I am not Chinese and this is a highly technological subject, and unlike maybe some writers that rewrite AI responses or directly insert AI text as their own, I will heavily use AI in this article, inserted in the form of screenshots, in order to gain relevant key points on the technicalities of the business in China.

In order to get the most accurate results, I will use Chinese AI: mostly Qwen from Alibaba and some Baidu own’s AI to try to challenge it.

I used screenshots from chats with AI when I wrote an article about Genomics in 2024, because I am not a genomics expert. The returns since that article are spectacular, although it was only a side portfolio for me.

So let’s embrace AI and use AI to analyse an AI company. This time I take my value investor approach and apply this successful framework to AI investing. This means that I want to see what we pay for, and what we get for that price, without making too many predictions about the future.

Baidu offers this kind of optionality where we do not pay much for AI.

Baidu is a search engine in China, but also a fast growing cloud and autonomous driving company. They lead in AI chips for efficient model running, and even have a service where they create an AI human.

Therefore, it makes sense to embrace the use of AI to try to understand this complex foreign company. I nevertheless spent hours trying to piece the puzzle back together. Report below:

Subscribe now

Read more

At the time, I argued that Allegro was undervalued in regards of total market cap divided by the potential to monetize a large user base, (one of my unique ways to look at valuations), as well as reasonably valued based on current earnings.

Luckily, this time the market reacted quickly and recognized the value, after a long period of drift. At the time I argued also that the Polish investor base was momentum and trend based rather than fundamentals based, and that a reversal could come but I did not know when.

Well, it happened.

H1 Results:

Subscribe now

Read more

Emerging markets, hated market and cheap.

I have never seen a cheaper environment for good stocks in Emerging Markets and serial acquirers, similar to cheap stocks in Europe in 2020-21 like Banks and Oil companies. (Repsol and Bank of Ireland my winners in that space).

We have a company that multiplied revenues by five since 2020, trades as an ADR, but is valued at four times forward earnings.

The market sees a risky company in a risky country.

I see a solid business that is slowing down after a large expansion phase. As a Financial company, I know that you do not invest for one year of profit and one year of growth, but for profitability over the cycle.

Despite being in a tough point of the cycle, we have a totally depressed valuation and profits deployed into capital returns.

We have a perfect macro backdrop which I will detail below in the short presentation. As this is a buy alert to inform my readers on time, a full investment analysis is not yet done.

Subscribe now

Read more

It follows the Amazon business model, even though the founder claims that they don’t and that they have a unique philosophy. Sure, the competitive environment in China is more difficult than the Amazon competition in the west.

Nevertheless, the valuation was ridiculous and the company remains impossible to displace.

The stock is slightly down over the period. It is fair that I review the company results and developments.

As a reminder, JD is a Chinese e-commerce company that is not number one in the market (Alibaba and PDD plus Tiktok shop), but has an integrated, high quality and fast delivery business model.

In my article, I argued that an older, richer China will eventually tend to favor this type of shopping experience over “the scrape for pennies” and gamified gaming of PDD and the less quality controlled Alibaba.

China still has a lower GDP per capita than Mexico and sits between Malaysia and Argentina.

Imagine the potential. Investors already tend to see China as a no growth developed country and this is a big misconception.

In Q2, JD grew revenues at a 12% CAGR in the past five years, and saw a decrease in 2026.

Net revenue was down -2.9% Yoy in Q2 2026.

• JD Retail is -5%, with a small decrease in operating income.

• JD Logistics is a good growth story, with 24% growth and a 16% increase in operating income (I reviewed JD Logistics last week).

• New business was down 45% but the loss, still important, is down 33% but remains substancial at 9,9 Billion RMB.

There was a reclassification of new business to JD logistics, therefore the decrease is lower in reality.

If we look at specifics, it is impacted by a subsegment: electronics and appliances.

The appliance revenues benefited from government subsidies in Q2 2025, therefore having revenues go down in this segment is normal, after the government subsidies.

New business:

When I reviewed JD, the new business of food delivery was not even really started. It had a dramatic rise but created huge losses due to competition in food delivery. I did not like it, but it does not affect the long term of the company.

