Build your Sleep Well Portfolio today

Hi all,

After a quiet July, August saw some very good earnings for our Sleep Well Portfolio. The purchases we made during the ‘SaaS-apolyse’ are now looking excellent. Veeva and Pro Medicus are in particular nearly a bagger in just a few months. Meanwhile, the ecommerce/fintech basket, Sea Limited and Mercado Libre, rallied steadily.

The AI narrative is turning from disruption to ‘value creation’.

You can read six core position reviews below, including (Wisetech, Pro Medicus, Cellebrite, Sea Limited, Mercado Libre, and Grab).

Veeva Systems, Fortinet, and CrowdStrike also reported very good Q2 earnings, and I’ll add my comments in this write-up.

AI’s ROI

In the last few months, there were doubts around the ROI of AI investments, and hyperscaler AI buildouts only went one direction: UP. Q2 results show some are already monetizing purpose-built AI agents, but others are only starting to use AI to improve internal efficiency.

Cellebrite AI - still early

Cellebrite’s AI product, Genesis, was only released in Q1 and opened for general use in Q2. It saw early uptake from 6 customers, with total ARR of $400,000. Less than 0.1% compared to the total ARR of $570M. Even smaller when considering the $12.5 billion four-year investigative AI addressable market that the new CEO, Shiven Ramji, cited in the Q2 earnings call. However, it is unlikely to accelerate in the near term given that Cellebrite’s clients are large agency buyers who have consistently elongated sales in the past 3 quarters, which is also why ARR guidance was lowered. Stock now trades at 24x FCF ex stock-based compensation for 15% growth.

Next,

Wisetech AI - still focusing internally

While the overall posture is to build AI capabilities into the deepest layers of CargoWise to automate import/export compliance, data entry, and exception management. So far, however, the impact is mostly internal and not expanded to customers (yet). Notably, AI was cited as having cut Wisetech's global workforce by 30%, helped write 90% of the new codes, and contributed $34M in annual run-rate savings. The company hopes its AI product will cut customers’ labor costs by 50% in 18-24 months. So AI monetization is either not an immediate push or an indication of an incredible market position. I still believe no other platform is as complete as CargoWise.

Veeva and CrowdStrike AI - clear signs

Among all sleep-well companies, I thought Veeva was the most AI-insulated business (mission-critical workflow of regulated data); AI was firmly on the product roadmap and acquisitions (e.g. Ostro, Copli, Yonalink)

In particular, Veeva created Falcon based on the Copli acquisition to automate "agentic labor" and handle repetitive, manual tasks in drug development.

Falcon has a lot of interest right now because it's very clear that that's high priority. Quick cost savings and compliance and efficiency. That's high on everybody's priority. I think there's a lot of interest in Falcon. We're the rate limiter right now. We have to get that product ready, start working with the early adopters, but interest in Falcon is very high.

As a result, the AI-fear we saw in Q1 was indeed overblown. My March purchase at $150/share (17x FCF) has now paid off handsomely at $290/share.

Now looking at Q2, it was another step in putting the AI fear to rest. Veeva posted the highest revenue growth in nine quarters! It’s business as usual.

In detail, I was pleased with R&D’s 19% YoY growth. Many new users adopted products across clinical, safety, and quality, despite AI. CRM also won big clients in Eli Lilly, Biogen, and Regeneron, reaccelerating growth despite the rollout of IQVIA+Salesforce CRM for life sciences solutions at the end of last year.

All in all, I feel lucky to have added more Veeva shares 6 months ago. Now, the multiple has bounced back to pre-AI SaaS-apocalypse level. So, I’ll be waiting till we see another multiple compression event to add more.

CrowdStrike - Security of AI

CrowdStrike also reported a reacceleration of growth.

Sales grew 26% to $1.47 billion, and adjusted operating margins expanded 350 basis points to 25.3%. Free cash flow margin stands at 26%, and the Falcon platform added a record $333 million in net new annual recurring revenue. ARR was also up 51%, and RPO grew 49%.

The underlying theme shows CrowdStrike’s consolidation strategy is working.

Falcon Flex grew 101% YoY. Non-endpoint solutions grew ARR by 39% YoY and now contribute 37% of total ARR. Customers with multiple modules continued to buy more modules, as shown below, making it harder to leave.

Cybersecurity is one of the clearest AI tailwinds, and CrowdStrike is poised to benefit.

“Ever since [Anthropic’s] Mythos, we have seen growth in our business, not measured by meetings or calls, but measured by ARR, and we do not see the threat landscape subsiding. Far from it. As demand for AI grows, so does our TAM, driving the need for CrowdStrike.”

“What we’re hearing from customers is they want to deploy more AI, but they’re being held back because of security, compliance, privacy, data protection type issues.” - CrowdStrike CEO George Kurtz

However, I can’t imagine buying more shares at the current ~130x free cash flow. The market is already pricing in a lot of perfection. I will dig in the latest investor conference (Fal.con) but for now, CrowdStrike is a target for trimming in the coming weeks.

Portfolio summary of Aug 2026

Here is how the portfolio looks today:

After 9 months of me ‘looking wrong’, the last three months are starting to prove that the hardest thing in investing is actually the waiting, and even harder is slowly buying when everyone is selling. No one quite knows when the bottom hits, but you have to know what is a reasonable deal for you personally, i.e., what is your financial goal, what is enough, what is the risk, how long can you wait, how confident are you about the value of the business you are buying, etc.

As always, time will tell. I have survived the 2008, 2015, 2020, 2022, and 2025 down years. Market cycles are a feature of long-term investing, not a bug.

By 2037 (when my daughters will have full control of the Sleep Well Portfolio), none of this would matter. I continue to put aside some savings for them and own more winners - ‘winner keeps winning ’- in both bear and bull markets.

If you are new to SWI, check out our FAQ and Owner Manual.

Subscribe now

If you are already an annual member, you can access the Sleep Well Portfolio and Thesis Tracker via the link below.

Take me to the Sleep Well Portfolio

Read more

My original thesis on Wisetech around Q1 2024 was simply that it was the hardest-to-replicate vertical SaaS business.

But after a year-and-a-half-long struggle around the ex-CEO-founder Richard White’s personal conduct and alleged monopolistic behavior after the e2open acquisition before the competition authority (Australian Competition and Consumer Commission - ACCC), shares have cratered to AUD 39/share from $141/share at peak.

Wisetech’s investment thesis is now a ‘special situation’.

I was wrong to catch the falling knife with my 3 additions on the decline. So, FY2026 results a few days ago provided vital information to set up a more methodological buy/sell plan going forward.

We will evaluate

(i) where the business stands after the corporate governance ordeal,

(ii) the e2open acquisition,

(iii) the pricing model transition, and

(iv) AI prospects.

Before that, let’s review my original thesis to see what’s changed.

• Wisetech is a very difficult-to-replicate Software-as-a-service (SaaS) business that solves the most complex logistics problems in a highly regulated industry, resembling the role of Veeva Systems (we own) in the life science industry.

• a backbone of the global trading system, with >50% of global trade going through CargoWise, >90% of the top freight forwarders and 3PLs are customers (47 out of top 50)

• a platform-based business that can extend organic and inorganic growth for decades, from sea freight to landbase, and from freight forwarding to customs and compliance, and warehousing.

• a rockstar CEO founder who has built the business to be profitable from day 1, then grown the company’s revenue, income, and free cash flow by 14x, 25x, and 21x since 2015, respectively. (stats as of FY2024 below)

• 85%+ gross margin and 40%+ EBIT margin, close to 0% attrition for the past 12 years

• valuation at the time was 75x free cash flow at the first buy and 100x FCF at the deep dive.

FY2026 results and how they compare

First, is Wisetech still the backbone of global trade? Mark the month of September 2028.

Read all WTC reviews here. The tracker spreadsheet is reserved for annual members only [Link].

If you are new: The Sleep Well Portfolio consists of time-tested, market-leading companies. We screen them with a rigorous checklist and regularly track their theses to determine when to buy. So far, we have made one mistake, but are happy with the current 13 holdings. All about us here.

