is the usual seasonality, as net revenue also decreased from Q4 2024 to Q1 2025.

*(Some have questioned why I do it differently than Nu reports. I do this to for me and to make Nu bank more comparable to other banks. Net revenue is comparable to what Nu calls Gross profit)*

The main causes of this lower growth were a 41.6% increase in interest costs and a 76.5% increase in provision for loan losses.

Because of this strong increase in interest costs, the risk-adjusted net interest margin (risk-adjusted NIM) decreased from 10.5% in Q4 2025 to 9.5% this quarter, with net interest income (NII) growing 63.7% to $3B.

The interest costs growing slower than interest income (41.6% vs 56.5%) was largely due to Nu’s weak deposit growth this quarter.

Let’s remember that because of banking regulations, deposits from one country can’t be used for loans in another. This means that Nu must build a new capital base in each country.

Provision for loan losses increased more than the NII (76.4% vs 63.7%) because of a strong growth in the overall loan portfolio and a larger share of unsecured loans in the portfolio (27% vs 23%).

IFRS 9 requires banks to estimate and recognize future loan losses immediately when a loan is issued.

When a bank issues a lot of loans like Nu did, it books high loan loss provisions before it generates any meaningful interest income from these loans.

Additionally, unsecured loans default at a higher rate, thus they require a higher upfront loss provision.

“Credit Loss Allowance or CLA closed at $1.79 billion in the quarter, up 33% quarter-over-quarter (76.5% Y/Y) on an FX neutral basis, mostly driven by three very specific dynamics already mentioned by David Vélez. Number one, seasonality. Number two, portfolio growth. Number three, portfolio mix” Guilherme Lago, Nu CFO Nu Q1 2026 Earnings Call

An interesting point is that interest income grew faster than interest costs for the first time in 5 quarters.

This is what I said in my Nu Q4 2025 review:

“interest cost has grown faster than interest income for 4 quarters in a row. I think this trend could stabilise now, as the company builds out its deposit base.” Ray Myers, Global Equity Brefing, Nu Q4 2025 Earnings Review

This is exactly what we saw this quarter.

7. Profitability

In Q1 2026, Nu generated $872M in net income, up 56.5% and 41% FXN Y/Y!

The chart above shows that Nu is a king of profitability with no signs of slowing down. The net income margin increased Q/Q from 43.5% to 44%.

ROE, which basically measures a bank’s return from its net assets, increased from 27% in Q1 2025 to 29%.

However, Wall Street didn’t like the sequential decline, as ROE was 33% in Q4 2025. It seems investors don’t understand seasonality, as ROE also declined from Q4 2024 to Q1 2025.

Most banks can only dream of such a high ROE. This level of profitability can only be possible with incredible operational efficiency.

Despite the strong ROE, the street didn’t like the increase in the average cost per active customer.

As we already discussed earlier, the cost per active customer increased 19% Y/Y on FXN basis, but still remains at an incredible $1.

Another metric to monitor is the efficiency ratio!

It measures operating expenses as a share of net revenue, and as you can see in the graph above, it has been steadily going down from 21% in Q1 2025 to 18% today.

Nu is better at using technology to serve a large number of customers than possibly any other bank in the world!

Nu’s efficient operating model allows low prices and high profits. This is a potent combo for explosive global growth!

However, Nu did miss the street analyst’s EPS estimate of $0.19 by 1 cent.

This was due to Nu spending heavily to prepare for the next leg of growth.

• Customer support costs grew by 58.8% to $205M.

• SG&A grew by 69.8% to $492M.

• Marketing grew by 42.6% to $63M.

• Other expenses grew by 58.8% to $170M.

Nu is investing in AI to automate customer support and operations. They are putting AI at the center of the organization, and that is incredibly expensive. It will take some time for efficiencies to appear.

Let’s not forget that Nu just sponsored MLS team Inter Miami, where Messi plays.

They signed a 5-year deal that cost many millions, and it will take time for the company to recoup this investment once they enter the US market.

Overall, Nu delivered another incredible quarter of profit growth.

However, net margin will likely not grow as fast as before, due to the investments that the company must make to grow in the US. But I don’t care much about the few basis point difference in the margin, I care about the absolute dollar EPS growth.

I have high confidence in the Nu and this management team.

8. Conclusion

In conclusion, Nu delivered another strong quarter of strong profit growth, yet in this market, investors and algos are looking for the tinniest reason to sell a stock.

As I mentioned in the introduction, there is a FOMO trade happening right now in the market.

Sell quality, especially SaaS and fintech, and buy AI stocks!

The AI boom is real, and there are many stocks that are delivering real results and deserve the AI hype, but there are many that do not. Some stocks have run over 1,000% this year, and a lot of investors are selling their SaaS and fintech stocks that are down 20-40% to buy AI stocks up 100-500%.

I don’t believe that this is a profitable trade. Selling quality companies at such depressed levels to buy hype stocks is a recipe for disaster.

I believe that this trade will end soon, as fintechs such as Nu, DLocal, Klarna, and Sofi are trading at unsustainably low valuations.

Companies with strong balance sheets, increasing TAM, and 30-40% topline growth don’t deserve to trade for sub-20 FWD P/E. Nu trading for a FWD P/E of 14, despite such strong growth and quality characteristics, is objectively unbelievable.

Eventually, the market will return to wanting profits, as it did in 2022 after the 2021 EV/SaaS bubble burst.

I am investing accordingly.

Leave a comment

Become a Premium Member!

Here is what my Premium Members can expect:

Portfolio Review - Each month, I will present the portfolio performance and discuss my stock watchlist and my best ideas.

• Recent developments.

• Unwarranted pullbacks.

• Insider activity.

• Potential catalysts.

Deep Dives – 8,000+ word detailed analysis of a company, delivered in 3 Parts.

• Part 1 – Brief History of the company and its Business Model.

• Part 2 – Management, Moats, Competitors, and Risks.

• Part 3 – Opportunities, Financial Analysis, and a Valuation Model.

You can expect a comprehensive research report that is educational, interesting, and provides actionable insights!

To see what you can expect, read my Palantir Deep Dive!

Members of the Premium service get access to my library of 12 Deep Dives and to all future Deep Dives, which will be released on semi-monthly basis.

Investment Cases – A short, concise report with actionable insights.

This report is about the size of a single part of a Deep Dive.

• Focused Investment Thesis

• Main drivers of the Bull Case

• Valuation Model

To see what you can expect, read my Oscar Health Investment Case!

Earnings Reviews and Updates – For companies that are of great interest to me and my readers, I will provide regular quarterly or semi-annual updates after earnings reports.

• Financial performance

• Business Update

• New developments

• Updated Valuation Model

To see what you can expect, read my Google Q2 2025 Earnings Review!

Equity Research Report List

1. Micron

2. Nebius

3. Meridian Holdings

4. Weight Watchers

5. Lumentum

6. Sivers Semiconductors

7. Klarna

8. Ondas

9. Cipher Digital

10. SanDisk

11. Novo Nordisk

12. Sofi Update

13. UiPath

14. Mercado Libre Update

15. Nu Update

16. QXO

17. ASTS

18. Pagaya

19. IREN

20. Rubrik

21. Oscar Update

22. Arteche

23. Nebius Update

24. WhiteFiber

25. Harrow Health

26. Root

27. Zeta Global

28. SentinelOne

29. Golden Matrix Group

30. ASML

31. AMD

32. UnitedHealth Group

33. AppLovin

34. Adyen

35. TransMedics

36. Evolution

37. Nebius

38. Oscar Health

39. SoFi

40. Dlocal

41. Sea Limited

42. Microsoft

43. TSMC

44. Hims and Hers

45. Meta

46. Nu Bank

47. High Tide

48. Tesla

49. Grab

50. Google

51. Amazon

52. Hello Fresh

53. Mercado Libre

54. Celsius

55. Robinhood

56. Palantir

57. Ferrari

You can follow me on Social Media below:

X(Twitter): TheRayMyers

Threads: @global_equity_briefing

LinkedIn: TheRayMyers

Disclaimer: Global Equity Briefing by Ray Myers

The information provided in the “Global Equity Briefing” newsletter is for informational purposes only and does not constitute financial advice, investment recommendations, or an offer or solicitation to buy or sell any securities. Ray Myers, as the author, is not a registered financial advisor, and readers should consult with their own financial advisors before making any investment decisions.

The content presented in this newsletter is based on publicly available information and sources believed to be reliable. However, Ray Myers does not guarantee the accuracy, completeness, or timeliness of the information provided. The author assumes no responsibility or liability for any errors or omissions in the content or for any actions taken in reliance on the information presented.

Readers should be aware that investing involves risks, and past performance is not indicative of future results. The author may or may not hold positions in the companies mentioned in the “Global Equity Briefing” report. Any investment decisions made based on the information in this newsletter are at the sole discretion of the reader, and they assume full responsibility for their own investment activities.

This was yet another stellar quarter from Arkady and the Nebius team:

• Capacity guidance increased from 3GW to 4GW.

• New 1.2GW site in Pennsylvania.

• AI Cloud revenues up 841% Y/Y.

• 40% ADJ EBITDA margin target for 2026.

• 3 major acquisitions to strengthen the AI Cloud offering.

A month ago, I published my Nebius 2030 valuation model, which by now seems already outdated.

As such, I have decided to create a new one, with improved assumptions.

But before that, let me look at a few of the most important things announced in this earnings release.

1. Q1 2026

• Revenue: $399M +625%, vs $389M estimate.

• ADJ EBITDA: $130M +306%, vs $91M.

• ADJ EPS: -$0.39 vs -$0.78.

In short, Nebius grew faster than expected and demonstrated significantly better profitability.

1.1. ARR

ARR grew by 671% Y/Y, 54% Q/Q to $1.92B!

Simply put, Nebius’ March revenues were $160M, compared to $104 in December 2025.

The company reiterated its end-of-year 2026 ARR guidance of $7-9B!