The food delivery is JD entering not only into food delivery but instant commerce, with orders delivered within 30 minutes. This creates huge losses in the short term but also brings new active users to the ecosystem.

In Q2, New business had a more than 50% reduction in losses in food delivery. The company is “confident that the food delivery business will continue to see meaningful year-on-year loss reduction throughout the rest of the year.”

Joybuy, the competitor to Amazon in western Europe, “delivered encouraging sequential revenue growth in Q2 as our overseas supply chain strength and differentiated service offerings continue to gain traction among European consumers. While absolute operating loss expanded as Joybuy entered a rapid scaling phase, its loss margin narrowed sequentially”.

Joybuy is probably going to be loss making for several years but is building an exciting business. Unfortunately I cannot test it in my country of residence.

Jingxi group buying platform is in its infancy but experiencing strong growth.

Profitability

Recent investments in new businesses reduced the profitability of the group from a peak 4.1% in 2024 to 2.1% in 2025. The profitability is recovering to 2.6% as of Q2 2026, and we expect it to continue to grow in 2026.

“On JD Group’s profitability, Q2 marked a clear inflection in its trajectory, returning to healthy year-on-year expansion. Looking ahead to the second half of the year, supported by core business health and investment discipline, we are confident in driving accelerated profit growth at the group level.”

Investing in a long term growth company like JD is not an investment in quarters. It is an investment in decades. This is similar to an investment in Walmart or Amazon. I don’t want to predict that JD will be as successful as Amazon, but it is a similar investment, with a strong core retail business and many start up businesses that are starting now, with aims to become profitable in 5, in 3 or even 10 years only.

Shareholder returns:

On buybacks, we have repurchased around 17% of our outstanding shares since 2023.

Conclusion and investor psychology

When investing like this, you need to remember to have infinite patience and a cool head.

In the mean time, enjoy the shareholder returns in terms of dividends, and benefit from the buybacks.

The valuation is at 7 times forward earnings, despite losses in new initiatives. 4.3 times EV/EBITDA A complete bargain for a great business.

I heard many people saying “BUT.. BUT”. I hear these people all the time.

There is always a “BUT” when people are not value investors but trend followers.

The value investor buys when it is cheap. Holds, and has a large winner when the narratives turn from fear to apathy.

I have many such winners that were hated at the bottom. After they double or triple, they are not hated anymore but seen as regular stocks, and nobody talks about them anymore.

There are many other investments with similar risk rewards in my portfolio and premium section of the newsletter: it is available for upgrade below.

Subscribe now

I don’t really have too much things to write about JD, the thesis was the same 5 years ago, 1 year ago and it will be the same in 5 years. Entry valuation matters as much as holding philosophy.

Disclaimer: This is not an investment recommendation. Please do your own due dilligence before taking any investment decision.

• Stock market: Hong Kong

• Symbol: 2618

• Market Cap: $12 Billion USD

Focusing on JD logistics, the company is on the forefront of AI and robotisation.

JD has listed its logistic arm in 2021 and the stock is down 65% since the IPO.

I became interested in 2025-2026 as I constantly track the cheapest growth companies in emerging markets and worldwide. (upgrade to follow all of my work).

Unfortunately, as you can see in the end of the chart, it rebounded in 2026 before I could analyse the company.

Nevertheless, I see a 10 times multiple, 8.5 forwards, and earnings growth rates of 12% approx expected in 2027 and 2028, with leading AI and robotics technology. This is on top of a rebound in 2026 expected by analysts.

Let’s dive in what could be a massive opportunity.

Upgrade below for the full coverage on Emerging markets and a following on European hidden champions.

Subscribe now

Read more

The portfolio returned 12.87% year to date, after returning 30.8% last year in Euros, beating the SP500 handsomely. (9% and I think flat in real terms in 2025).

This is achieved even as I see the portfolio as an undervalued rocket ready to jump as emerging markets and small caps remain cheaper than ever relative to large cap SP500, because the fair value of most of the stocks in the portfolio is probably 80% above their current value.