Read more

You can read my deep dive here,

And why I bought it here.

Today, I’ll discuss why future growth will be even higher than FY2026, backed up by larger contracts signed and favorable renewal terms. I’ll discuss key-man risk, AI developments, and, last but not least, the premium valuation.

Read all PME reviews here. The tracker spreadsheet is reserved for annual members only [Link].

Subscribe now

1. Future growth will be higher

PME’s topline growth may be at 24% YoY, but forward growth is likely in the 30% range for the next few years. Contracts are signing faster than revenue is being recognized. We saw record 2025 and 2026 for new and larger contracts, record renewals 6 out of 6 all at higher per-transaction fees, and higher minimum volumes.

As it stands, PME has a total contract value of A$1.75B and a 5-year contract value of A$1.35B, showing the business's floor revenue.

Below are the 10 contracts signed during the year. 9 are in the US, and 1 is at a German university with Europe's largest cancer research center. The large ones are ‘full stack,’ some are ‘full stack +1’ (cardiology imaging, such as UCHealth Colorado), and all are cloud-based.

Below are the 6 renewals (1 extended - VISN23), with higher per-transaction fees and some with increased minimum volume. The company also maintained the 100% retention rate.

These price increases show clear pricing power and directly improve PME’s already impressive margins (74% EBIT and >50% FCF), 3 times higher than the next competitor (Sectra).

Looking at the contract details also shows that PME is successfully transitioning into a true platform, cross-selling core function viewers into archive, workflow modules, and into other imaging scopes from radiology: CT, MRI, X-ray to cardiology imaging: heart and vascular to lung, digital pathology, oncology, and ophthalmology.

Operating as a ‘full stack’ platform (viewer, archive, workflow, and since 2026, reporting) across multiple imaging scopes is exactly why PME will be the gatekeeper for AI applications.

2. The AI playbook

First, I just want to quickly review the three key risks I flagged in the original deep dive: key-man, AI, and valuation.

Key man

There are no changes in senior executive levels, but the list of directors again shows how important Dr. Sam Hupert, Anthony Hall, and Peter Kempen have been to the company. Any of them leaving would be a blow. An interesting read I found is that 30% of the total workforce is female, but 37% are female in senior executive roles. Additionally, voluntary staff turnover is only 2% in FY2026, far below the tech industry average of 10-15%. So not only are customers staying, but employees are staying put too.

On to AI

I find PME not only incredibly insulated, but also able to use AI to improve the business in many ways.

Read more

Understandably, Mr. Market didn’t like it. I didn’t like it. Shares sold off 30%.

Now, at ~10/share and a 14x Adj.EBITDA 2027, are these issues likely to recur? Has the bad prospect been priced in? What about the high stock-based compensation? What’s the real FCF ex SBC valuation? Has the quality of the business changed?

I’ll reserve the analysis for premium members. As much as shares seem like a good deal, I think owners must know the new CEO is inheriting a long-dated execution issue, and there must be a process to reset and improve.

To refresh your memory, here is the link to the deep dive, past reviews here. The tracker spreadsheet is reserved for annual members only [Link].

Let’s get straight to the point.

Q2’26 ARR miss flags a few issues.

1. The timing of Tom Hogan’s replacement is abrupt.

2. A recurring elongated sales cycle issue.

3. Guidance forecast cut after already being tightened in Q1.

Read more

The Sleep Well Portfolio consists of time-tested, market-leading companies. We screen them with a rigorous checklist and regularly track their theses to determine when to buy. So far, we have made one mistake, but are happy with the current 13 holdings. All about us here.

Shares are now up 70% from 2026 lows, and all it took was a stabilizing Adj.EBITDA margin rate of 0.7% from 0.6% in Q1’26 and Q4’25. The increase in Shopee's take rate was impressive given all the talk about competition. But what’s clear to me is that Shopee is self-sustainable and can turn that profitability tap at will, putting Garena in the back seat, and the focus on Monee, just like Mercado Pago of Meli, is all about their exposure to risk. We’ll focus on Shopee in this write-up.

Below, we see that Shopee has been Adj. EBITDA positive since Q3’24, but a deliberate deceleration in Q3 and Q4 last year threw Mr. Market's confidence in Shopee’s future into doubt. The stock fell from $200/share to $77/share.

We added twice during this period. Sea is now over 55% of my PA.

Read all Sea Limited reviews here. The tracker spreadsheet is reserved for annual members only [Link].

Q2’26 shows the first glimpse of a reversal in Adj. EBITDA. It was only growing by 0.1 percentage point to a 0.7% margin in Q2’26, and that was enough for the market to ‘value’ the business at $131/share. Management already spelled out that Q3 and Q4 will be even better!

With this solid momentum we are optimistic that Shopee will achieve the milestone of $1 billion in adjusted EBITDA for the full year. - Forest Li, Q2’26 cc.

Q1 and Q2 were $478M, meaning Q3 and Q4 will be at least $522M. And compared to H2 of 2025, which was $388M, implying 35% growth YoY.

I think the market will like it.

Now let’s skim through the headlines, and then I’ll focus on the key takeaway in each segment.

Q2’26 Group financials

• Revenue grew 48%YoY to $7.8B,

• Gross profit grew the same to $3.6B

• Adj EBITDA grew 11% to $917M

The drivers were Shopee’s revenue growth at 48%, Monee loans book growth at 63%, and Garena’s bookings growth at 15.5%.

Below, you can see that revenue growth for all three segments has been incredible over the past three years. Shopee accelerated from 20% to 48% this quarter, Monee kept growing at a high-speed lane at 60% for 8 quarters, and Garena has recovered from post covid lows.

But all eyes are on Shopee’s long-term margin. Let’s focus on this before moving to Monee and Garena.

Shopee - Long-term margin: 2-3%

Management's goal is to get Shopee margin to 2-3%. We are at 0.7%. How will they get there?

• Cost improvements (SPX and fulfillment cost structure improvements over time)

• Better take rate (ads or other forms)

a. Cost improvements

The roadmap has always been reducing costs through volume. Building a logistics network that is cap-light and dynamic. In Q2, the focus of the discussion was on fulfillment. Like Amazon FBA, Shopee shoulders all the logistics and warehousing for merchants in exchange for a fee.

Fulfillment is a double-digit business, and economics are improving

Improvements were driven by

(i) optimizing the cost structures.

(ii) operating leverage as it scales up. More sellers and buyers joining spread more of the cost out.

(iii) integrations between fulfillment and SPX reduced the frictions between how items in the warehouse move across the entire value chain.

Overall, management reiterates the strength of Shopee’s capital-light fulfillment strategy coupled with mid-size warehouses.

We can realize the cost synergies and cost benefits by running both the warehouse and logistics together. All these things help us to build the fulfillment businesses and helping us to gain the advantage of the overall platforms. Again, we run fulfillment business in relatively light CapEx fashions that we do not own the land, we do not own the warehouses. […] We are also experimenting more automation with our fulfillment center recently, which actually reduce our cost to run as well. That is still in the early stage.

b. Take rate

Take rate is a balancing act between

• how much reinvestment is required to grow the ecosystem,

• price competitiveness compared to other platforms and versus offline pricing.

• sellers’ profitability

Management see higher take take from commissions and from advertising.

Advertising - a clear gateway to 2-3% margin and beyond.

Ad revenue was up more than 70%, and ad take rate improved by over 90 basis points year-on-year, driven by more Ad-paying sellers, which rose around 45%, and higher average ad spend per seller, which increased more than 15% year-on-year.

The growth rate is on par with Mercado Libre and much higher than Amazon (20%) today, and higher than Amazon in the early days, which was 50%, but for a long period between 2013 and 2020. Today, it’s already 8% of Amazon’s GMV and 10% of its revenue, with a 60% operating margin.

So, Shopee’s ads are only in the second year running (started in 2024) and have a long runway to improve Shopee’s overall profitability.

Advertising is one of my core theses on why Sea will be a big winner in the future. Everything else (logistics, livestream, AI) is all catered to bring more sellers and buyers to the platform so they can use more ads on Shopee. In Q2, these numbers look strong.