This implies December 2026 revenues of $583-$750M. So, the company expects to grow its monthly revenue by 264-369% in just 9 months. Incredible.

1.2. Capacity Guidance

Most importantly, Nebius increased its end-of-year 2026 contracted capacity guidance from 3GW to 4GW!

This is a 4x increase in guidance since August 2025!

The increase comes from Nebius announcing a new greenfield, fully owned 1.2GW data center project in Pennsylvania.

“Today, we announced a new site in Pennsylvania to support 1.2 GW of power once fully lit live. This is our second owned gigawatt scale site in the United States. Our platform is most efficient when we own the full stack, and we are building towards that. Our owned contracted capacity now accounts for more than 75% of our total power.” Arkady Volozh, Nebius CEO

In this statement, Arkady revealed an important piece of information concerning a criticism often directed at Nebius. Owned data centers account for 75% of their contracted capacity.

Previously, some analysts criticised the company for relying too heavily on rented colocation facilities. It is much faster and cheaper upfront to develop a colocation data center, but it’s more expensive in the long term, as Nebius needs to pay rent and the landlord has to make a margin.

But colocation is just 25% of contracted capacity!

Much lower than many analysts estimated.

However, let’s remember that contracted power is the energy that the utility has agreed to deliver to Nebius, but a lot of the equipment still needs to be built and installed.

Meanwhile, connected power is what is already connected and ready to draw from the grid.

Active power is the power currently being consumed by Nebius clients for AI workloads, and this is what ultimately generates the company’s revenues. So contracted power guidance is great, but only as long as the GPUs actually start running.

Notice that Nebius didn’t give an active power guidance, because there is a lot of uncertainty about when exactly the company could have the facility built, GPUs delivered, and energy connected for active data center activities.

It will be a few years before the 4GW mentioned will be fully active!

Nevertheless, the pace at which this guidance increases clearly suggests that my earlier 3GW active power target for 2030 was too low.

If your head is spinning from this 4x increase in contracted power target, then you are not the only one. This is an absolutely insane target, and very ambitious. Understandably, this means that Nebius will have to invest significantly more to meet this target.

1.3. Capex

Thus, they increase their 2026 capex guidance from $16-20B to $20-25B!

“That is what is driving us to build more and to raise our 2026 CapEx guidance to between $20 billion and $25 billion, which is up from our prior range of $16 billion- $20 billion. This increase reflects investments in our 2027 capacity that will come online early next year. We expect these investments to contribute positi vely to revenue in the first half of 2027, where we already have customer and commitments in place. Meta is one such customer. We need to invest to fully realize it. This requires capital, which is our fourth dimension. We’re doing a very good job in tapping the market at scale.” Arkady Volozh, Nebius CEO

You can’t increase capacity without increasing capex, so this was expected and is totally normal.

I wouldn’t be surprised if capex guidance increases further.

As just $2.5B were spent in Q1, more than $20B will be spent in the next 9 months.

Where will this cash come from?

1.4. Cash Situation

Nebius finished Q1 2026 with an incredible $9.3B cash balance!

This already covers almost half of the needs, and a massive increase from $3.7B in Q4 2025. Nebius was able to raise funding from:

• $2B from the Nvidia investment.

• $4.3B from convertible note issuance.

• $3.2B in customer pre-payments.

The $2B Nvidia investment and $4.3B from convertible notes were already known by the market, the $3.2B customer pre-payments were not.

Do you like the chart above?

FiscalAI makes managing investments smarter, faster, and stress-free.

• Visualizations

• AI-powered insights

• Financial data

• Earnings transcripts

• Portfolio analytics

All in one place. Instead of wasting hours digging through filings and spreadsheets, Fiscal.ai helps you get to the important information in minutes.

Save time, stay organized, and let Fiscal.ai handle the heavy lifting so you can focus on growth.

Join with my link below.

https://fiscal.ai/?via=global

In fact, as of Q1 2026, Nebius has collected $4.8B in customer pre-payments!

This is extremely impressive and again destroys a bear argument. As there were no detailed announcements regarding pre-payments from specific customers, some bears claimed that Nebius won’t be getting any. This clearly disproves that.

Most importantly, only $686M of the deferred revenue liability is current and to be recognized in the next 12 months as revenue.

So Nebius has already received $4.1B in pre-payments from customers for services that will be delivered in 2027 and 2028. Again, this is extremely impressive and demonstrates how strong the demand is for their services.

Which company collects payments from customers more than a year before delivering the services?

Most of the time, in the corporate world, it is the opposite, with clients paying suppliers 60, 90, and even 180 days after receiving the service.

1.5. Acquisitions.

2026 has been busy for Nebius on the M&A front, as they did 3 important acquisitions:

• Tavily

• Eigen AI

• Clarifai

Tavily is an Agentic AI search business. Essentially, Tavily helps companies find the right data and format it in a way that AI Agents can easily understand and process it. Basically, Google Search for AI Agents.

Eigen AI is an inference and model optimization platform that specializes in making AI deployment faster and cheaper.

Clarifai is a deep learning and computer vision company that strengthens Nebius’ managed inference and orchestration offering.

All 3 of these acquired companies will be integrated into Nebius AI Cloud, offering to build a full suite of services that could rival Hyperscalers!

This is how Nebius plans to reach its ADJ EBITDA target!

1.6. ADJ EBITDA Target

Nebius just revealed an 40% ADJ EBITDA target for 2026!

This is a massive improvement from the 10% ADJ EBITDA margin Nebius has in Q2 2025.

The company is achieving scale and operating leverage, thanks to its AI cloud and bare metal offerings.

Despite 684% revenue growth in Q1 2026:

• Cost of sales grew only 320%

• SG&A expenses grew 136%

• R&D grew 85%

• SBC grew 102%

• Depreciation grew 332%

As it integrates the acquired capabilities and develops new ones, its AI cloud value proposition will only get better and better, supporting margin improvements.

2. Valuation

After rising 148% in 2026, Nebius now trades for a market cap of $53B!

Read more

I am Ray, a passionate investor and equity analyst. And today I am covering Micron!

For 50+ years, the memory industry was viewed as a brutal, cyclical commodity market.

Companies in this space were prone to extreme boom-and-bust cycles, where periods of high profitability were quickly followed by devastating oversupply and price collapses.

However, the rise of AI has fundamentally changed this dynamic.

Micron is transitioning from a cyclical commodity manufacturer into a structural enabler of AI.

Last quarter, the company grew revenues by an insane 196% and net income by 770%, reaching a net margin of 58%!

These types of results are unheard of for an industrial company of such scale and remind us of the stratospheric growth of Nvidia in 2024.

In this Micron Deep Dive, I will examine what is driving the growth and whether the company remains an attractive long-term investment opportunity.

Let’s begin.

1. Business Model

2. Manufacturing Strategy

3. Customers

4. Risks

5. Opportunities

6. Financials

7. Valuation

8. Valuation Model

9. Conclusion

1. Business Model

Micron is the only major American manufacturer of memory chips.

During the 1980s and 1990s, the memory industry went through brutal price wars initiated by Japanese and later South Korean competitors. Dozens of American memory makers were forced into bankruptcy or exited the market.

Micron survived through a relentless focus on cost-efficiency and strategic acquisitions. The 1998 purchase of Texas Instruments’ memory business and the 2013 acquisition of Japan’s Elpida Memory were key moments.

The Elpida acquisition, in particular, doubled Micron’s capacity and secured its position as a primary supplier for Apple’s iPhone.

In 2026 its Micron’s core products include:

• DRAM

• HBM

• NAND Flash

• NOR Memory

Let’s expand on them.

1.1. DRAM

DRAM is the short-term memory of any computer system.

It provides a high-speed workspace where the processor stores data for immediate access. Unlike storage, DRAM is non-permanent and volatile, so the data is lost when power is removed.

DRAM is by far the largest product category, generating $45.6B in LTM and 78.5% of total revenues!

This is also the primary growth area for the company, with Micron growing DRAM revenues by a mindboggling 207% in Q1 2026.

The growth is driven by the fact that the memory requirement for AI servers is much larger than that of traditional enterprise servers. While a standard server might require a few hundred gigabytes of RAM, an AI-optimized server requires up to 8 times the DRAM to handle the massive datasets and parameters of AI models.

Simply put, thanks to the AI boom, DRAM has transitioned from a supporting component to the primary enabler of LLMs.

Micron’s current leadership in this space comes from its 1-gamma node, which is the most advanced DRAM manufacturing process in the world.

The 1-gamma node is expected to become the highest-volume node in company history and is on track to represent the majority of the DRAM bit mix by the middle of calendar 2026.

This technology provides several benefits for AI:

• Bit Density: 45% increase on 1-beta node, allowing for more memory in the same space.

• Latency: 17% lower latency for AI inference tasks.

• Energy Efficiency: 24% better energy efficiency, reducing data center power costs.

As AI models grow in size, they require more DRAM capacity per server. Previous generation servers use several hundred gigabytes of DRAM, but new AI-optimized platforms are pushing these requirements into the terabyte range.

This change creates a structural increase in the total addressable market for DRAM, improving on the simple replacement cycles.

In short, by utilizing ASML’s EUV machines, Micron has been able to increase bit density and improve power efficiency, which is a must for data centers where electricity costs are a major driver of costs.

1.2. HBM

High-bandwidth memory (HBM) is a specialized type of DRAM that uses a vertical stacking architecture to achieve extreme data transfer speeds.

Instead of placing memory chips side-by-side, HBM stacks multiple DRAM dies on top of each other.

This stack is then placed on the same package as the AI processor, allowing for a much wider interface and shorter signal paths.

The main use cases for HBM include:

• AI training

• High-performance computing

• Advanced graphics rendering

In AI systems, HBM is the most critical component because it feeds data to the GPU.