Buying the cheapest stuff works, despite the occasional error. The agressive re-balancing I did in early 2026 sets it up for a good return in the future.

There are a few positions that did maybe extremely well and are not undervalued but they represent a tiny portion of the portfolio.

In July, I stopped being active around July 15 to focus on writing and organising summer trips, but I will share my last movements below, and companies updates. Writing this feels good as I love a portfolio review and checking back on companies.

The markets.

We have seen the AI bubble start cracking and prodigies like Leopold Aschenbrenner rise and fall already. I explained numerous times that we were in a bubble and it’s not over.

It is depressing that every “bubble boy” is popular and attracting millions in capital while having investing lifespans of less than three years and are all momentum traders that end up blowing up peoples funds and dreams.

On the other hand, serious investors like me and others on Substack, and most of you readers remain unknown and with small followings. We get little traction on the posts we put on social media, while we are beating these bubble guys over the long turn (Actually they are beating themselves), and we will probably be very popular after 10 or 20 years runs of doing serious value/quality investing. I have no doubt about this, fellow serious writers.

The movements:

(followed by top 10 and company news).

As always, I repeated the process that works, but being even more strict on entry valuations, adding to quality companies below ten times earnings, bordering 5 to 8 times earnings, and letting winners run, and mostly in Emerging markets.

The opportunity set in 2026 remains excellent:

Subscribe now

Read more

I saw a tweet saying that various of these AI companies are at 40 times sales and 200 times earnings or whatever. The meltdown coming is similar to the one we saw in Peleton and Zoom in 2022 after they were bubbles during covid.

Meanwhile, South East Asia contains some of the cheapest stock markets in the world, and investors going for Emerging markets will be making bank and in the hall of fame after it’s too late for the herd to follow.

5-Sea Limited.

(Nasdaq)

Sea limited is an e-commerce company with business in South East Asia and Latin America. It also owns video game assets like Garena Fire.

A fast growing fintech arm completes the lot.

Contrary to most of the deep value names in this list, it is a fast growing company.

In Q1 revenues were up an astonishing 46.6%.

However net income was only up 6%.

This is because the company is in the growth phase of its lifecycle. For example the ebitda of the E-commerce segment is down 15% in Q1.

SEA has 666 million quarterly active users.

The best way to valuate this company is by EV/SALES which is at 1.7 times, and EV/EBITDA at 13.3 times.

I think it is not the cheapest company, but it has fast growth and it is an essential holding if you want exposure to South East Asia. It is also cheap on an EV/Sales and on market cap (61 billion USD) vs (666 million) active users.

“We have had a strong start to the year. 2026 is a year where we are leaning in to deepen our competitive moats, while maintaining financial discipline. Our strong revenue growth reflects the effectiveness of these investments, and we are already seeing unit economics start to improve for some of these initiatives. We believe this is the right approach to maximize long-term value, given the significant runway for growth still ahead of us in our markets,” said Forrest Li, Sea’s Chairman and Chief Executive Officer.

Who is ready for my other top picks? They are not growth companies, but Illiquid defensive companies under 5 times earnings. Things that nobody wants but I hold. I have added two of the ideas this year. Subscribe for the Emerging value portfolio.

Subscribe now

Read more

A Tech and AI company at ten times earnings.

Whether It is a AI winner or loser remains to be seen in the long term, but AI is creating more competition for Text.

It is a former market darling from Poland that I keep an eye on.

This is a company known for integrating text chat boxes into your websites or apps, and now AI into the workflows.

But before that, a word about last week article. It contained an error in the way I understood the NAV calculation page, and I have corrected it. The opportunity is more at 8 times earnings rather than 5 times earnings. The writing on the company materials does not have the clarity that we need as investors. I am however happy to hold or add.

I was lately doing a survey of the Agentic AI marketplace for professional needs and went back to look at what the major AI companies can do: short answer: not so much yet. Agentic is definitely overhyped since what I found is that it remains inferior to old school automation (Zappier/Python/Javascript). The reason why it is inferior is probably not technological. It is that the AI companies are limiting processing available to users. Prompts are just getting refusals: “no I won’t do all this work”. Or worse “I did this work” - “Well no actually, I didn’t and invented the outcomes”.