• Average monthly new active buyers grew more than 35% year-on-year, a significant acceleration from previous quarters.

• Average monthly active buyers increased 18% year-on-year, and overall buyer engagement also continued to improve, with purchase frequency increasing by 8% year-on-year.

Other discussions - Market share, VIP, Livestream

Management observed that Shopee is maintaining market share, given its price competitiveness and wide assortment of items.

Competition was also relatively stable in Southeast Asia and Taiwan.

In Brazil, they claimed growth is well above the market growth levels.

Shopee was still very price competitive, a lot stronger than the competitor in the region, even after there is change on the take rate and the free shipping threshold. So we believe that for e-commerce businesses, the fundamental still holds. It is the price competitiveness of our assortment.

Additionally,

It is the completeness of our assortment. It is the fundamental structure of cost to serve, and it is experience of how the buyers can discover the product on our platform. All those things help us to grow faster in the market in Brazil. If you look forward, we still believe that Brazil has a long way to go in term of e-commerce growth. We are hoping to grow in Brazil in a profitable fashion, with the growth rate outpace the market in the coming quarters.

Overall, Shopee management believes it is growing faster than Meli (and perhaps TikTok) in Brazil, even after Meli’s drastic free shipping investment (R19, down from R79)

VIP Members: 15M and contribute 24% of GMV

• Q1 2025: >1m

• Q2 2025: 2m

• Q3 2025: 3.5m

• Q4 2025: 7m

I am very excited about the Shopee VIP program. That’s the e-loyalty program. And it is the most effective way to grow GMV, given that Shopee’s monthly active buyers and market share are already >300M, making it much harder to acquire new Shopee members now. But enticing existing members to spend more and more frequently improves the quality and predictability of Shopee’s revenue. Q1 showed that each conversion increases their spend per user by 30-40%.

Members are still ~4% of total monthly active buyers (est. 300M in 2023) but already contributed 24% of GMV in Asia in the quarter. Average monthly retention remains strong at around 80%, and members continue to show higher engagement, spending meaningfully more after subscribing.

Brazil went live in April and has already surpassed 1M.

Live stream - bringing in traffic (35% of physical goods)

Live streaming and partnerships with YouTube and Meta (Facebook, Instagram) are Shopee’s counter to TikTok. It’s been going on for a year, and Q2’26 shows it’s working. Traffic from outside of Shopee is ringing in.

Orders from live streaming and short-form video grew more than 50% year-on-year, outpacing the 28% total GMV growth and accounting for more than 35% of physical goods orders in Southeast Asia. Unit economics also improved sequentially as Shopee further optimized the marketing spend.

Shopee affiliate orders generated by these creators on Facebook increased by more than 85% quarter-on-quarter, with Facebook Reels proving to be a very popular channel to drive purchases. We have now extended our Instagram collaboration to all eight of our core markets, and we are seeing promising early results from Indonesia, the first market where we launched the partnership

Monee - Brazil expansion and credit risk

Revenue grew 59% to $1.4B and Adj. EBITDA grew 13% to $288m, driven by growth in the

• Loan book at 61% to $11B,

• Active credit users >40m, +34%YoY

There is a growth gap between revenue and adj. EBITDA. Management attributed this to loan mix, promotional loans to new borrowers, and investment in Brazil.

Monee has been a solid source of profitability and growth for the last 2-3 years. However, like Pago, we can see Monee is reinvesting aggressively to acquire new users in off-Shopee SPL lending (20% of SPL today), but that’s naturally lower-margin than on-Shopee.

Monee is also expanding rapidly in Brazil due to the vast potential. Pago has been operating for over 20 years and continues to post 40-50% revenue growth. Q2 saw a 68-75% growth in assets under management and the credit portfolio.

Excerpt from my Meli’s Q2 review

Assets under management grew 68% YoY to $23 billion. The credit portfolio reached $16.4 billion in Q2, growing 75% YoY. Total Payment Volume exceeded $100 billion for the first time. Monthly active users rose 30% YoY to 88M, with even faster growth in Brazil and Mexico.

Hence, Monee launched the ShopeePay app in Brazil, similar to Pago and NU Bank.

Going forward, Monee will no longer be in a supporting role to Shopee but will drive much of future growth and profitability. For now, we can expect Monee’s Adj.EBITDA to grow more slowly and leverage its clear data advantage from Shopee.

With our e-commerce user base, our e-commerce data, and also with our better credit scoring algorithm that's proven in Asia already, but of course, we customize for Brazilian flavor. We are able to broaden our product in Brazil over time. If you compare what we offer and what the other player offer, there are many low-hanging fruits that we believe that we can capture just by doing the right product structures, inputting the right data in our platforms to better credit scoring users, and just also with the license we applied, which is kind of as good as the others already in the market.

Importantly, Monee seems to have risks under control, thus far.

While the loan book grew by 61%, the NPL ratio of 90+ contracted to 1%. Management sees no correlation between loan book growth and the increase in credit risk. My only pickle is that they haven’t shared much about the composition of the loans, the size of the loans, the country, or the NPL of 15-90 days, which can indicate how credit risk is evolving early on. With the 90+ day, it could already be written off and may not show on the 90+ day metric.

One way I see it is that the provisions going up by 71.5% is worth watching. Not a red flag now, but I’ll be monitoring this.

Finally Garena,

Garena - a steady quarter

Nothing special in Q2, with stable 666M quarterly active users and 66M paying users. I’ll post two slides here for our future reference.

Free Fire is 8.5 years old and still strong, with partnerships and event adaptations. It’s still to prove if Garena and develope and publish another block buster game.

Conclusion

Sea is executing expertly, dominating a very fragmented region in Southeast Asia, and now growing profitably in Brazil. Lots of investments required but also a lot of potential. I am happy management is playing the long game.

I am most impressed by the sustained VIP growth and Livestream's contribution to traffic and order conversion, which proves buyers and users are more engaged than ever and suggests Shopee might have cracked the formula to play offense against TikTok Shop.

With ~4B Adj. EBITDA and an $80B market cap (~75B EV), 20% growth over the next 5 years, and management has started buying back shares ($90/share), Sea remains my largest position and is always a good place to add more shares on dips.

A gift for reading the whole analysis and not falling asleep!

30% discount forever

Disclaimer

As a reader of Sleep Well Investments, you agree to our disclaimer. You can read the full disclaimer here.

For a solid background of the business and industry, read all previous updates here. The tracker spreadsheet is reserved for annual members only [Link].

The Sleep Well Portfolio consists of time-tested, market-leading companies. We screen them with a rigorous checklist and regularly track their theses to determine when to buy. So far, we have made one mistake, but are happy with the current 13 holdings. All about us here.

I don’t want you to miss out on the portfolio's fast recovery, so here is a 30% discount for you to hop on the train before it takes off.

30% discount forever

Before I review Q2’26 results. Let me be clear about my point of view.

I believe Meli is the best technology business in Latin America. It is an incredibly customer-obsessed, adaptable, and high-growth business that has become the commerce infrastructure of the region; therefore, it is almost impossible to replicate.

The fragmented economies and less-penetrated ecoms and fintech in Latin America mean all parts of Meli above have 5-10 years of high growth ahead, yet they are high barriers to entry to new competition.

If I have to rank, Mercado Pago remains the biggest and clearest opportunity, then Advertising. Meanwhile, Envios (logistics) and Libre (marketplace) are the sticky mousetrap that acquire and retain customers.

I will focus on analyzing Pago today as it has become a ‘durable growth’ lever, but the market is questioning its risk management as it expands too rapidly to customers and merchant credit, away from its original setup, which was to process payments between merchants and customers, online and offline.

Ambition - largest digital bank

Meli’s ambition is to be the largest digital bank in the region.

The roadmap to get there is what Meli calls principalities: it wants to be the primary banking partner with users, serving all of their daily financial needs.

That’s payments, transfers, lending, savings, investments, and insurance.