Without HBM, the GPU would be starved of data, making it effectively useless for advanced AI. Micron entered the market with HBM3E, which offers approximately 1.2 TB/s of bandwidth per stack and up to 36GB of capacity.

Micron’s HBM3E is particularly competitive because it uses 30% less power than similar products from competitors. But the industry is now moving toward HBM4, which Micron began shipping in volume during Q1 2026.

HBM4 has a major architectural change, with more than doubling in bandwidth to 2.8 TB/s, and max capacity to 64 GB.

AI desperately needs HBM4 because the size of model parameters is growing faster than memory technology can keep up.

OpenAI’s and Anthropic’s trillion-parameter AI models are hitting a physical limit with the older 1024-bit interfaces. By doubling the interface width to 2048-bit, HBM4 allows for massive throughput without needing to increase the electricity voltage.

This helps manage the thermal wall, which is the heat trapped inside the stack that can damage the chips. Micron’s HBM4 is designed into the Nvidia Vera Rubin platform.

HBM revenues are reported within the overall DRAM segment, so we don’t know exactly, but it is likely a large share and the majority of the growth.

But some independent analysts estimate that last year Micron made about $6-10B in revenues from it, about 16-27% of total group revenues.

Subscribe now

1.3. NAND Flash

NAND flash is the permanent, non-volatile storage technology that retains data even when power is turned off.

It is the core technology used in solid-state drives, smartphones, and memory cards.

Unlike DRAM, which is built for speed, NAND is built for high-capacity storage. The main use cases include storing operating systems, applications, and large datasets.

For AI, NAND is used to store the massive amounts of data that are needed to train models.

The NAND segment generated $12.1B in revenues in LTM, about 21% of total revenues!

Do you like the chart above?

FiscalAI makes managing investments smarter, faster, and stress-free.

• Visualisations

• AI-powered insights

• Financial data

• Earnings transcripts

• Portfolio analytics

All in one place. Instead of wasting hours digging through filings and spreadsheets, Fiscal.ai helps you get to the important information in minutes.

Save time, stay organized, and let Fiscal.ai handle the heavy lifting so you can focus on growth.

Join with my link below by Thursday, 14 May, and get a 25% discount.

https://fiscal.ai/?via=global

Similar to DRAM, the NAND segment has also experienced stratospheric AI-driven growth, with revenues growing by an incredible 169% in Q1 2026.

As AI data centers are increasingly moving away from previous-generation hard disk drives and toward all-flash storage to speed up data access, Micron has released several products to support this trend:

• 245TB 6600 ION SSD: This is the highest-capacity data center drive available, allowing for a massive reduction in the number of server racks needed for storage.

• 122TB E3.S SSD: This drive offers 67% more server rack density than previous form factors and 37% better energy efficiency than using multiple hard drives.

• G9 NAND: Micron’s latest NAND technology provides the high speeds necessary for AI data lakes and vector databases.

One of the growing uses for NAND in AI is in inference.

During the inference process, AI models generate a large amount of temporary data called the KV cache to remember the context of a conversation.

As context windows grow to millions of tokens, this cache can become too large to fit in DRAM. High-speed NAND SSDs are now used to store this cache, which prevents the model from having to recalculate everything from scratch when users ask a question.

This offloading helps improve the efficiency of AI systems and reduces the total cost of ownership for data center operators.

1.4. NOR Memory

NOR memory is another type of non-volatile storage, but it is optimized for high-speed reading of small amounts of code rather than large datasets.

It allows a processor to execute code directly from memory, a feature known as execute-in-place. The main use cases for NOR are storing boot code for computers, firmware for automotive systems, and control software for industrial equipment.

Simply put, while NAND is used for large files, NOR is used for instant-on performance.

When an AI-powered car starts, the safety-critical microcontrollers must boot up in milliseconds to ensure sensors and cameras are active.

Micron’s automotive NOR flash is engineered to meet these demands, providing the durability and speed required for advanced driver assistance systems, sensors, and gateways.

Meanwhile, in the AI data center, NOR memory is used for the fundamental boot-up process of servers and networking switches.

The revenues from this product are reported within the NAND flash segment. While they are not experiencing explosive AI-driven growth, the demand trends still look strong and favorable, thanks to a growing usage in automotive systems.

Furthermore, the market for NOR flash is currently experiencing a supply squeeze.

Because major manufacturers like Micron and Samsung are focusing their factory capacity on more profitable HBM and advanced DRAM, the availability of NOR is tightening. This has led to longer lead times and firmer pricing for NOR products across all end markets, including industrial and medical sectors.

For Micron, this supply-demand imbalance in the broader memory market helps maintain strong profitability even in its smaller product lines.

1.5. Cyclicality

Micron operates in the semiconductor memory industry, which has long been dealing with high capital intensity, rapid technological obsolescence, and a notorious boom-and-bust cycle.

This cycle is driven by the incredibly strong demand elasticity of the memory industry.

When demand is high, prices rise rapidly as users of memory chips race to secure supply. This is exactly what we are seeing now. Memory is an input in the final product, which can’t be shipped without memory.

It takes years and costs $10-20B to build new fabs, so memory makers can’t quickly increase supply, further driving higher prices.

When prices are high, manufacturers look for ways to reduce memory requirements, reducing specs or changing designs.

Let’s not forget that high prices also incentivize suppliers to increase production. So, memory manufacturers all at the same time race to invest in capacity to increase supply. While it takes years for the additional supply to reach the market, it does at the same time.

Supply explodes, and pricing collapses.

The same factors that make increasing supply quickly difficult also make decreasing supply a non-starter. Once built, foundries must operate constantly, as the equipment that they spent billions on degrades while interest costs pile up.

In this industry, raw material input costs are a lower share of total costs than the depreciation of equipment.

Simply put, by shutting down foundries, memory makers will lose more money than by selling each chip at a loss.

Furthermore, Micron’s business is also complicated by the bullwhip effect!

It’s a supply chain phenomenon where small changes in consumer demand for end-products (like smartphones or PCs) result in increasingly larger swings in orders as one moves down the supply chain to the semiconductor manufacturers.

Read more

The company delivered strong growth across all key metrics. Find that article here.

Today, I am sharing my new Grab’s 2030 Valuation Model.

• Current Valuation

• Analyst Estimates

• 2030 Valuation Model

• GMV

• Segment ADJ EBITDA

• Profitability

• 2030 Target Share Price

Let’s begin.

Read more

I am Ray, a passionate investor and equity analyst. And today I am covering Grab.

Grab’s stock was already not doing great at the start of the year, but the Iran War accelerated the decline.

The company’s primary delivery and mobility businesses are very dependent on the steady stream of affordable oil to Southeast Asia. The region is one of the biggest customers of Middle East oil, and its economy is very exposed to oil shocks.

Despite these disruptions, Grab delivered great results, continuing to demonstrate that its business model is more resilient than many analysts anticipated.

• Best revenue growth in 9 quarters.

• ADJ EBITDA growth of 46% Y/Y.

• $120M in net income vs $10M last year.

• Accelerating GMV growth.

Overall, the start of 2026 was strong for Grab. Even though the first quarter is usually the slowest for the company due to holidays like Chinese New Year and Ramadan, Grab saw its growth accelerate.

The company focused on improving its operations while dealing with the Iran Oil Shock.

1. Overall Results

• Revenue $955M +24% Y/Y

• ADJ EBITDA $154M +46% Y/Y

• EBIT $22M vs -$21M last year

• Net Income $120M vs $10M last year

• EPS $0.03 vs $0.01 last year

• ADJ FCF $98M vs -$101M last year

1.1. Revenue

Total revenue grew by 23.5% Y/Y and 19% in FXN to $955M!

Grab benefited from a weak dollar this quarter, as such, its local currency growth was lower. Recent Trump’s actions, such as the Trade War and Iran War, have significantly weakened the USD, so this quarter the company benefited from that.

As you can see in the chart above, this was a good acceleration from 18.59% growth in Q4 2025 and 18.4% growth in Q1 2025.

The total revenue of $955M beat what many analysts on Wall Street expected, which was around $938M.

The growth was driven by all segments, and because more people were using the app and they were using it more often, we will expand on this later.

1.2. ADJ EBITDA and ADJ FCF

ADJ EBITDA was $154M, up from $106M in Q1 2025, and 46% increase Y/Y.

This beat the Wall Street consensus estimate of $142M!

Moreover, the ADJ EBITDA margin improved to 16.2% from 13.7%.

This clearly demonstrates how Grab is getting more efficient and making more profit for every dollar it earns.

Furthermore, looking at cash flow, this quarter, ADJ FCF turned positive at $98M vs $101M loss in Q1 2025.

On a trailing twelve-month basis, which looks at the last full year, ADJ FCF reached $489M.

Grab financial services segments distort normal FCF, so Grab shows ADJ FCF, which removes these distortions.

As you can see in the above picture from the earnings presentation, the company adds back the changes in loan receivables and deposits. The loan receivables position increased by $192M vs last year, which simply means that the company issued more loans.

By including it in FCF, it would make it appear that Grab lost money, which is not an accurate representation of its business.

1.3. Net Income and EPS

Grab reported a Net Income of $120M for the quarter, which is a big increase from the small $10M profit they had a year ago.

This was driven by an operating profit of $22M, compared to an operating loss of $21M in Q1 2025. The company also saw a big gain in the fair value of its financial assets, which added net $95M to the bottom line.

The net $95M gain consists of $163M gain on convertible note fair value and $68M of other loss adjustments.

The $163M gain relates to the the $1.5B convertible note Grab issued last year.

Under IFRS rules, Grab must reevaluate the fair value of this convertible note because they are categorised as a derivative financial liability and record gain or loss in each period.

Essentially, if the share price rises, Grab must record a loss as that means the holders of the convertible notes get to convert them into shares at a more favourable price. If the share price falls, as it did during Q1 2026, Grab must record a gain, as the convertible notes become less valuable to the note holders.

US GAAP doesn’t have this, thus Grab is exposed to additional income volatility each quarter.