We have the technology to control screens and operate actions with screen image analysis. But it is inefficient compared to running code that connects services through APIs, or analyses data in the database itself, and will always remain inefficient unless Ram and processing power drops like 1000x in price. The takeway is that Memory and efficiency is a real bottle neck, and that manual orchestration of automation will remain key in the future due to the need for efficiency.

Back to Text SA:

Text Allows you to integrate Chatboxes onto your website, use AI chatbots, have a ticketing system, all integrated with automation and reporting.

Before AI, Text was not unique but more a rarer type of company as it focused on seamless integration with different backend workplaces and websites, to allow your agents to talk to customers and help with service or ordering.

Now the field is seeing an explosion of startups.

Text rebranded, with a new website, a new ugly unreadable logo and a newer positive focus on sales increase rather than customer service.

They claim that adding these chatboxes with AI or human agents increase conversion on websites. It’s pretty obvious. The problems are however numerous: product offering, competition and go to market. We will see that later.

Let’s review the latest KPIs for the company and the competitive environment.

After quarters of declining MRR or monthly recurring revenue, the most recent update showed an increase. This was not by a growth in customers but rather by a discontinuation of legacy pricing.

The share of customers with high MRR is growing while small customers are dropping in terms of MRR.

Article continues below. Upgrade below to access the emerging market opportunities with extreme discounts available at the moment

.

Subscribe now

Read more

The market is changing, my articles are changing. The opportunities are more clear.

They don’t need very long articles.

Tencent is the company behind Wechat in China. This includes WechatPay and many services.

I never claim to be a China expert. I am a numbers guys and big picture thinker.

The number and big picture are clear: This company is a monster.

• Fintech: 1st in China and the world by active users.

• Social networks: 1st in China by number of users.

• Games: 1st in China by users and revenue, and globaly by revenue.

• Video, Music: first by paid subscribers

• Cloud: Here the company is not first but it remains a large player in a key growing industry.

You can see an overview of Tencent positioning in the following slide from the investor presentation.

Sometimes, you feel that a company is too big to be analysed in details. This is definitely the case with Alphabet and Tencent.

The revenue growth year on year is a solid 10-12% on average.

Social networks are about 16% of revenues as of Q1, gaming 33%, marketing services 19%, Fintech and business services 31%.

The company is powering the Chinese online communications, and being invested in it is investing in the Chinese online ecosystem. It will provide services and collect revenues from the long term growth drivers of the Chinese online economy. And of course from the AI economy in China and abroad.

The installed user base, the network effects and the AI engine ensure continued success in this sector.

I expect China to lead on AI due to higher innovation rate and lower cost of chips and eventually memory too in a few years.

Tencent is investing in AI heavily, with its own cloud operations and LLMs.

I always believed that AI in itself is not going to be a moat, but AI applications and applications to users and data at scale will be a moat. Tencent has everything to be a key player in AI, and workbuddy for example can generate extra revenue, even globally. The key advantages of Tencent AI will be the automation of workloads within the Tencent ecosystems, which is facing customers daily.

The best way to understand where Tencent is is through the annual report introduction:

We sustained healthy growth rates in 2025, as AI capabilities improved our ad targeting and supported more engagement with our games, and as our cloud business delivered improving revenue growth and profit at scale. Our highly resilient and cash-generative core businesses provide us with the resources to fund our increasing investments in AI, including recruiting top-tier AI talent and upgrading our AI infrastructure. The increasing intelligence of our HY 3.0 large language model, and the utility of our AI products such as Yuanbao, WorkBuddy, and QClaw, are encouraging early signs that these investments will unlock new opportunities. People enjoy consuming and being entertained, and derive satisfaction from creating and being productive, and it is Tencent’s privilege to provide AI services that can enhance our users’ capabilities across these dimensions.

Like Meta, AI is improving the core business ad targeting and engagement. The cloud business is growing. The company is in an AI investment phase. It will roll out AI within wechat and it will increase costs, while making the service even better, soon becoming a real life personal assistant of sorts.