How these products evolve is driven by its customer feedback measured by the annual NPS survey. That's the Net Promoter Score, showing how satisfied users are with Pago’s service.

In Q2, 70M active users said they were highly satisfied with the service; Meli scores an NPS (net promoter score) of 77, the highest in the industry in Brazil, Mexico, and Argentina. The feedback drives Pago’s roadmap and user experience.

But the conviction to expand Pago so rapidly comes from Meli’s scale, something we know as an ‘ecosystem’.

Together with the marketplace, Meli has 150M active users. A distribution advantage that allows Pago to expand at near zero acquisition costs. Then, a 26-year advantage of historical user data that are high-frequency, real-time, and granular. That’s the underwriting advantage that no one has in the region. I want to key on this a bit more before we move on.

• The data for an ecosystemic user is full access to individual navigation patterns and transactional data down to SKU levels.

• Meli knows not just how much customers are spending on each category each month, but also if they are buying running shoes, vitamins, and which reviews they are researching.

• The longitudinal data shows who is more creditworthy and what levers (credits, installments, cash back) allow the individual to make a purchase.

• All this data allows LLMs (AI) to excel even more than other banks, combining transactional, purchase intent, and industry-available data.

Importantly, in challenging or black swan events, Pago can assess which borrowers are most likely to pay and which merchants are likely to recover in challenging times. So when we think about the credit business, they are not giving out credit cards to users who need them; they are giving out the product to customers who shop and interact on Meli and find cashbacks, installments, or other features valuable. Pago can capitalize on this extra information.

Lastly, I can’t emphasize it enough: to achieve the ambition, Meli needs to think and invest long-term.

That’s what we have been seeing over the past year with results that look just like Q2. Strong top-line growth but lower profit and free cash flow.

Q2 results - 30th consecutive quarter of revenue growth above 30%

Net revenue and financial income grew 50% YoY, surpassing the $10 billion mark for the first time, the fastest in four years and also the 30th consecutive quarter of growth above 30%. But profitability compressed: operating margin narrowed 550bps to 6.7%, continuing the slide from Q4’24. Management cited continued spend on free shipping and credit cards.

The market didn't love it — shares fell the next few days (only to recover as Sea Limited reported a recovery of Shopee’s margin), reflecting investor caution about margin sustainability given the pace of investment, even as underlying business momentum is acknowledged.

Here are my good and bads from the quarter.

Good - deepen engagement across segments

• Commerce - saw deeper engagement, users purchasing more across categories, since the drastic investment in free shipping 4 quarters ago. Unique active buyers rose 26% YoY to 90M, and items per buyer increased 14% YoY. The trend seems durable.

• Fintech also saw deeper engagement as users parked more of their money in Pago. Assets under management grew 68% YoY to $23 billion. The credit portfolio reached $16.4 billion in Q2, growing 75% YoY. Total Payment Volume exceeded $100 billion for the first time. Monthly active users rose 30% YoY to 88M, with even faster growth in Brazil and Mexico.

Both segments did well, in my view, but active users who use both products- Meli calls them ecosystemic users- were even more engaged.

• Ecosystemic users were the most engaged group by far, making them the company’s most valuable segment. In details:

an ecosystemic user has 70% more GMV on the marketplace and 90% more TPV, double the assets under management. Very much engaged, and that results in better profitability. We also mentioned that the rate of growth of those type of users is the highest of any kind of users.

I think it’s growing 37% year-on-year. We don’t disclose the actual share of users. We just want to make sure that those are important users to us, growing very fast.

• Advertising: Revenue grew 73% YoY in USD, powered by greater seller engagement with AI tools and growth broadening beyond Self-Service as momentum strengthened in the cross-border trade (CBT) and Top Brands channels.

• Cross-border GMV is growing approximately 60% year-over-year, with triple-digit growth in Brazil, Argentina, and other markets, and above-average growth in Mexico, the largest CBT market by far.

• AI: Mercado Libre completed the rollout of its AI-powered marketplace search architecture across its five largest sites, with higher conversion and click-through rates more than offsetting the cost of third-party LLMs.

AI is definitely contributing to cost efficiency.

AI Customer service shrunk to 7,000 reps, from 10,000 four years ago, even though the business grew by 3x in that period. 90% of the interactions are done without a human participating in the issue.

2026 is probably the first year in many years in which Meli was not growing the engineering team. They have 20,000 developers that are using AI. A year ago, they were coding with help from AI. Today, human-written code is the exception.

A few optimistic quotes from the management:

Product development scaled from 8.4% of revenues to 7.2% of revenues year-over-year, even though it has this incremental cost of AI within it.

We are convinced that the 27 years of proprietary data that we have accumulated across commerce, payments, credit, and logistics position us in a good place in order to capture the opportunity that AI is bringing.

Everything to me was good; now the bad, or what I will be monitoring.

Bad - we need to monitor

Margin compressed by almost half to 6.7% compared to a year ago.

Driven by higher expenditure and investments in free shipping (4.7% operating margin impact) and building the credit portfolio (2.4% margin impact).

Behind the margin compression are actually two moving parts.

• Expansion of the loan book and a shift toward credit cards.

• Mixed picture of the loan quality.

Those two drove higher bad-debt provisions. Let’s look at both in detail.

Read more

I am excited to update you on Grab’s business, a position I acquired in October 2024. So much has been going on this year that may impact the business's long-term trajectory, such as the fuel crisis and FX headwinds, and the rise of EV fleets. I’ll discuss that elephant in the room - Autonomous driving and AI impacts.

But first, I’ll dig into the core flywheel customers and drivers, groceries expansion, advertising, and that promising fintech segment.

For a solid background of the business and industry, read my deep dive here and all previous updates here.

Disclaimer: As a reader of Sleep Well Investments, you agree to our disclaimer. Full details here.

If you are new to SWI, check out our FAQ and Owner Manual.

Build your Sleep Well Portfolio today

1. Core business: more users, more engagement.

The most important metric for a marketplace like Grab is the number of its transacting users and drivers. In the quarter, it delivered.

• 54M MTU, 16.7% YoY growth driven by a 19% increase in monthly active drivers.

Grab ended the quarter with 54 million monthly transactive users, growing 16.7% YoY and higher than the last 5-year compounded annual growth rate of 13%. There is a lot of room to keep growing, considering Uber also reported “adding more first-time users in the last twelve months than in any period in the last five years".

Grab doesn’t disclose how many drivers, but in Q1, a study found Grab had ~ 6 million partners (drivers + merchants); this number grew 19% in Q2. Very healthy supply and demand growth.

Also healthy is that daily transacting users grew faster than monthly transacting users, indicating higher engagement in the period. Grab highlighted that groceries (GrabMart) grew 42% YoY, 1.7 times faster than Food, and is still only 14% of Food’s user base; we’ll go deep on the revenue contribution and the likely significant size of this vertical to Delivery later, but the key takeaway now is that Deliveries have plenty of room to grow.

At the core of Grab, the flywheel is working nicely.

More users, more partners (drivers, merchants), and more engagement (more products).

2. Group financials and segments

The flywheel turns into growing financials

Group

At the group level, revenue grew 22%, GMV grew 21%, and Adj. EBITDA reached $168M, growing 54% YoY.

It’s great to see steady and resilient growth despite the fuel crisis (2% of Mobility revenue impacted) and the recently announced commission cap. This is the 18th straight quarter of EBITDA growth.

The management sees $4.125B in revenue and $730M in Adj. EBITDA for the year. There is plenty of growth left, so I believe the business is priced very reasonably for a $10B EV (15B market cap) today.

Ok, let’s go into a little more detail.

Mobility: 8-9%Adj. EBITDA margin longterm

Mobility GMV grew 18%, transactions grew 28%, but revenue only grew by 12%.

Management explained the lagging revenue growth was due to the net-off from the savers program (cheaper fares) and the targeted incentives to drivers to support the fuel surge.

The goal is to make Grab a sustainable workplace for drivers (more earnings - 4%YoY growth, and less impacted by the fuel surge - $7M targeted support) and an affordable place for customers at the same time. That meant revenue grew more slowly due to more discounts to passengers and more incentives to drivers. We also see that reflected in rising incentives, but it’s something I expect to decrease as the marketplace matures.