This is why, even with the $120M profit, the fully diluted EPS was reported as -$0.01 despite basic EPS being $0.03.

Basic EPS is calculated by dividing net income by basic shares outstanding.

Diluted EPS is calculated by assuming that all outstanding warrants, notes, and employee stock options get converted into equity. This excludes the $163M gain on the fair value of the notes and thus the negative EPS.

This means that without this gain, Grab would not have been net income profitable!

I don’t like this, this essentially makes P/E not a usable metric for evaluating Grab, because the net income swings widely.

1.4. Guidance

Grab decided to keep its guidance for the full year of 2026 the same.

They are confident that they can meet these targets despite the challenges in the region. The company believes that its strong performance in Q1 provides a solid foundation for the rest of the year.

The company expects revenue to grow between 20% and 25% to $4.04-4.10B.

They also expect ADJ EBITDA to grow by 40% to 44% to $700-720M

By keeping this guidance, Grab is telling investors that they have plans in place to handle the fuel crisis and the new rules in Indonesia.

They also mentioned that they have enough cash to buy back up to $500M of their own shares, with $400M already being used in an accelerated program.

Subscribe now

2. GMV

In Q1 2026, On-Demand GMV grew by 24% Y/Y to reach $6.1B.

On FXN basis, the growth was 21%. This is a decent result because Q1 is usually a seasonally soft quarter.

GMV Growth accelerated from 20.86% in Q4 2025 and 16.27% in Q1 2025!

The growth in GMV was largely driven by two things.

First, the number of users grew by 16% to 51.6M.

This was the fastest user growth 5 quarters, and an acceleration from 15% in Q4 2025 and 15.6% in Q1 2025.

Second, each user is spending more money, with the GMV per monthly active user growing by 7% Y/Y to $130.

This indicates that once people start using Grab, they tend to use it for more things, like ordering groceries or taking more rides.

However, another important driver was incentives, which grew by 29.7%. Ideally, we would want the GMV to grow at a faster pace than incentives, but I will expand on incentives later.

Do you like this report?

Become a Paid Premium member to see dozens of exclusive pay-walled equity research reports such as this.

Curious about which stock I own?

Full portfolio with all holdings, cost basis, and unrealised gains is visible to Premium members. I do weekly updates and detailed monthly reviews. Premium members are also notified of all sales and purchases using the pay-walled Substack Chat.

The annual plan is available for 50% cheaper per month than the monthly plan.

Subscribe now

3. Delivery and Mobility

Both delivery and mobility segments saw strong growth in Q1 2026, despite significant outside challenges.

Deliveries GMV grew by 25% or 22% FXN to $3.9B while revenue grew 23% or 17% FXN to $510M.

Mobility GMV grew by 23% or 19%, FXN to $2.2B, while revenue grew 19% or 16% FXN to $337M.

For both segments, GMV grew faster than revenues, which clearly means that Grab’s take-rate shrank this quarter.

On a Y/Y basis, deliveries take-rate fell from 13.3% to 13.1%, while mobility take-rate fell from 15.6% to 15.2%.

However, on a Q/Q basis, both increased slightly from Q4 2025. This is not that spectacular, as the take rate in Q4 2025 was lower due to seasonally higher incentives related to the Christmas and New Year festivities discounts.

The deliveries segment benefited from the growth in GrabMart and a strong advertising business.

In mobility, the number of active driver-partners grew to an all-time high, increasing 16% Y/Y.

This helps ensure that users can always find a ride quickly. However, the high cost of fuel and new rules in Indonesia became the main topics of concern during the quarter.

One interesting highlight was the growth of GrabMart, which continues delivering strong results.

GrabMart GMV is now 10% of the total deliveries GMV and is growing 1.7 times faster than food delivery.

People who use the app for groceries tend to order 1.8 times more often than people who only order food. This helps Grab build a stickier relationship with its users.

Looking at Segment ADJ EBITDA, deliveries grew by 40% Y/Y to $88M, while mobility grew by 24% to $198M!

In the graph above, we see that despite various challenges, the mobility segment continues to be a profit driver for Grab, delivering ADJ EBITDA that’s more than double that of deliveries.

However, the delivery segment is growing at almost double the pace and will likely overtake the mobility in the long term.

There were 2 huge changes that affected Grab’s business this year:

• Iran War

• Take-rate law change in Indonesia

3.1. Iran War

On 28, February 2026, the Iran War began, causing a major shock to the global energy market.

The Strait of Hormuz, where about 25% of the world’s oil passes through, was closed/ This caused crude oil prices to jump over close to $100, reaching peaking at $120, but stabilising at around $100.

This fuel crisis had a direct impact on Grab’s driver-partners across its core markets.

When fuel prices go up, drivers make less profit from each ride.

To keep the drivers working and the marketplace healthy, Grab had to take several countermeasures:

• Targeted earnings support: Grab gave extra money and fuel rebates to drivers to help them with the high costs.

• Incentive peaks: Because of the fuel crisis and the holidays, total incentives reached a record $650M this quarter. Management expects this to be the peak and that incentives will go down in the next quarters.

• Turbo AI mode: Grab introduced a feature called Turbo for drivers. Drivers who used this AI tool to find better routes and rides saw their earnings per hour go up by 23%.

• EV transition: The high fuel prices are forcing Grab to move faster toward EVs. They are partnering with companies like BYD and GAC to help drivers switch to EVs, which are cheaper to run than gas cars.

• Price and fees: If fuel prices stay high, Grab said they might have to raise prices for customers or add new fees. However, they want to do this carefully so people don’t stop using the app.

Management said that despite the war and high fuel prices, the demand for rides remained strong in April.

Weekly average transaction volumes were up 32% compared to last year, and total weekly active driver partner online hours were up 19%.

Meanwhile, delivery daily transacting users were not only up compared to Q1 2025, but also up from December 2025 seasonal highs.

This is a clear signal that there is massive untapped demand that Grab still can reach, despite temporary cost shocks.

3.2. Take-Rate law change in Indonesia

On May 1, 2026, Indonesia’s President Prabowo Subianto announced a big change for the ride-hailing industry, signing a new rule that cuts the maximum commission that companies like Grab can take from drivers.

• Commission reduced from 20% to 8%

• Mandated health and accident insurance

This is potentially a 60% decrease in Grab’s revenues!

However, Grab confirmed in the earnings call that this rule applies only to 2-wheeled mobility drivers.

Grab pointed out that motorcycle rides in Indonesia account for less than 6% of their total mobility GMV. This means they have other parts of the business that are not affected by this new rule.

Nevertheless, this is a major change because it means Grab gets much less money from every motorcycle ride in Indonesia. This move was a response to years of protests by drivers who wanted better pay and working conditions.

Grab is working on several potential countermeasures to reduce the impact on profitability, including:

• Shift to advertising: Grab is moving away from making all its money from ride commissions. They are now selling more ads to merchants and other brands. The average spend per advertiser grew 44% this quarter.

• New fees for customers: To make up for the lost commission, Grab might increase platform fees or service fees that the customers pay. It is unclear yet if the new rule prohibits this or not.

• Higher prices: Fares in Indonesia might have to go up. Grab’s CFO said they will need to recalibrate the business model for two-wheelers in Indonesia.

• Reduce incentives: Grab pays hundreds of millions of dollars to drivers in various bonuses, which are not regulated by the government. If the government forces Grab to give a larger take-rate, one option is to simply reduce these incentives.

Grab is currently talking to the Indonesian government to understand exactly how and when the new rules will be put in place.

This could become a bigger issue than it might seem, depending on how the rules are implemented and how other regional countries react.

This could be nothing but theater by the president to improve his approval rating with millions of drivers and lower-income people who don’t understand how businesses work.

If Grab can simply increase platform or processing fees, then the real take-home pay and take-rate for drivers won’t see meaningful increases.

However, if the president is serious and bans additional fees, then Grab could be in trouble.

While Grab said it applies only to 6% of Indonesia’s mobility GMV, which could be about 1-2% of total GMV. However, the issue with government regulations, especially in low-income emerging countries, is that it usually tends to go one way, and that is worse. Regular taxi drivers and food delivery drivers also want higher pay and will likely demand similar measures.

Then drivers in the Philippines, Vietnam, Thailand, Malaysia, and other countries will hear about this new rule and demand similar changes.

This could be the beginning of a deteriorating regulatory environment.

4. Financial Services

Southeast Asia is a very promising region to grow a technology-first financial services business. Depending on estimates, around 50% of the population doesn’t have access to a bank account, and another 20% only have access to limited financial services.

Overall, this cohort of adults contains close to 300M people, thus, Grab has made it their key growth pillar.

Financial services revenue grew by 43% Y/Y to $107M!

We can see in the chart above that this is a meaningful acceleration from 33.8% in Q4 2025 and 36.4% in Q4 2025.

However, honestly, I am a bit disappointed that they didn’t hit 50% growth. At $107M a quarter scale, this segment is still small. Considering the TAM the company has, I would expect them to grow much faster.

But what impressed me the most is the strong growth of the loan disbursals, which expanded by 67% to reach $1.05B!

This is what I said last year in my Grab company Deep Dive regarding this segment.

“Grab’s mobile-first, automated, digital platform has the potential to originate billions of dollars in loans from its network of restaurants, drivers, merchants, and customers at much lower origination costs than other financial institutions. “

This quarter, Grab demonstrated how this is done, as gross loan portfolio grew by 130% to $1.4B. The company is targeting $2B by the end of the year.

Furthermore, Grab reported that customer deposits grew from $1.4B to $1.6B, only 14% Y/Y.

Also, a bit disappointing.