Goldman Sachs has developed a bottom-up estimation framework that, by assessing the daily average token consumption of WeChat’s AI agents, estimates that if fully rolled out, the incremental inference costs would amount to roughly 5% to 17% of Tencent’s forecasted adjusted operating profit for the fourth quarter of 2026.

What is interesting is that the capital expenditures did not grow too much like the hyperscalers in 2025:

In Q1, Capex was up 16%.

Investment portfolio:

Tencent maintains a large investment portfolio compassing leading tech companies worldwide, and earns good returns on it historically:

As at 31 December 2025, our investment portfolio amounted to approximately RMB957,219 million or $141.25 billion USD — or 26% of the market capitalisation as of July 12.

We can see PDD, Sea limited, Spotify, Kuaishou (content platform), Reddit, etc, and many startups in the portfolio;

How about buying all of this at a crazy discount? Is it worth it? Searching for the cheapest EM stocks is what I do at the emerging value newsletter.

Subscribe now

Read more

My finances are composed of

1. The Emerging Value portfolio

2. The Dividend growth portfolio

3. Some bonds

4. Cash

The goal for the dividend growth portfolio was:

• low volatility

• capital preservation

• small income

• completely passive

• alternative to bonds and cash

The way to do this, was to buy defensive dividend companies that are undervalued. I build the portfolio during the Iran war crisis.

The results are good. The names did well on average, faster than I expected. The names that did badly only lost about 2%. There was some dividend income, although very small.

Below I will quickly review each holding. I did not make any additions.

Read more

I am starting to see good value in Poland again.

Regarding Central Asia, the Uzbek sovereign wealth fund was IPOed and listed in the London stock exchange but it does not trade at a discount and is not interesting right now, on top of having heavy politicised companies.

5-Allegro (Poland).

The Amazon/Ebay of Poland

The Q1 results were very satisfying with grow on all sides, leading to a 24% Adjusted EBITDA Growth. The international segment is still loss making.

Allegro started AI initiatives including a partnership with OpenAI to enable AI powered search and shopping experience.

Allegro entered health services and package travel markets: having such a large customer base enables the company to offer more services at a low acquisition cost: A core thesis to many of my investments in superapps.

As noted in my analysis, Allegro Pay fintech is a huge opportunity for a relatively small market capitalisation with a large customer base: we saw continued growth in this segment.

Loan origination up 37% YoY to PLN 3.9bn in Q1

Consumer loans contributed PLN 48m to Other Operating Income in Q1, up 125% YoY, as the share of self-funded loans grew up to 59% (+8pp YoY), reaching PLN 1.2bn

On June 25, it announced a share buyback program.

“25 June, 2026 | Allegro.eu announces the implementation of the Share Buyback Program (Phase 1) via open market transactions, for the purpose of share cancellation and reduction of the Company's share Capital”

The stock rebounded and is in my opinion set to become a darling of the polish stock market, after a disappointing IPO. While not extremely cheap at 18 forward times earnings, it is cheap on a market cap basis compared to the user base, especially considering the fintech business that is just starting out.

https://about.allegro.eu/wp-content/uploads/2026/05/Allegro-investor-presentation-Q1-2026.pdf

4-Halyk bank (Kazakhstan)

Listed in London, the second largest bank of Kazakhstan trades at 4 times earnings and an approx estimated yield of 12-14%.

To adjust for the economic cycle, we see that the price to book value is 1.1, reasonably in the cheap range but not ultra cheap. These Kazakh banks seem to be over earning compared to the rest of the world.

It has a super app with 8 million monthly active users, but most of the income is net interest income, with b2b income very important at 57% of loans, providing exposure to the energy and mining focused economy.

I did not release a deep dive of Halyk. If you are interested, I am sure you could find articles and reviews online.

Below are 3 quality companies, each time under 10 times earnings. This is what I do at emerging value. I find and buy the cheapest earnings out there.