Despite the above, mobility margins this quarter were still at 8.6%, within the 8.5%-9% guidance range. Going forward, with the rollout of EVs, fuel costs will have less impact.

EVs actually have lower costs once the initial investment is made in the vehicle. This quarter, for example, we announced nine new fleet partnerships in Thailand. We've announced a new partnership with Wuling in Indonesia, and we've expanded charging access through the app for drivers in the Philippines. In nearly every market, we have examples of how we're becoming a positive catalyst for the introduction of cleaner vehicles into Southeast Asia.

Now to the key discussion,

Autonomous driving (AV)

AV is a structural risk to Grab; we can already see what is happening with Uber. The many partnerships with OEMs (Zoox - Amazon, Rivian or Lucid) in the US don’t guarantee that Uber will be the aggregation app for AV. It makes sense for the industry, but Waymo and Tesla could go alone and own the whole stack (physical and app). Why wouldn’t they? They have even better distribution than Uber, and they also own the supply of cars and technologies around it. They can also just acquire Lyft if they needed something like Uber.

Looking at Grab, the risk of disruption is less immediate.

First, what Grab has is time. The Southeast Asian region is highly fragmented (8 distinct government policies for AV), chaotic roads, low cost per ride, and low-wage drivers that make it less cost-effective for AV to compete. Most importantly, what is less discussed is…

The next section is for premium members; I discuss the idiosyncrasy of Grab, mapping capabilities, hybrid approach to AV, preparedness for AV. Then I also discuss the growing influence of Groceries and GrabFin. Subscribe for the real details.

Read more

Build your Sleep Well Portfolio today

Hi all,

In July, we saw more uncertainties and speculations around AI spending and return on investment (ROI). Micron (MU), Nebius (NBIS), and other AI infrastructure names fell as much as 50% from the recent peak. Mr. market was highly unstable. We, on the other hand, had a quiet month. The portfolio recovery continued, but not much happened. August and September will be more eventful months to report as the Q2 reports roll in, so I’ll keep this short.

AI’s ROI not clear

There is no doubt AI will revolutionize everything digitally in the next 5 years. Elon Musk goes further, predicting AI will also dominate the physical world by 2036, driving the greatest economic transformation. Combining both means we will enter a ‘quasi-infinite economy’, where digital intelligence with vast numbers of physical robots creates a framework where goods and services become so abundant that money and traditional jobs may become optional by 2036.

If you share Elon’s vision, I guess AI’s ROI will be substantial for all.

For now, even as hyperscalers continue to expand backlog (AWS expects to reach $500B by 2027), I can’t say ROI is clear.

First, these hyperscalers are also expanding their CAPEX budget at a rapid clip (Amazon increased its capex budget to $220B for 2026). Then, we also saw SpaceX and Meta early in July announce they were renting out excess capacity (on a 90-day basis), and thirdly, Ramp’s latest AI Index shows the median firm spends $11.38 per employee per month on AI, which is ~700x less than the top 1% of firms spending ~$7,500 per employee per month.

So, if there is ROI, AI usage and monetization are only tightly concentrated to a small group of power users doing important work.

For the majority, AI usage and ROI are still unclear, in my view.

That said, I remain closely watching Amazon, Microsoft, and the likes as an investment directly in AI. Certainly, the constraints in chips (China) and power and cooling (US) are very interesting to explore. I wish to share some of these names in the future.

Portfolio recovery continues

Despite the volatility and liquidity pull from AI, the portfolio’s recovery has kept the pace from June. It is happening faster than I thought, as it’s been just 6 months since the bottom of the software apocalypse, which was also when we evaluated the AI risk for all holdings (link here). The takeaway is that whenever the worst headlines are on the news, that’s an indication of the bottom. Since then, most of the software holdings have shown no signs of disruption, and shares have rebounded 20-30% (Veeva, Fortinet, CrowdStrike, Kinsale, Cellebrite). Our ecommerce and fintech (Sea Limited, Mercado Libre, Grab) are lagging, but I remain very optimistic about their future.

If I hadn’t had a large allocation for these names, I would be adding more already. History shows opportunities may come. The last drawdown of this scale lasted a year and a half after the war in Ukraine started. And, learning from the last period, all in all, we made 10 purchases and 1 sell. Almost all have been highly accretive to the portfolio so far. I believe patience will win.

Before we review the portfolio. I have a small change to make. I find the monthly Best Buys, Thesis Tracking, and Portfolio updates counterproductive to long-term investing. They encourage overthinking and tinkering in such a short period that makes very little difference to the business. So going forward, I will focus on the quarterly updates (Thesis Tracking), keep the monthly portfolio update brief, and transition to a longer-form quarterly portfolio update. My primary goal remains building a sleep well portfolio for my daughters, and the frequency of documentation makes it less productive, in my view.

Summary of July 2026

July was a non-event month after an incredible resurgence of software in June. Intuitive Surgical reported Q2; nothing changed in my view, despite a 10% drop, so the in-depth writeup on why I bought below the company remains relevant.

Deep dive here in case you want to know everything about the company.

Here is how the portfolio looks today:

After 8 months of me ‘looking wrong’, the last three months are starting to prove that the hardest thing in investing is actually the waiting, and even harder is slowly buying when everyone is selling. No one quite knows when the bottom hits, but you have to know what is a reasonable deal for you personally, i.e., what is your financial goal, what is enough, what is the risk, how long can you wait, how confident are you about the value of the business you are buying, etc.

As always, time will tell. I have survived the 2008, 2015, 2020, 2022, and 2025 down years. Market cycles are a feature of long-term investing, not a bug.

By 2037 (when my daughters will have full control of the Sleep Well Portfolio), none of this would matter. I continue to put aside some savings for them and own more winners - ‘winner keeps winning ’-in both bear and bull markets.

If you are new to SWI, check out our FAQ and Owner Manual.

Subscribe now

If you are already an annual member, you can access the Sleep Well Portfolio and Thesis Tracker via the link below.

Take me to the Sleep Well Portfolio

Read more

Build your Sleep Well Portfolio today

Summary of June 2026

June saw an incredible resurgence of our latest pick and buy - Pro Medicus is up 75%, even when bought at 55x free cash flow, similar to my CrowdStrike first buy level. I also bought Intuitive Surgical, at 50x fcf, so not far off but I believe these two will vastly improve the portfolio going forward. Detailed account why I bought these two recently here.

Other non-AI companies, including our SaaS picks, remain fundamentally strong, and their share prices have started to recover by 10-20%. As of writing, the portfolio is ready to make a comeback, despite not owning any pure-play AI.

After 8 months of me ‘looking wrong’, last month and this month are starting to prove that the waiting and, better yet, slowly buying in maximum fear are the most difficult things but are the right action to take.

As always, time will tell. And we have been through this before; recall that the last prolonged decline was a year and a half — it began with the start of the war in Ukraine around February 2022. But every time we practice discipline and patience, we improve as long-term investors.

Market cycles are a feature of long-term investing, not a bug.

By 2037 (when my daughters will have full control of the Sleep Well Portfolio), none of this would matter. I continue to put aside some savings for them and own more winners - ‘winner keeps winning ’- in both bear and bull markets.

You can expect Sleep Well to continue studying and adding irreplaceable businesses, patiently waiting for buys at reasonable prices (often at historic lows). I don’t promise monthly/weekly ideas or regular buy/sell signals. I provide only business-focused research; you need to make your own portfolio decisions.

Read my most recent picks and buys - SWI #24:

SWI #23:

Subscribe now

If you are new to SWI, check out our FAQ and Owner Manual.

If you are already an annual member, you can access the Sleep Well Portfolio and Thesis Tracker via the link below.

Take me to the Sleep Well Portfolio

Read more

This post covers two new positions in the Sleep Well portfolio: Intuitive Surgical (ISRG) and Pro Medicus (PME.AX). Both are up since purchase — 3% and 50%, respectively. More importantly, both are the kind of businesses I’ve spent years looking for: zero customer churn, near-unassailable competitive moats, and decades of growth runway that have nothing to do with macro noise or interest rate cycles.