“we actually don’t have any issue at all in raising deposits….. If you look at the pricing of our deposits, we are never the most aggressive in the market. We’re able to actually gather sufficient deposits to create the right shape of balance sheet. There’s no point in having excess deposits, particularly in this yield curve environment. What you’re seeing is us carefully managing the level of deposits to make sure that we optimize for P&L purposes. If we needed to raise more deposits, we’re very confident that we can do that.“ Alex Hungate, Grab COO

Grab said that they are satisfied with their deposit base and are not chasing deposit growth with aggressive discounts. Essentially, Grab has enough cash to satisfy the demand it has for its loans, and the company doesn’t want to overpay for deposits.

This is because the financial segment is unprofitable, and Grab wants to reach profitability soon.

The segment lost $17M in Q1 2026, which is an improvement compared to $26 it lost in Q4 2025 and $30M in Q1 2025.

ADJ EBITDA margin improved from -26% in Q4 2025 and -40% in Q1 2025 to -16% in Q1 2026!

I am not sure if prioritizing the segment AJD EBITDA break-even on the account of deposit and market share growth is the right long-term strategy. The management clearly believes this to be the right strategy.

It could be that the current disruptions in Iran are affecting their thinking, as stronger investments to create growth in the financial services segment could lead Grab to be unprofitable on the group level, and the management clearly doesn’t want that.

5. Incentives

Incentives are the discounts Grab gives to customers and the extra money they pay to drivers. Both are designed to drive GMV growth by driving customer demand and driver supply.

In Q1 2026, total incentives were $650M +29.7% Y/Y over $501M spent last year!

A reminder that under IFRS accounting, these incentives are not recorded as costs on the income statement, but rather as a reduction of revenue. So by reducing incentives, Grab could drive stronger revenue growth, but we are not seeing that right now.

Below, we can see how incentives are split across mobility, deliveries, drivers, and consumers.

Driver partner incentives grew by 42% Y/Y to $305M, largely because of higher fuel bonuses from Grab!

Meanwhile, consumer incentives grew by 21% to $345M, largely because of the 31% growth in delivery consumer incentives.

There were two main reasons why incentives went up this quarter.

First, the holidays caused supply pressures, as Grab had to pay drivers more to stay on the road. Last quarter, there were the Chinese and Thai traditional calendar new year celebrations, when drivers from major cities often return to their home provinces and villages to visit family. Additionally, there was Ramadan, which affected some muslim drivers.

Second, the fuel crisis caused by the war in Iran made fuel very expensive, so Grab gave drivers extra support to make sure they could still afford to work.

As a percentage of On-Demand GMV, incentives rose to 10.5%, compared to about 10% a year ago.

Grab’s management believes that Q1 2026 was the peak for incentives.

They expect that as fuel prices normalize and they use more AI tools like Turbo, they won’t need to spend as much on incentives in the future.

6. Taiwan

A reminder that a few months ago, Grab acquired Foodpanda’s Taiwanese food delivery business.

This was a smart deal and an interesting opportunity.

I analyzed the deal in my previous report. Please find it below.

7. Conclusion

Despite the oil shock, Grab had a strong start to 2026!

The company reached a record $154M in ADJ EBITDA and saw its revenue grow by 24%.

They were able to manage a sudden fuel crisis caused by the war in Iran and still keep their driver supply at an all-time high. The growth of financial services into a $1B+ quarterly lending business shows that Grab is successfully becoming a fintech.

However, the growth of the fintech business is too slow considering the scale.

At $107M revenues a quarter scale, it should be growing much faster than 43% Y/Y. Grab must meaningfully accelerate the fintech business, or it risks missing the fintech opportunity.

Furthermore, the future has some big challenges, especially with the new 8% commission cap in Indonesia and the ongoing high fuel costs.

However, Grab is using AI and advertising to build new ways to make money that don’t depend only on ride commissions.

Additionally, by keeping its guidance the same, which was given before the Iran war, Grab is telling the investors that it is built to survive and grow even when times are tough.

Investors are still cautious, but the company’s progress toward becoming a highly profitable regional player in mobility, food delivery, and finance remains very clear.

It’s been 6 months since I published a Grab valuation model, I think it's time for an update.

So I will publish a new one tomorrow for my premium subscribers!

Become a Premium Member!

Here is what my Premium Members can expect:

Portfolio Review - Each month, I will present the portfolio performance and discuss my stock watchlist and my best ideas.

• Recent developments.

• Unwarranted pullbacks.

• Insider activity.

• Potential catalysts.

Deep Dives – 8,000+ word detailed analysis of a company, delivered in 3 Parts.

• Part 1 – Brief History of the company and its Business Model.

• Part 2 – Management, Moats, Competitors, and Risks.

• Part 3 – Opportunities, Financial Analysis, and a Valuation Model.

You can expect a comprehensive research report that is educational, interesting, and provides actionable insights!

To see what you can expect, read my Palantir Deep Dive!

Members of the Premium service get access to my library of 12 Deep Dives and to all future Deep Dives, which will be released on semi-monthly basis.

Investment Cases – A short, concise report with actionable insights.

This report is about the size of a single part of a Deep Dive.

• Focused Investment Thesis

• Main drivers of the Bull Case

• Valuation Model

To see what you can expect, read my Oscar Health Investment Case!

Earnings Reviews and Updates – For companies that are of great interest to me and my readers, I will provide regular quarterly or semi-annual updates after earnings reports.

• Financial performance

• Business Update

• New developments

• Updated Valuation Model

To see what you can expect, read my Google Q2 2025 Earnings Review!

Equity Research Report List

1. Nebius

2. Meridian Holdings

3. Weight Watchers

4. Lumentum

5. Sivers Semiconductors

6. Klarna

7. Ondas

8. Cipher Digital

9. SanDisk

10. Novo Nordisk

11. Sofi Update

12. UiPath

13. Mercado Libre Update

14. Nu Update

15. QXO

16. ASTS

17. Pagaya

18. IREN

19. Rubrik

20. Oscar Update

21. Arteche

22. Nebius Update

23. WhiteFiber

24. Harrow Health

25. Root

26. Zeta Global

27. SentinelOne

28. Golden Matrix Group

29. ASML

30. AMD

31. UnitedHealth Group

32. AppLovin

33. Adyen

34. TransMedics

35. Evolution

36. Nebius

37. Oscar Health

38. SoFi

39. Dlocal

40. Sea Limited

41. Microsoft

42. TSMC

43. Hims and Hers

44. Meta

45. Nu Bank

46. High Tide

47. Tesla

48. Grab

49. Google

50. Amazon

51. Hello Fresh

52. Mercado Libre

53. Celsius

54. Robinhood

55. Palantir

56. Ferrari

You can follow me on Social Media below:

X(Twitter): TheRayMyers

Threads: @global_equity_briefing

LinkedIn: TheRayMyers

Disclaimer: Global Equity Briefing by Ray Myers

The information provided in the “Global Equity Briefing” newsletter is for informational purposes only and does not constitute financial advice, investment recommendations, or an offer or solicitation to buy or sell any securities. Ray Myers, as the author, is not a registered financial advisor, and readers should consult with their own financial advisors before making any investment decisions.

The content presented in this newsletter is based on publicly available information and sources believed to be reliable. However, Ray Myers does not guarantee the accuracy, completeness, or timeliness of the information provided. The author assumes no responsibility or liability for any errors or omissions in the content or for any actions taken in reliance on the information presented.

Readers should be aware that investing involves risks, and past performance is not indicative of future results. The author may or may not hold positions in the companies mentioned in the “Global Equity Briefing” report. Any investment decisions made based on the information in this newsletter are at the sole discretion of the reader, and they assume full responsibility for their own investment activities.

I am Ray, a passionate investor and equity analyst. And today I am covering Sofi.

Sofi just reported a stellar earnings report, yet the market went mad, and the stock fell 16%.

Friday, I published my review of the quarter and explained why I believe that completely doesn’t make any sense. Please read that paywall-free article if you would like to know more.

Today, I am going to expand on why I have made Sofi one of my biggest positions and believe the company is significantly undervalued.

• Valuation Analysis

• 2030 Valuation Model

• Margins

• Dilution

• EPS

• Target Share Price

Let’s begin.

Read more

I am Ray, a passionate investor and equity analyst.

Today, I am reviewing the April performance of the Global Equity Portfolio.

A reminder that I track all aspects of the Global Equity Portfolio for my Premium subscribers.

• All sales/buys are revealed as they occur in the Substack Chat

• Portfolio composition

• Performance

• Strategy

• Stock watchlist

• Potential catalysts

Since its founding, the portfolio has returned 151.76%, a CAGR of 16.58%, beating the S&P500 return of 118.69%, CAGR of 13.88%!

In the above image, we see a graph from Portseido that compares the performance of the Global Equity Portfolio to the S&P500 if invested in the index at the same time as I bought the stocks.

I have been using Portseido to track the performance of the Global Equity Portfolio for some time now. They are an investment performance tracking and visualizations platform that is extremely easy to use. I especially like their smart benchmarking graphs and tables you can see in the picture above.

This is exactly what I was looking for, and I would recommend them to everyone looking to approach investing seriously and beat the market in the long term.

Use my link below to begin your 2-week trial.

https://www.portseido.com/?fpr=global

Let’s take a look at April’s performance.

1. Performance

The Global Equity Portfolio rose 7.01% in April.

Read more

I am Ray, a passionate investor and equity analyst. And today I am covering Sofi.

I am only 30 years old, so I have not been investing for that long.

But the reaction to Sofi’s earnings is the biggest disconnect between what the reality is and how the market reacted I have ever seen in my investing career.

• 43% Revenue Growth

• 62% ADJ EBITDA Growth

• 100% EPS Growth

Yet the stock was down 16%!

This is especially insane, considering the stock was already down 30% YTD and traded at a FWD P/E of 28 before the earnings.

Find me another stock that is delivering such rapid top and bottom-line growth at such a valuation, with such a strong TAM and incredible innovation in the last 3 years.

So why would the market react in such a way to these earnings?

There are some culprits we can blame.