Subscribe now

Read more

Since I started the newsletter I am at 17% CAGR but the SP500 also.. interesting times where I really like my performance but the SP500 did extremely well also.

The market:

I believe that there is an insane bubble in the semiconductor buildup sector, worse than everything I saw in 15 years, but that the rest of the market is pretty fairly valued. Quality and “compounders” are undervalued.

Activity:

I did a lot of reshuffling in this half, mainly to position it for the years ahead and to focus increasingly on the current opportunity set: in my opinion emerging market tech is way too cheap with multiple names under 10 times earnings. I don’t care really much about the market and the economy if I see 100% upside to most names just on current earnings.

I wrote in the past that buying and never selling is a good strategy, but greed is making me more active because I see that I can capture some extra gains by positioning my portfolio cheaper. I want to go for that other 100% and then the million.

I kept concentrating in the highest opportunities, despite my large number of historical positions.

Movements:

I sold some slow performing nano caps because of the low relative attraction, higher capital increase or key man risks, versus companies with > 1 billion USD in market capitalisation.

I sold some smaller positions in small caps: Portmeirion (large loss). Portmeirion is a small cap holding. I believe Portmeirion is still cheap, and it conducted a capital increase. I did not have the capital to contribute to both Intrum and Portmeirion capital raises and I focused on the better one.

Savonnerie de Nyons: Largely flat, nano cap with growing earnings and dividends but I prefer larger caps now in general due to liquidity and less governance risk.

So, legacy European small caps is a segment I am reducing.

Ly Corp: Still bullish on this Japanese name, but concentrating the portfolio a bit further: Japanese tech conglomerate managed by Masa Son that owns Yahoo Japan, Line Corp, PayPay and grows with good management and acquisitions. AI play with Line corp. Cheap on price to cash flow because earnings are not showing the real earnings power. P/E 15

Additions to: 3 growing EM companies under 10 times earnings (details below). But one of them is recently reviewed Inter&Co.

In the full review, I will show my recent additions, the whole portfolio with Emerging market value stocks and hidden champions, then a few sentences on each position.

The weighted average P/E ratio is 14, EV/EBITDA is 6.8, net income margin 13.6% and estimated revenue grow 19%. (Koyfin). I believe that the most correct number from this Koyfin estimate is the EV/EBITDA average.

There are about 50 positions and the top 10 is 48%, and the top 20 is 69%.

Subscribe below if you are interested in the cheapest global value names.

Subscribe now

Read more

Unfortunately or fortunately, we have been taken out with a small premium, at around 10 times earnings. This is why I am getting tired of French nano caps, the full upside is often taken by the controlling family with these take private offers. Before the take private offer, there has probably been some insider trading as the stock price broke out in a strange way.

Over a long time period, the breakout is pretty clear in early 2026. While these cannot be directly the majority owner of the company, these can be people who heard that something happens, or guessed correctly after witnessing some meetings and contacts prior to the take over offer.

This phenomenon is often talked about by French value investors such as William Higgons, a legendary investor that noticed this effect before.

In his fund reports (such as those for France Small Indépendance), Higgons and his co-managers frequently track performance catalysts. They regularly document stocks that begin to climb purely on takeover rumors before an official bid is materialized. When the market starts to realize a high-quality company is trading far below its intrinsic value, volume and price often drift upward before an official OPA announcement. (Gemini).

I can’t find the exact quote because I heard it in Video interviews.

I have never written such an email before, but these days as I was reviewing my holdings for the H1 2026 portfolio review, I see that another family controlled French nano cap is having a strange breakout in my portfolio, and with volume.

Exclusively for premium subscribers, due to privacy needs, here are the details of the buying pattern I have noticed in this nano cap, with a strong price rise:

Subscribe now

Read more

Market cap: 2.4 billion USD

Price: 5.30 USD

Brazil is one of the most exiting Emerging market in the world valuation wise. I reviewed companies like Sendas Distribuidora, Pics, and Brookfield corporation lookalikes such as Patria and Vinci.

All these companies are quality financial businesses with growth and deep discounts to global peers.

Today, I want to review Inter&Co, a another Brazilian Fintech, similar to the company reviewed two weeks ago, Pics NV.