After a difficult stretch, this is what a portfolio upgrade looks like. The full thesis, valuation work, and reasoning behind both positions are below — for paid subscribers. If you’ve been on the fence, this is the one to read.

Links to the full deep dive and valuation are provided here for context:

SWI #24: Pro Medicus (PME.AX)

SWI #23: Intuitive Surgical (ISRG)

Subscribe now

Intuitive Surgical (ISRG)

A year and a half ago, my family faced the biggest challenge we’d ever been through. My then five-year-old daughter broke her femur and needed either surgery or traction treatment. We were told the clinical outcome would be the same either way. What nobody told us — what I only understood much later, after spending months researching Intuitive Surgical — was how different the road there would look.

We chose traction. What followed was 8 weeks of my daughter lying completely still in a hospital bed. 8 weeks of me sleeping on a chair beside her, 24 hours a day, while my wife was at home alone with our second child, who was 6 months old at the time. I remember begging the nurses to let us recover at home. Not only was the care poor, but our family was fracturing under its weight.

Robotic-assisted surgery would have changed the experience entirely. That 8-week hospital stay: cut to 1–2 weeks. The 5-month full recovery: cut to 2–3 months. The NHS, stretched as it always is, would have freed up a bed worth at least £500 a day. And two parents would have got their lives back months earlier.

I didn’t know any of this then. I know it now. So having robotic-assisted surgery would have helped us return to our normal life much quicker.

Intuitive’s slides summarise the benefits to all stakeholders well:

And so, the firsthand experience and points above form a core thesis for why I wanted to own Intuitive Surgical and Pro Medicus (more on this later).

As a business, it’s incredibly easy to see why a long-term investor would want to own it.

***If you are new to Sleep Well, please read our owner's manual.***

Read more

Build your Sleep Well Portfolio today

Sea Limited (SE) and Mercado Libre (MELI) have approached my 6 years of ownership. In this period, I have experienced two major stock declines. While the first was broad-based due to the post-COVID hangover, the second and current one is more driven by a new, elevated reinvestment cycle as both lock horns in Brazil. The two positions constitute over 50% of my net worth, so I write from a very biased but very critical and long-term point of view. Past reports below if you want fuller coverage:

SWI’s SE coverage: [deep dive, 4th buy, 3rd buy, 2nd buy, 1st buy, AI risks, Sum-of-parts Valuation, tariff review, Q4'25, Q3’25, Q2’25, Q1’25, Q4’24, Q3’24, Q2’24]

Meli: CEO transition, AI risks, Q3’25, Q2’25, Q1’25, The Most Sleep Well Investment Of LATAM, Top pick for 2025, Q4’24, Q3’24, and Buy Alert.

In this article, I evaluate their investment decisions to determine whether they are justified and whether the Latin American e-commerce sector market is a zero-sum game.

I’ll start with Sea Limited (SE), going in-depth, then Mercado Libre (MELI).

1. Sea Limited (Decentralized and capital-light logistics investments, 2% ad take rate, 3% penetrated VIP program, Monee, Garena, Investments to counter TikTok Shop)

2. Mercado Libre (Cross-border and verticalization investments, Margin compression, Long-term impact, Battles against Shopee)

If you are new to SWI, check out our FAQ and Owner Manual.

If you are already an annual member, you can access the Sleep Well Portfolio and Thesis Tracker via the link below.

Take me to the Sleep Well Portfolio

Our most recent picks and buys are already up by 35%, and all buys since January are back in the green. Patience and discipline are finally showing results.

TLDR (too long didn’t read):

Sea is the largest e-commerce company with ~53% market share and the largest digital-only bank in Southeast Asia. Meli is the largest e-commerce company with >35% market share and >50% in retail digital media, and a top 10 digital bank in Latin America.

Both are clearly dominating their respective markets, born of building the largest distribution moat (the most users and traffic). The next leg of growth can be simplified to:

One, acquiring new users (buyers and merchants), and two, enticing existing ones to spend more, focusing on retention and operational depth.

In my view, acquiring new users from such a dominant position is harder and less productive than the second option. The emergence of TikTok Shop, Temu, and other cross-border platforms also forced Sea and Meli to double down on reinvestment, playing both defense and offense.

I believe both are reinvesting in the right tools (moving up to higher-value SKUs, building cross-border trade, improving service through in-house logistics and fulfillment, focusing on retention through VIP programs, and integrating financial services), tools that TikTok Shop can’t easily replicate. Early results are strong, and I am confident that both businesses’ free cash flow per share (the ultimate financial metric) will be higher than it is today. And as free cash flow per share diverges from the stock price (see the orange lines below), I will continue to accumulate shares.

Sea Limited

Between May 2022 and Aug 2024, Sea Limited traded around $75/share, with a market cap of $40B.

At the time, the market viewed the company as a grow-at-all-costs business, underpinned by subsidies, marketing intensity, and logistics partnerships. It burned $2B/year to support the e-commerce segment, Shopee. All the while, the main source of cash, its gaming segment, Garena, was shrinking rapidly. That narrative was enough for the market to forget that Sea also had a third segment, a very promising ‘digital-only bank’ in Monee, formerly SeaMoney. At this point, in the worst case, I valued Sea at $44B: 

• Garena at $12B (12x EBITDA, low range of industry 12-20x EBITDA),

• Shopee at $25B (2.5x EV/Revenue, lower than peers at 3x), and

• SeaMoney (Monee) at $7B (15x EBITDA).

Still, the stock dropped to a record low of ~$40/share, or a ~$30B market cap.

Forrest Li issued a rallying cry in September 2022 to turn the company into a self-sufficient, profitable business. He ruthlessly cut costs, canceled all non-core growth initiatives, and enforced a ‘zero cash compensation’ scheme for executives. It was an all-out change and marked the bottom of Sea’s stock at ~$40/share.

Six quarters later, Shopee reached ‘self-sufficiency’ and was $22M from breakeven. Then, two quarters later, in Q1’25, the company reached the $1B EBITDA mark ($4B annualized).

It was one of the most impressive turnarounds and showed the company can turn the profit tap on and off when required.

Q1’26 marks 7 quarters of stable profitability. The company also posted:

• >$7 billion of revenue, representing 47% year-on-year growth.

• Adjusted EBITDA exceeded $1 billion for the first time ($4B annualized)

With a market cap of $53 billion, that’s ~13x Adj.EBITDA for a company that is still growing at least 20% for the next 5 years. This is 2x cheaper than when I first pitched the company in Q1’2024 at ~25x Adj. EBITDA.

The question for me is: Can Sea earn multiple times the $4B in Adj.EBITDA in 2037, when I pass on the portfolio to my daughters? How much does it need to reinvest to exist alongside TikTok Shop and Mercado Libre (in Brazil)?

In other words, I care about Sea’s future earning power in 2037 and beyond. That’s very different from the market’s focus.

Sea’s Q1’26 financial commentary

On Shopee:

Gross merchandise value (GMV) grew 30% YoY and posted a $223M Adj EBITDA.

Compared to Q1’2024, when the share price was about what it is today:

• GMV has increased by 60%, to $37B vs $23B

• Revenue has increased by 64%, to $5.1B (take rate increased from 12-13%)

• Adj. EBITDA has increased to $233M, from breaking even.

Everything has gone in the right direction.

Underneath the numbers is a richer ecosystem of broader SKUs at lower prices, with financing options, faster delivery, and refunds, which broadens to more diverse shoppers and VIP shoppers. AI also continued to accelerate the business

around 80% of customer queries are now handled by our AI chatbot. AI usage helped reduce customer service cost per contact by around 30% year-on-year while maintaining high satisfaction rates — Q1’26

When the ecosystem creates more value, the company can improve monetization.

That’s why we saw a gradual increase in take rate to 14% from 12%.