1. Tech Platform segment is performing so badly that Sofi decided to rebrand it.

2. Deceleration in the Loan Platform Business.

3. Didn’t increase the guidance.

4. Dilution.

In this Q1 2026 earnings review, I will look at the quarterly performance and analyze all 4 of these reasons to determine whether they are valid or not.

Let’s begin!

1. Overall Performance

• Revenue $1.1B +43%

• ADJ EBITDA $339.9M +62%

• Net Income $166.7M +134%

• EPS $0.12 +100%

• Tangible Book Value $9.2B +83%

• Tangible Book Value per share $7.21 +57%

1.1. Revenue

The company brought in $1.1B in GAAP revenue, which is a 43% Y/Y increase from the $771.8M it made in Q1 2025.

As you can see in the graph above, Sofi demonstrated strong acceleration from the 19.65% growth it had in Q1 of 2025.

This growth is much faster than what legacy banks and other fintechs are seeing right now and is largely driven by the lending business, on which we will expand in a bit.

A big part of this money came from Net Interest Income, which reached $693M, up 39% Y/Y.

Sofi is making more money from interest because it is using member deposits to fund its loans, which are cheaper than other ways to get money.

Meanwhile, gee-based revenue reached $386.8M, up 23% Y/Y. This includes money from the loan platform business and fees from credit cards and investing. While this is growing, it is not growing as fast as the interest income.

1.2. ADJ EBITDA

ADJ EBITDA is Sofi’s preferred way to see how much money the business makes from its daily operations.

In Q1 2026, ADJ EBITDA was a record $339.9M, which is 62% higher than the $210.3M made in Q1 2025.

The ADJ EBITDA margin was 31%, showing that the company is getting more efficient with each quarter.

Sofi also emphasizes the Rule of 40 metric, which is popular with tech companies, where you sum the growth rate and the profit margin. If the score is above 40, it is considered that the company is achieving a great combination of growth and operating leverage.

Sofi had a score of 72 in Q1 2026. This was the 18th quarter in a row that Sofi has beaten this goal.

Furthermore, Sofi reported and incremental EBITDA margin of 41% means that for every new dollar of revenue, 41 cents went straight to profit.

This shows that Sofi’s business model is working. The more members they get, the more profit they make because they don’t have to spend as much to serve each new person.

1.3. Net Income and EPS

For Q1 2026, GAAP net income was $166.7M. This is a massive jump of 134% Y/Y from the $71.1M made last year. This resulted in an EPS of $0.12.

This number was exactly what analysts expected, but it was not a beat like in previous quarters.

EPS grew by 100%, significantly lower than the 134% net income growth. This is because of Sofi issuing new shares and diluting shareholders.

The net income margin was 15%, which is double what it was in Q1 2025. Management also mentioned that if the stock price had not changed, the EPS would have been $0.13.

This is because the way they account for some stock-based costs depends on the share price.

The Tangible Book Value per share is now $7.21, up 57% Y/Y.

This is a very important metric for banks. It shows the actual value of the assets the company owns.

The fact that this is growing so fast suggests the company is becoming much more valuable, even if the stock market is currently disagreeing.

1.4. Guidance

As I mentioned in the introduction, guidance was the disappointment that caused the stock to sell off so much.

For Q2 2026, Sofi expects ADJ net revenue to be about $1.115B, which is 30% growth Y/Y. They expect an ADJ EBITDA margin of 30% and an EPS of $0.10 to $0.11.

The problem for investors was that Sofi did not raise its full-year 2026 guidance.

They still expect about $4.655B in revenue and $0.60 in EPS for the whole year.

Usually, if a company has a record Q1, they raise their goals for the year.

By keeping them the same, it looked like they expect a slowdown later in 2026.

In an interview with Bloomberg, CEO Anthony Noto explained that they made the guidance before the Iran War. At that point, they expected 2 rate cuts to happen in 2026 and modeled their performance accordingly.

Due to Trump’s actions in Iran, that seems unlikely now, as oil, natural gas, fertilizer, and other commodity prices have risen substantially.

Subscribe now

2. Tech Platform

The “AWS of Fintech” bull case is officially dead!

Revenue for the Tech Platform was only $75.1M. This is a 27% drop from the $102.9M it made in Q1 2025.

Even worse, it was a 38% drop from the $122M it made just last quarter in Q4 2025.

This part of the company is supposed to be the tech in fintech, providing the backend systems for other banks.

The main reason for this failure was the loss of a very large customer. While Sofi did not name them in every report, it is known to be Chime.

Chime moved its business off Sofi’s platform at the end of 2025, and the loss of that money was felt heavily this quarter.

The number of accounts on the platform also fell 16% Y/Y to 133M.

This segment was supposed to be a high-margin business that didn’t need a lot of capital, but right now it is struggling to find its footing. Management says that if you ignore the lost customer, the segment actually grew 12% Y/Y, but even that is patheticaly slow growth.

Because the performance was so bad, Sofi is rebranding the whole segment.

It will now be called Sofi Technology Solutions.

They are also changing how they sell the product. Instead of just selling to small fintechs, they are trying to sell to big banks. They have broken the business into four areas:

• Processing

• Banking Core and Services

• Payment Hub

• Risk and Fraud

Management hopes this change will help the segment grow by 20% to 25% in the future.

They also noted that 13 new clients started using the platform this quarter, but they are clearly much smaller than Chime was.

For many bears, the performance of this segment clearly shows that Sofi might just be a bank and not a fast-growing tech company.

Unless the SofiUSD revenue scales and gets reported in this segment, I am skeptical that things can be turned around. We have been hearing one excuse after another for 3 years now.

First, it was because the fintechs didn’t have the budget for upgrades after the 2022 stock market crash.

Then, it was because inflation and high interest rates reduced capex budgets.

Now, it’s because of the Iran War.

The execution in this segment has been poor and extremely disappointing from Anthony Noto and the Sofi team.

It could be that the TAM for this segment is much lower than the company hoped, or that many fintechs and banks don’t want to help fund their competitor.

Do you like this report?

Become a Paid Premium member to see dozens of exclusive pay-walled equity research reports such as this.

Curious about which stock I own?

Full portfolio with all holdings, cost basis, and unrealised gains is visible to Premium members. I do weekly updates and detailed monthly reviews. Premium members are also notified of all sales and purchases using the pay-walled Substack Chat.

The annual plan is available for 50% cheaper per month than the monthly plan.

Subscribe now

3. Member and Product Growth

While the technology business is struggling, the banking business is growing very fast.

Sofi is very good at getting new members and selling them more than one product. Sofi calls this the Financial Services Productivity Loop. The more products a member uses, the more money Sofi makes, and the less it has to spend on ads.

In Q1 2026, Sofi added 1.1M new members. This is a record and brings the total number of members to 14.7M, which is up 35% Y/Y.

It is extremely impressive for a company of this size to keep growing its customer base at such a rapid pace.

The total number of products reached 22.2M, up 39% Y/Y.

This means that the average member has about 1.5 products with Sofi. A very important number is the cross-buy rate. This shows how many new products were opened by people who were already members. In Q1 2026, this was 43%, up from 36% a year ago.

This is great for Sofi because it’s much cheaper to sell to an existing customer than to find a new one.

As you can see in the chart above, this is a 10.5% improvement from 2020. However, it has not been growing very fast. This is largely because Sofi keeps growing the number of customers, which is their priority.

Once it becomes harder to acquire new users, Sofi will focus more on driving up the product per member.

Sofi’s brand is also getting much stronger.

Unaided brand awareness hit 10% this quarter. This means that when you ask people to name a bank, 10% of them say Sofi without being prompted.

This is double the level of 2022. Sofi was also named the #1 U.S. Bank by Forbes and ranked #1 for DIY investor satisfaction by J.D. Power.

Much of the growth is coming from financial services products like Money and Relay. These products grew by 34% and 44% Y/Y.

But interestingly, Crypto products reached 239.5K after launching in Q4 2025.

This is a really strong growth in a single quarter, especially considering that crypto and growth stocks have not been doing that well since the Iran War.

Meanwhile, lending products also grew by 33%, driven by personal and home loans.

4. Lending

The lending segment is still the biggest part of Sofi’s business, and in Q1 2026, the segment was running at full speed.

Total loan originations hit $12.2B, +68% Y/Y from the $7.2B originated in Q1 2025 and +16% Q/Q from $10.49B in Q4 2025.

As you can see in the chart above, it was the strongest Y/Y lending growth in 9 quarters and an acceleration from Q4 2025.

Personal loans were the largest part of this, reaching $8.3B, up 51% Y/Y.

Whilst student loan originations accelerated massively, jumping 119% Y/Y to $2.6B.

This is the result of the Trump administration restarting student loan repayments that the Biden administration stopped during COVID. People are looking to refinance their student debt as they notice higher interest charges hitting their bank accounts.

This is how Sofi acquires high-value customers. By offering affordable loan refinancing, it is acquiring university graduates early in their careers. Then Sofi moves them down in the funnel, upselling them new products, like mortgages, which was suge growth this quarter.

Home loans also saw huge growth, up 137% Y/Y to $1.2B.

This is an area that the company is increasing its focus on, in preparation for a possible real estate boom, which could happen if interest rates come down in the US.

Recently, Sofi announced its expansion into home equity lines of credit (HELOC). This allows people who own a home to borrow money against the appreciated equity value, even if the mortgage has not been fully paid off. This has become very popular because many people have a lot of equity in their homes as prices have risen in the last decade, but don’t want to sell

Overall, the revenues of the lending segment grew by 55.4% Y/Y to $642.4M.

In the chart above, you can clearly see that the company delivered strong Q/Q growth as well, growing by 28.9%.

This was the best growth in over 9 quarters!

4.1. Loan Platform Business

The loan platform business has been a huge success and the company’s biggest innovation in the last 2 years.

In Q1 2026, Sofi originated $2.96B on behalf of its partners, up 85% Y/Y.