As a prelude to this review, at emerging value, I focus on valuation first rather than narrative. I don’t hype companies that have XX% growth and extrapolate this growth infinitely in the future. I try to have a cautious view and take into consideration bearish arguments. I look for investments where the growth is the extra bonus that provides optionality.

Inter&Co is one of many that have these characteristics in Emerging Markets at the moment. While everyone looks at “1999 tech bubble 2.0 stocks”, the rest of the market in quality, growth and quality growth in Emerging markets is incredible. I don’t invest for anything else than shareholder yield and these yields only are bargains.

Upgrade below for access to the whole oportunity set

Upgrade

History

Inter&Co was started in 1994, which is old for a fintech. This is because it was not a fintech but a financial institution that was originally part of a larger group. It started offering business loans, payroll credit.

In 2002 it left the larger group, and in 2008 it received a bank licence in Brazil.

In 2012 it formed Inter Group with the company Inter Seguros (insurance).

2015 was a milestone when Inter launched the first digital checking bank account in the country.

in 2016 the app was launched, and branded cards appeared.

in 2018 Banco Inter was IPOed in Brazil.

It raised money again in 2019 and 2020 and moved towards a super app.

in 2022, it moved to the Nasdaq and started offering services in the USA.

Along the way, there have been small tuck in acquisitions to acquire assets and product offerings.

This historical development was successful in terms of active clients additions.

In terms of financials, it is only since 2024 that Inter&Co reached the scale to enable full profitability.

Business overview

Similar to Pics, Inter&Co is a Brazilian fintech, an online bank.

It offers accounts for buyers and merchants, together with credit cards.

It also offers savings and investing products, and generates interest income from credit given to customers.

A home page could look like this (they are personalised for the clients), with Pix (P2P payment network in Brazil), Pagar (Pay), Cartoes (cards), some news feed with social features, an investing tab, a shopping tab.

Inter&Co intents to become a super app.

The app is personalised based on client activity, asset base and product activations.

Now that we know the basics about the business, let’s review the historical revenue development. Inter&Co shows strong growth quarter on quarter.

Net interest income is 77% of revenue, and fees are 23% of revenue. In the interest income side, that includes revenue from customer loans and revenue from investments held by the company.

For example in Q1, it held 46 billion in loans to customers and 27 billions in securities, as well as 4 + 4 billion in cash and funds due from financial institutions.

Inter&Co claims to focus on what they call sustainable lending: loans with lower interest rate than Card revolving credit and personal loans that can carry outrageous 100% and more interest rates in Brazil.

Like Pics, it is now turning more more on Secured loan markets: car loans, payroll and mortgages.

Beyond the positive self description of the business, we can look at the latest financials (Q1 2026) to see the composition of the loan portfolio:

• Mortgages are 24.5%

• Home equity is 10.3%

• Payroll+Personal loans : 25.6%

• Credit Card: 31.3%

• Small and Medium businesses: 8.3%

Together, 67% of the loan portfolio is secured.

The company is moving towards a super app strategy, with shopping, insurance, investing and many other functions. It has an integrated loyalty program.

Like Pics, the extra products carry frequency, offer monetisation opportunities, and the Inter&co customer base is still not fully penetrated (44 million clients).

None of these Fintech super apps are very differentiated. All the Brazilian players are trying to become superapps and I don’t see Inter&Co becoming a monopoly or a duopoly. Nevertheless, even if no winner takes all, Inter&Co and other fintech can take marketshare away from traditional banks and offline financial services.

Strategy

The main focus is to increase the client base, and to sell more products to the existing client base. For example it has only 9 out of 44 million clients who have credit products.

There remains ample potential to convert the other 35 million clients to credit products, as well as onboard new clients.

However, the Brazilian customer is reported to already be over leveraged, and this is going to be the difficult to fill that gap.

The plan to increase the credit penetration is supported by the following products:

The ROE is at 15.5%, but they plan to improve this ratio to 28% by 2029, notably by improving underwriting, increasing leverage, cost efficiency, growth, etc.