At the center of it is ad revenue, which grew 80%, accelerating the pace of growth since it was first discussed in Q1’24. This is one of the most forward indicators of a marketplace's health. Why? Ads reflect merchants’ intention to invest in their business on Shopee. If they see strong traffic and strong buyer retention, merchants will invest more.

Ad revenue grew 80% and ad take rate increased by more than 90 basis points year-on-year.

Ad-paying sellers and their average ad spend both increased by around 35% year-on-year, reflecting the strong value sellers see in our ad offerings.

Average monthly active buyers increased 16% year-on-year, and the buyer purchase frequency grew around 12% year-on-year.

Looking back at my original thesis and the quarterly commentary summary below, we can see the rapid progression toward a 5-7% ad take rate.

Read more

Build your Sleep Well Portfolio today

Summary of May 2026

The relentless decline in software stocks is showing signs of coming to an end. As of writing, the portfolio is ready to look ‘hot’ again, despite not owning any pure-play AI.

CrowdStrike, the first pick, is now up 282%; Fortinet is up 154%; other software platform picks have also bounced 30%; I am most pleased with the latest pick, no. 24, which is also up 30%. It was arguably a tougher BUY decision given I entered at a 55x FCF valuation, similar to my first CrowdStrike buys.

After 8 months of me ‘looking wrong’, the waiting and slow buying-in are finally showing results. As always, investing is simple but incredibly hard. Discipline and patience are required but cannot be taught; they must be ‘absorbed’ slowly. Everyone will make mistakes, and the last 8 months have not been easy, but failing and going through pain are the only ways to improve as a long-term investor.

Going forward, my approach remains the same. Continue to study and add irreplaceable businesses, patiently waiting for buys at reasonable prices (often at historic lows). I don’t promise monthly/weekly ideas or regular buy/sell signals. I provide only business-focused research; you need to make your own portfolio decisions.

One of the examples is my most recent pick and buy - SWI #24:

SWI #23:

Before we celebrate, recall that the last prolonged decline began with the start of the war in Ukraine around February 2022. It lasted a year and a half for some of our businesses. We are roughly halfway there. The pain may return. But owning the best businesses is the best pain prevention. Their stock rebounded spectacularly between mid-2023 and mid-2025, and this time won’t be different.

Market cycles are a feature of long-term investing, not a bug.

Stay focused on your long-term goals. For me, it’s 2037 (when my daughters will have full control of the Sleep Well Portfolio). I continue to put aside some savings and own more winners - ‘winner keeps winning ’- in both bear and bull markets.

Subscribe now

If you are new to SWI, check out our FAQ and Owner Manual.

If you are already an annual member, you can access the Sleep Well Portfolio and Thesis Tracker via the link below.

Take me to the Sleep Well Portfolio

Updates on the latest buys and SaaS names

Read more

In this report, I dive straight into four underlying growth drivers that require owners’ attention.

Before you start, you can skim through the Q1’26 results here — I won’t discuss the obvious. We will focus on:

• Drivers and customers’ incentives, fuel prices, and commission decree.

• Financial services reinterates breakeven by H2’26 - showing risk control

• Advertising monetization - active advertiser growth slowing, durable average spend growth

• Capital allocation showing positive direction - $5B cash position.

• Positioning to future-proof from AI and AV.

Then we’ll revisit why Sleep Well owns Grab and whether a ~20x free cash flow valuation (~10B EV) is attractive/unattractive for the Sleep Well Portfolio.

Read our previous write-up on Grab for more context.

• Q3’25 - Earning update

• Q1’25 - Tariff Review, Q1 Update

• Q4’24 - Update, New Risks, TopSWI

• Q3’24 - Update

• Q2’24 - Deep Dive, 1st BUY

Disclaimer

As a reader of Sleep Well Investments, you agree to our disclaimer. Full details here.

If you are new to SWI, check out our FAQ and Owner Manual.

Build your Sleep Well Portfolio today

Read more

Build your Sleep Well Portfolio today

Summary of April 2026

The relentless decline in software stocks has stabilized in April, even as the uncertainties of the Iran War linger.

I did ‘nothing’, avoided following the emotion of the crowd, and instead focused on researching the SWI #24 pick.

Going forward, I see opportunities to add to several businesses (more details later), and I am very close to adding to our two recent picks, #23 and #24. Both irreplacable, AI-accelerated, and trading near historic lows.

Yes, they are still expensive compared to the average business, but in my view, the investment choice I make is akin to buying beachfront properties over ‘well-functioning apartments’. One will be even more undersupplied, while the other will be aplenty.

Read more here SWI #24:

SWI #23:

Of course, the challenge is that my bullish views will make me look wrong. The last time the market suffered a prolonged decline was the Ukraine war, which started around February 2022 and lasted a year and a half for some of our businesses. But then their stock also rebounded spectacularly between mid 2023 and mid 2025.

To me, these are mini market cycles, a feature of long-term investing, not a bug.

I can’t time them. So my strategy is always to aim for 2037 (when my daughters will have full control of the Sleep Well Portfolio). Focus on leaving out savings consistently, and own more winners - ‘winner keeps winning’. In 11 years, as beachfront properties become even more undersupplied, I will have more shares of them.

Subscribe now

If you are new to SWI, check out our FAQ and Owner Manual.

If you are already an annual member, you can access the Sleep Well Portfolio and Thesis Tracker via the link below.

Take me to the Sleep Well Portfolio

Updates on GRAB, VEEV, WTC, SE and plan BUY orders

Read more

Patience is required.

The good news is that my investment period runs until 2037, so what has happened in the past 6 months might not matter much in a few years, and definitely not in over a decade. I still believe that consistently (i) putting aside emergency funds, (ii) investing excess in high-quality companies, and (iii) staying invested through tough times will help me build a reasonable capital and knowledge base for my daughters’ future.

Today, I am putting money where my mouth is - I will initiate a starting position in SWI Pick #24 and add more over the next few quarters.

Learning from my mistake, this is a more predictable business and will be a smaller starting position.

• #24 is a technology leader whose alternatives are legacy solutions and AI as an accelerator instead of a competitor.

• The product is mission-critical to top-tier hospitals, 100% of whom renews long-term contracts lasting 5+ years.

• Durable growth opportunity with a large whitespace and a market cap of under $10B.

• Clean financials and prudent accounting, no debts, 50% of cash allocated to dividends, and capex-light.

• Risks include high valuation, looming succession planning, legacy solutions fighting back, and AI still bringing uncertainties.

• Why now? The business has been perennially expensive, but it has been bundled with the ‘AI SaaSapocalypse’, pulling the stock down by nearly 60% in the last 7 months from its peak, giving it a good time to buy in stages (dollar-cost-average) into a position, or at least to do the work now before we get a better entry point.

If you are new to SWI, here is the link to all write-ups.

What is Sleep Well Investments?

1. Deep research on a few exceptional niche leaders.

2. Patiently waiting and thoroughly following up for buying opportunities.

We will go through:

• Investment Thesis

• Why does it help us sleep well?

  • Mission-critical product, slow-changing customers

  • Founder-led management

  • Solid moats and barriers to entry

  • Organic growth opportunities

• Limited competition

• Risks - legacy fighting back, key man

• Valuation

• Sleep Well Scores

Read more

Build your Sleep Well Portfolio today

Summary of March 2026

It’s been three months of relentless decline of software stocks due to the ‘AI kills software’ narrative, and of course, recently, the Iran war.

The impact of wars is broad and deep when they're prolonged and out of our control; however, AI-fear narratives are defensible, particularly for businesses with high switching costs and are hard to operate, even if AI could replicate the software. In February, our detailed AI risk assessment found that 10 out of 12 holding businesses are AI-immune.

As a result, in March, I cut the most vulnerable - SPS Commerce and used the proceeds to add to the three most immune. I am planning to add to my latest pick, Intuitive Surgical #23, and my new pick #24, which I will have the write-up for in a week's time.

The challenge, of course, is that I will continue to look wrong. The market can remain irrational longer than my patience. The last time software companies had a prolonged decline was the Ukraine war, which coincided with the post-COVID SaaS business hangover. It lasted a year and a half for some of our business. But then they rebounded spectacularly between mid 2023 and mid 2025.