While the growth looks impressive on a Y/Y basis, it was not as strong on a Q/Q basis, as you can see looking at the orange line in the graph above. Originations declined 19% sequantally, and this was part of the reason that added to the stock sell-off.

LPB generated $138.3M in revenues, which was a 88% Y/Y increase, but a 28% Q/Q decrease. Take-rate also went down from 5.22% in Q4 2025 to 4.67% this quarter.

Sofi CEO Anthony Noto later commented to Bloomberg and Yahoo Finance that this decline was not due to loss of interest from funders, but rather a focus on growing their own loan book. The fact that overall orignations grew so strongly, 68% Y/Y and 16% Q/Q, gives credence to this explanation.

The take rate may have fallen because there could be some volume incentives from the credit partner to issue more loans.

4.2. Credit Quality

Credit quality has always been a big topic for Sofi, especially since the short-sellers Muddy Waters recently claimed that Sofi’s loan book is riskier than the company claims.

They alleged that Sofi misstated $312M in debt and that the real charge-off rate for personal loans is 6.1%.

Sofi says these claims are “factually inaccurate and misleading”.

They reported a personal loan net charge-off rate of 3.03% for the quarter. This is up from 2.80% last quarter but down from 3.31% a year ago.

If you include the delinquent loans that Sofi sold, the rate would be 4.4%.

Management says their losses are well within the 7% to 8% range they planned for.

Furthermore, looking at historic loan quality vintages, we see that Sofi loans today are performing much better than the loans the company issued in 2017.

As of Q4 2025, 43% of the unpaid principal remained from issued loans.

3.94% of that was in default, compared to 5.93% at the same maturity for loans issued in 2017, a gap of 159 basis points.

As of Q1 2026, we see that the gap is improving, with 4.64% of loans in default compared to 6.32%, 168% basis points.

Simply put, Sofi defaults are lower than in 2017, despite issuing multiples more loans, because they are issuing less risky loans to richer people.

There is nothing to worry about, as the average Sofi personal loan borrower has an income of $154,000 and a credit score of 745. These are very high-quality borrowers. The student loan charge-off rate is even lower at just 0.65%.

Overall, Sofi’s credit performance seems stable, but the market remains nervous because of the short-seller reports.

5. Financial Services

The Financial Services segment Revenue was $428.5M, up 41% Y/Y.

This was a meaningful deceleartion, and the worst growth quarter in over 9 quarters!

This was one of the reasons the stock sold off, but it doesn’t make much sense to me, as all other segments performed strongly.

I have to note that the LPB revenues are reported here, not part of lending, as lending revenues are where Sofi actually holds the loans on its balance sheet. The $50M Q/Q decrease in LPB revenues was essentially the main reason why this segment decelerated.

Share

6. Dilution

The narrative is that Sofi is a heavily diluting company!

But that is simply not the case, as all dilution needs to be looked at in the context of the financial results delivered.

In the last 2 years, Sofi has been steadily increasing the number of shares outstanding by 10% per year.

During this time, revenues have grown with a 30.6% CAGR, while EPS has grown with 27.5% CAGR.

It is clear that Sofi dilution is smart, balanced, with a clear strategy, and most importantly, not financially dilutive.

However, by now, it is also clear that the last 2 dilutions were a strategic mistake, despite it not being a financial mistake.

It’s the only criticism that I see repeated again and again in comments, “Noto diluted,

is a terrible company, dilution is bad”.

I suspect that this dilution was actually the real reason why Muddypants attacked Sofi. They saw that this bear narrative is strong and persistent, and wanted to cash in on it by adding gasoline to it.

There is no objective metric by which Sofi could be a “terrible company”, yet this narrative persists because of the dilution and stock performance.

During the earnings call, they said that these note issuances were not economically dilutive to the EPS. This is because Sofi used the $3B+ it raised to grow the EPS.

Simply put, the EPS is larger today than if they hadn’t diluted, even though the share count is higher.

Here is the real issue, investors don’t care about it.

They only care about the share count going up.

It doesn’t matter what Sofi does with the money, selling stock to raise funds is “bad, full stop” and the market doesn’t want Sofi to do that.

If Sofi had not done these raises, the share price today would be higher, even though the EPS would be lower.

That’s why I am saying that it was a strategic mistake, despite it not being a financial mistake.

However, in the long term, Sofi is delivering strong results, releasing new products, and growing EPS.

Just looking at the number of shares outstanding, without taking into account the incredible EPS growth the company is delivering, is simplistic and stupid.

The market will realize this as Sofi continues to execute quarter after quarter, and this dilution narrative will die.

7. Conclusion

The Q1 2026 report for Sofi shows a company that is doing two very different things.

On one hand, the banking business is a massive success. Members are joining in record numbers, they are using more products, and the company is making a solid profit.

The lending business is also breaking records, despite high interest rates, persistent inflation, and the Iran War.

On the other hand, the technology business is in trouble.

Losing Chime was a big hit, and the rebranding to “Sofi Technology Solutions” shows that the company is attempting to change its strategy to survive.

But most importantly, even though Sofi hit its profit goals, the fact that they didn’t raise their guidance for the rest of the year made investors think that the best growth might be behind them.

For the rest of 2026, Sofi has to prove that it can keep growing without the technology platform being a major drag.

They also have to show that their loans are as safe as they say they are, especially with short-sellers watching every move.

If they can do that, the company could be very valuable in the long run.

But for now, the stock remains very volatile as the market weighs the record growth against the risks.

However, I firmly believe that Sofi is one of the most undervalued long-term opportunities on the market.

So I will be releasing a new updated Sofi 2030 valuation model, with my earnings estimates and updated share price target.

It will be available for my premium subscribers on Sunday!

Leave a comment

Become a Premium Member!

Here is what my Premium Members can expect:

Portfolio Review - Each month, I will present the portfolio performance and discuss my stock watchlist and my best ideas.

• Recent developments.

• Unwarranted pullbacks.

• Insider activity.

• Potential catalysts.

Deep Dives – 8,000+ word detailed analysis of a company, delivered in 3 Parts.

• Part 1 – Brief History of the company and its Business Model.

• Part 2 – Management, Moats, Competitors, and Risks.

• Part 3 – Opportunities, Financial Analysis, and a Valuation Model.

You can expect a comprehensive research report that is educational, interesting, and provides actionable insights!

To see what you can expect, read my Palantir Deep Dive!

Members of the Premium service get access to my library of 12 Deep Dives and to all future Deep Dives, which will be released on semi-monthly basis.

Investment Cases – A short, concise report with actionable insights.

This report is about the size of a single part of a Deep Dive.

• Focused Investment Thesis

• Main drivers of the Bull Case

• Valuation Model

To see what you can expect, read my Oscar Health Investment Case!

Earnings Reviews and Updates – For companies that are of great interest to me and my readers, I will provide regular quarterly or semi-annual updates after earnings reports.

• Financial performance

• Business Update

• New developments

• Updated Valuation Model

To see what you can expect, read my Google Q2 2025 Earnings Review!

Equity Research Report List

1. Nebius

2. Meridian Holdings

3. Weight Watchers

4. Lumentum

5. Sivers Semiconductors

6. Klarna

7. Ondas

8. Cipher Digital

9. SanDisk

10. Novo Nordisk

11. Sofi Update

12. UiPath

13. Mercado Libre Update

14. Nu Update

15. QXO

16. ASTS

17. Pagaya

18. IREN

19. Rubrik

20. Oscar Update

21. Arteche

22. Nebius Update

23. WhiteFiber

24. Harrow Health

25. Root

26. Zeta Global

27. SentinelOne

28. Golden Matrix Group

29. ASML

30. AMD

31. UnitedHealth Group

32. AppLovin

33. Adyen

34. TransMedics

35. Evolution

36. Nebius

37. Oscar Health

38. SoFi

39. Dlocal

40. Sea Limited

41. Microsoft

42. TSMC

43. Hims and Hers

44. Meta

45. Nu Bank

46. High Tide

47. Tesla

48. Grab

49. Google

50. Amazon

51. Hello Fresh

52. Mercado Libre

53. Celsius

54. Robinhood

55. Palantir

56. Ferrari

You can follow me on Social Media below:

X(Twitter): TheRayMyers

Threads: @global_equity_briefing

LinkedIn: TheRayMyers

Disclaimer: Global Equity Briefing by Ray Myers

The information provided in the “Global Equity Briefing” newsletter is for informational purposes only and does not constitute financial advice, investment recommendations, or an offer or solicitation to buy or sell any securities. Ray Myers, as the author, is not a registered financial advisor, and readers should consult with their own financial advisors before making any investment decisions.

The content presented in this newsletter is based on publicly available information and sources believed to be reliable. However, Ray Myers does not guarantee the accuracy, completeness, or timeliness of the information provided. The author assumes no responsibility or liability for any errors or omissions in the content or for any actions taken in reliance on the information presented.

Readers should be aware that investing involves risks, and past performance is not indicative of future results. The author may or may not hold positions in the companies mentioned in the “Global Equity Briefing” report. Any investment decisions made based on the information in this newsletter are at the sole discretion of the reader, and they assume full responsibility for their own investment activities.

The market liked the results, with the stock ending the day up 3.7% and 29% in the past week.

• Rebranding to Meridian is completed.

• Best revenue growth in 4 quarters.

• Strong ADJ EBITDA growth of 26%.

• Net leverage ratio down to 0.53x.

After a complex merger and a major rebranding effort, the company has proven that its diverse business model can generate cash while expanding into some of the most valuable gaming markets in the world.

This report provides a detailed look at the results of Q1 2026, exploring the performance of individual business segments and the broader strategic moves that have strengthened the company’s position.

Disclosure: This is Issuer-Sponsored Research

Let’s begin.