Today, the valuation of all these businesses is also at its lowest since their IPOs. They could still fall further; however, looking beyond 5-10 years, I believe they will benefit from AI, and I will be more prepared to capture the gains once it rebounds again.

Subscribe now

If you are new to SWI, check out our FAQ and Owner Manual.

If you are already an annual member, you can access the Sleep Well Portfolio and Thesis Tracker via the link below.

Take me to the Sleep Well Portfolio

Busy March, bought more AI-immune businesses

Read more

Build your Sleep Well Portfolio today

Hi everyone,

Yesterday, I bought a sleep well business that:

• Scored 15/20 points on my sleep well checklist.

• Possesses all four moats - scale/cost, network effects, switching cost, and brand.

• Competition is the best thing that has happened to this company, as FCF has compounded at 41% since this player entered its market over 10 years ago.

• On track to grow at least 20%+ for 5-10 years and available at <15x FCF and <3x Sales with a net cash balance sheet.

Subscribe to read more

If you are new to Sleep Well, please read our owner's manual.

Read more

I am buying more shares in the most AI-beneficiary businesses, one of which is trading at ~17x FCF (>5% free cash flow yield) and with a 15-20% free cash flow CAGR over the next 5 years. That means the expected return is at least 15-20% in the next 5 years.

The proceeds will come from the sale of SPS Commerce, the least AI-insulated business in the Sleep Well Portfolio.

Before we dive in, a few smarter people have already made the call about the current SaaSocalypse:

One of whom is NVIDIA CEO Jensen Huang, he said:

‘Wall Street has got it wrong’. - Q4’2026 earning call.

He argued that it was ‘the most illogical thing in the world’ to sell enterprise software based on AI-increased demand and utility. Jensen Huang said AI will not replace software but will increase demand for platforms like ServiceNow, as agents will run on them to boost productivity.

The Trade Desk CEO, Jeff Green, echoed with a few details.

Right now, Wall Street has an unprovable bear thesis — “software is dead.” There is fear that Claude and tools like it will enable developers to duplicate Salesforce or The Trade Desk. During a time when the market can’t prove that thesis wrong, the market and the sector go down. What I think they’re missing is that the best companies in the world work hard every day looking for an edge. We’re all looking for every possible practical application of AI, new tools and every other possible advantage. AI changed a lot, but it didn’t evaporate integrations, moats, trust, customers and tools — let alone people with esoteric expertise and obsessive passion to improve things.”

But are these views applicable to our businesses?

Absolutely. They also support my findings that AI is a net positive, particularly for businesses that possess proprietary & regulated data, platform solutions, operate a mix of physical and software operations, and are actively using AI to improve operations.

Link to detail findings in this article

TLDR (too long didn’t read):

For nine of my twelve positions, AI is actively expanding the value these businesses deliver. CrowdStrike benefits because more AI means larger attack surfaces. Veeva benefits because AI agents operating in biopharma need a validated, compliant data infrastructure. Cellebrite benefits because digital evidence becomes more voluminous and complex. GTT benefits because it is the only viable technology for the LNG infrastructure that powers the AI data center buildout. Dino Polska, Sea Limited, Meli, and Grab operate in physical-world markets where AI is a logistics optimizer, not a competitor. SWI’s AI Risk Assessment

Let’s dive into why I bought more Veeva Systems, and have ordered to buy more Mercado Libre, Cellebrite, and Sea Limited.

This trade alert write-up replaces my monthly Best Buys edition to remove duplications.

1. Why am I buying Veeva Systems?

Read more

Sea Limited (SE) holds a special place in my portfolio. The central thesis has always been a three-headed technology company growing more heads, leveraging the growing scale and data richness. When I first bought the company for my daughters (Sleep Well Portfolio), it had just risen (posting $1.5B in free cash flow) from the crushing depths of post-COVID. The group was burning $2B a year, the e-commerce arm, Shopee, was still loss-making; the gaming arm, Garena, was not seen as evergreen; and Monee (SeaMoney at the time) was still a ‘promising’ fintech venture.

Today, three years later, the company generated ~3.5B in free cash flow. Impressively, across the ~700M people in Southeast Asia (and +200M in Brazil), ~400M unique buyers and ~20M sellers are on Shopee, ~37M take credit (BNPL, personal loans, SME loans) on Monee, and ~100M play games daily on Garena. Half of the sellers were reported to earn their main income on Shopee (per Q1’25). And, 20% of total GMV was generated by just ~2% of buyers (VIPs). Incredible. Ah, then the current loan book is just $9B (growing at 80% rate), or <10% of a typical bank's loan book in Southeast Asia. Finally, only 10% of quarterly active gamers pay on Garena.

When you internalize the sheer impact of Shopee on sellers’ lives, the sheer purchasing power of VIP members (similar to Amazon Prime members), and the number of gamers spending time on the platform. I don’t see any issue with Sea's decision to invest in increasing the wallet share and penetration. Of course, it depends, but we will see what they have invested in (FY2025) and what they are going to focus on, which makes a stronger company.

To me, this is a multi-year, decade-long story, as the business's lifecycle is still in its early stages.

At the helm, Forest Li, CEO and founder, is a rare breed. I admire him for the ability to pivot (2015) to a platform, grow the company at the speed of light (2016-2021), then brake hard to rebalance and change course as soon as things didn’t work out (2022-2023). Every quarter, I listen to his call and see a humble, eager-to-improve CEO. So when he said that AI would help Sea become a $1 trillion company (15x its current size). I tend to believe in the story.

So, that’s the lens I use when I read the quarterly results. With shares declining to ~$88/share after the Q4’25 announcement, the road I am walking may not be shared with the majority of the market, but I am totally ok, as my additions will be more meaningful at these levels.

But let’s get one thing out of the way: why were shares down by -16%?

Shopee's Adj.EBITDA margin was guided to be

• FY2027: ~0.55%, compared to

• FY2025: 0.7% in 2025, and

• FY2024: 0.2% in 2024.

Further away from the 2-3% long-term goals.

But that’s not what I am focused on. There is an enormous underserved community in Southeast Asia and Brazil that Shopee monetizes on, but the time isn’t now.

Let’s go through

• 1. Brief FY2025 results

• 2. Shopee - where are the investments?

  • SPX Express

  • VIP program

  • Shopee Live + Youtube + Meta

• 3. Monee

• 4. Garena

SWI’s SE coverage: [deep dive, 4th buy, 3rd buy, 2nd buy, 1st buy, AI risks, Sum-of-parts Valuation, tariff review, Q3’25, Q2’25, Q1’25, Q4’24, Q3’24, Q2’24 update]

**Links to thesis updates, buy-and-sell, and deep dives—all in the Sleep Well Portfolio spreadsheet above.

If you are new, start here. Please use the FAQ alongside the SWI Manual to get the most out of your subscription.

Read more

Build your Sleep Well Portfolio today

Summary of Feb 2026

In January, I defended SaaS from AI risks for its incredible high switching costs and then further explained in the Best Buy series why they were also hard to operate even if AI can replicate the software part.

In Febuary, software continued to decline in Febuary. I have conducted a detailed AI risk assessment for all holdings. The finding showed AI is largely immaterial to 10 out of 12 businesses. However, I need to cut one holding - SPS Commerce (this week) and monitor one closely. I’ll also add more to two (patiently).

I could be wrong, of course! Valuation is at its lowest since IPO, and this particular business has also implemented its own AI capabilities.

However,

Its data structure is the problem! Deterministic, rule-based, and repetitive. SaaS with the same data structure, and with limited switching costs and network effects advantages, is likely to suffer the same fate.

Let’s look at the details of why I am trimming before we report the portfolio performance.

Subscribe now

• If you are new to SWI, check out our FAQ and Owner Manual.

If you are already an annual member, you can access the Sleep Well Portfolio and Thesis Tracker via the link below.

Take me to the Sleep Well Portfolio

Cutting SPS Commerce (SPSC)

Read more