1. Overall Performance

• Revenue: $50.1M +17.3% Y/Y

• Gross Profit: $28.1M +16%, margin of 56%

• ADJ EBITDA: $6.3M +26%, margin of 12.6%

• EBIT: $3.2M vs -$0.11M last year

• Net Income: $2.3M vs -$0.26M last year

• EPS: $0.18 vs -$0.02 last year

• OCF: $5.2M -33.4% Y/Y

The narrative for the Q1 2026 is one of growth, efficiency, and a successful return to profitability.

1.1. Revenue

Meridian reported total revenue of $50.1M, which is a 17.3% increase from $42.7M in Q1 2025.

As you can see in the chart above, this is the fastest growth in 4 quarters and the best performance since the merger consolidation!

This growth is quite impressive, considering it occurred while the company was undergoing a reorganization and merger integration.

The ability to grow sales by double digits while reorganizing the company suggests that the underlying business is very healthy and that customers are increasingly attracted to its betting and gaming products.

1.2. Profitability

Meridian reported an ADJ EBITDA of $6.3M for the quarter, a 26% Y/Y increase and a beat of the company’s guidance of $6.1M.

As you can see in the graph above, this was the best result in 5 quarters!

The ADJ EBITDA margin rose to 12.6%, up from about 11.7% in Q1 2025. This margin expansion is a clear indicator that the company is becoming more efficient and its growth strategy is delivering results.

However, the most important news is that GAAP net income reached $2.3M.

One year ago, the company recorded a net loss of $230K.

This is a welcome development after the large loss last quarter, driven by goodwill and intangible asset impairment.

A reminder that last quarter, Meridian recorded an impairment charge that reduced the balance sheet value of goodwill and intangible assets by $91.8M.

This was a non-cash accounting entry, meaning it does not impact the company’s actual bank balance or daily operations.

When a business gets acquired, it is usually for a premium to the accounting value of its net assets. In such situations, the acquirer records the difference between the purchase price and the net asset value of the acquired business as goodwill on the consolidated balance sheet.

Companies pay a premium to the accounting value of net assets because they expect the acquired business to grow earnings. And we are seeing that the Meridian is now delivering strong earnings growth.

As of Q3 2025, Meridian had $129M of goodwill and other intangible assets on the balance sheet.

US GAAP requires yearly tests to assess whether that goodwill position is appropriate. Rules require that the goodwill position be reduced in certain situations when the share price goes down, interest rates increase, or some business units underperform. This reduction appears on the income statement as a loss.

These impairments, while disappointing, don’t change the core investment thesis of the company as they mostly affect the less interesting parts of the business.

We see that the company has reached a tipping point where revenue is growing faster than expenses, allowing more money to flow to the bottom line.

Let’s see what enabled Meridian to achieve such an improvement.

To understand how the company makes money, we must look at its gross profit, which rose 16% to $28.1M.

For a company like Meridian, gross margin shows what’s left after paying the winnings and the direct costs of serving the games and bets.

As we can see in the chart above, compared to Q1 2025, gross margin remained steady at 56%, while gross profit improved by $4M.

This stability demonstrates that the company is managing its costs well, even as it grows in new B2C markets like Brazil and Mexico, and B2B markets like the US and Canada, which incur very high startup costs.

Moreover, the operating profit was $3.1M, a massive turnaround from the small operating loss seen in Q1 2025.

This improvement happened because the SG&A increased only by 2.8% to $25M, compared to the revenue growth of 17%.

This is a clear example of Meridian improving its operating leverage, as it uses its existing infrastructure, technology, and staff to handle more business without linear expense growth.

1.3. Balance Sheet

Meridian significantly improved its balance sheet stability, countering a key bear narrative of excessive debt.

The company has been aggressively paying down the debt it took on to merge Meridianbet with Golden Matrix.

As of Q1 2026, the company’s total debt was $29.7M, down 54% from $64.4M.

If we look at net debt, which is the total debt minus the cash the company has in the bank, the figure fell by 62% to just $13.4M.

Meridian now has a net debt leverage ratio of 0.53x, which is very healthy.

Simply put, the company is generating more than enough income to cover debt obligations, giving it meaningful financial flexibility to invest in its business or new M&A.

The company ended the quarter with $16.2M in cash, which is down 45.3% Y/Y as it pays down debt.

1.4. Cash Flow

Meanwhile, operating cash flow fell by $2.58M, -33.4% Y/Y to $5.16M!

As you can see in the chart above, OCF is a very lumpy metric that fluctuates a lot depending on profitability and cash outflow.

While this quarter, net income grew by $2M, Meridian paid $1.72M in accounts payable compared to last year, when it built up $4.3M in additional accounts payable during the quarter.

ADJ EBITDA is a great proxy for OCF, as it essentially shows what OCF would be if we ignored the timing of receivables and payables.

1.5. Guidance

Based on this strong performance, the company has issued positive guidance for the second quarter of 2026.

Management expects revenue to be between $51-53M, which would represent Y/Y growth of 18% to 23%.

This is slightly above the analyst consensus estimate of $51.6M.

Subscribe now

2. Meridianbet

Meridianbet segment is the core engine of the company, accounting for 70% of total group revenue.

This segment runs online sports betting and casino games across more than 18 regulated countries in Europe, Africa, and South America.

In Q1 2026, Meridianbet generated $34.9M in revenue +26% Y/Y!

This growth is coming from both the online side of the business and the company’s network of 740 betting shops.

The success of Meridianbet is driven by a massive increase in the number of people using the platform.

During this quarter, the brand signed up 428K new customers, a 41% increase compared to the same quarter in 2025.

This is a clear sign that the company’s marketing efforts are working well.

More importantly, these new users are becoming active players with the number of depositors rising by 27% to 283K, and the total number of active users grew by 21% to 334K.

This segment is also highly profitable, with a gross margin of 69.3% and an operating income of $6.6M. The 37% Y/Y growth in operating income for this segment suggests that the company is getting more value out of each customer.

3. Expanse Studios

Expanse Studios is the creative core of the company. While Meridianbet focuses on letting people bet, Expanse Studios focuses on building the actual online casino games.

In Q1 2026, Expanse continued to grow its footprint.

Its games are now available on 1,519 different operator sites, which is 175 more than at the end of last year.

The studio launched six new proprietary games this quarter, bringing its total collection of titles to 77.

Additionally:

• The number of players grew by 156%

• Revenue grew by 345%

• Pay-ins grew by 218%

• Gross gaming revenues grew by 133%

It is clear that this segment is delivering explosive growth. The revenues of Expanse are included in the Meridiabet segment revenues.

Furthermore, one of the most important moves this quarter was a new partnership with QTech Games, a leading distributor in emerging markets. This deal allows Expanse’s games to be offered to millions of new players in the Americas, where online gaming is just starting to take off.

The studio also introduced social casino versions of its games, which allow people to play for fun without betting real money, helping to build brand awareness and letting people test their gaming strategies.

To sell games in new regions, a studio must have its software tested and certified by the local government.

Expanse has been very successful in getting these stamps of approval.

During the first quarter, the studio secured new certifications in:

• Latvia

• Estonia

• Sweden

• Portugal

Furthermore, it is also waiting for certification in Ontario, Canada, and New Jersey, US, which are some of the biggest gaming markets in the world.

This business is very scalable, as by obtaining these certificates, Expanse can grow its revenue with limited capital investment.

Once scaled, the studio could become a potentially high-margin part of the business.

4. RKings and Classics

The RKings and Classics for a Cause segments are focused on prize competitions.

This is a unique type of e-commerce where people pay a small fee to enter contests to win luxury items like sports cars, houses, or vacations.

RKings is Meridian’s competition business based in the United Kingdom.

In Q1 2026, it generated $7.7M in revenue, a 12% increase over the previous year.

This was driven by the average order value, rising 29% to $16.91.

Furthermore, the value the company gets from each new person who registers went up by 15% to $17.72.

This segment is undergoing a transition in strategy, where they focus on attracting fewer but higher-value customers.

These improvements show that the company is getting better at marketing and is attracting these higher-value customers who are more engaged with the platform.

Classics for a Cause is an Australian business that focuses on car raffles.

Unlike RKings, which mostly sells single tickets, Classics for a Cause has built a promising subscription model.

People pay a monthly fee to be VIP members, which gives them regular entries into prize draws.

During Q1 2026, the number of VIP subscribers passed 10,000 for the first time in a year, ending the quarter at 10,750.

Meanwhile, the number of new users rose 18% to nearly 10,000, and the total number of transactions increased by 12% compared to Q4 2025.

Overall, this segment grew revenues by 4.3% to $4.4M!

5. GMAG

GMAG is the company’s casino game aggregator business focused on Asia, that’s been underperforming lately.

In Q1, the GMAG segment reported revenue of $3.1M, which was a 18.4% drop from the $3.8M it earned a year earlier.

This segment is not only a drag on the group’s revenue growth, but also on profitability. Gross margin is 33.9%, and the segment lost $0.4M this quarter.

To counter this slump, the company is focused on improving the GMAG offering.

During the quarter, the segment added 12 new game providers, which is a huge increase from only adding 2 providers in the same period last year.

It also added 1,382 new games to its platform.

By adding more providers and more games, GMAG is aiming to become more valuable to casino operators.

6. Mexplay

Mexplay is the company’s Mexican online casino and sports betting brand.

While it is currently a small part of the overall group, it is growing at an incredibly fast pace.

The number of people signing up for Mexplay grew by a staggering 271%, reaching a total of 74,000 registered users.

Even more important is that the number of first-time depositors rose by nearly 200% to 6,101 people, while first-time deposit value grew by 183% to $61K.

This shows that Mexplay is successfully converting people into paying customers.

The success of Mexplay is not just about the money it makes in Mexico, it is a test case for how the company can enter other Spanish-speaking countries.

By learning how to market to Mexican players and follow the local rules, the company is building a playbook that it can use as it expands across Central and South America.

7. Valuation

A market cap of $95M, TTM revenues of $190M, and adjusted EBITDA of $20M implies that the company is trading for 0.5x P/S and 4.4x EV/ADJ EBITDA.

That compares favorably to competitors.