(Pics plans to get the ROE to 30% in comparison - this shows that similar business models can aim for similar profitability targets and to me validate their targets).

International expansion: Inter expanded to the US first because there was a need from their clients to have a USD global saving account and investing account. 5 million customers already use this account. Brazilian expats in the US also use it, and Inter got the naming rights for a Orlando MLS club (soccer) stadium and is actively expanding across the Miami region with sponsorship activities.

In the mid term this global account capacity can attract customers in neighbouring countries, as heard in a recent CEO interview.

Capital allocation: Inter is focusing on growing its loan portfolio, but still pays a fast growing but small annual dividend.

Management and shareholding

This is another family owned company. Here, the Menin family.

The controlling group controls 26% of economic ownership, and 78% of voting power. Softbank has 13.70% of economic ownership but only 4% voting power.

The other investors own 56% of the company but only 17% of voting power.

We have here a company completely controlled by Costellis International Limited, the (Controlling Group).

Before being led by Joao Vitor Menin, it was led by the Co-Founder Rubens Menin. The Menin family ownership led the large growth in revenue and net income over the past decade and it’s fair to say that they have been excellent in execution.

There is no recent insider buying.

Risk

Risk 1-credit losses

As with any credit institution, the risk is that non performing loans -NPLs- eat the net income and make the bank loss making.

And we saw in Q1 2026 the NPL increase a lot, especially 15 to 90 days. It was officially driven by private payroll growth and seasonality. Seasonality that we can confirm on the chart.

Cost of risk if nevertheless rising: “Cost of Risk -- CFO Santiago Stel said, “we are expecting something closer to six percent [for the year]” compared to previous guidance of “five percent–five point five percent.” The company adds that this increase of cost of risk is also caused by a voluntary push into some products.

Read more on risk here:

Risk 2-macro economy: here is a copy and past of my section on Pics reviewed two weeks ago: note that macro economy is not what I view as my specialty as it is widely unpredictable, according to the GOAT Warren Buffett.

-Macro economic view: I view the Brazilian interest rate as too high for the past two years, caused by a public deficit and to defend the currency. However, this is causing a slow down in the economy and a wave of stress for companies and customers. I am expecting some stress to continue until the central bank decides to cave in an reduce the central bank rate or SELIC rate.

The central bank has indeed started cutting rates in 2026, but the war on Iran is hurting this normalisation route. In any case, a worsening of default rate and economy would lead to rate cuts. Rate cuts in the USA would also impact.

Net interest margin is very high because interest rates are high in Brazil. I expect that rates could go down, or stay high depending on war on Iran and public spending. Brazil has a leftist president with Lula.

However, funding costs will also decrease when interest rates go down and net interest margins should decrease but not that much. Credit losses should also decrease when rates are lowered to 10 or 5%.

A lower interest rate should make savings accounts offered by Inter less interesting, which is a negative. I see in reviews of Inter and competitors, that Brazilians are very focused on earning high interest rates. In Europe here nobody cares about earning 2% and focus on the stock exchange more. On the other hand, a lower interest rate will send all Brazilian stocks higher due to a search for yield.

Quipus Capital discussed the macro economic impact of Fintech in Brazil of interest rate changes, in an interview I posted on the newsletter in February. I really like this clear summary from the interview, from someone who has studied Brazilian economics extensively. I recommend his publication.

I

Risk 3-Competition

What I wrote about Pics also applies here: this is a very competitive industry in Brazil.

Valuation

Let’s review the Brazilian Fintech peers and some regional banking peers: (June 20)

The whole sector appears undervalued. Nu Holdings is standing out as the clear leader in Brazil and Latam fintech. The price to book reflects the superior profitability and the price to earnings a premium.

In my opinion Inter&co and Pics Nv represent a much more interesting opportunity due to the large undervaluation.

Valuation for a growing online bank like this could be on P/E ratio, but there is a certain cyclicality in the business.

I like to look at price to book, and at 1.2x, it is reasonable.

looking at P/B is a way to normalise earnings, avoid over earnings, and look at a cycle neutral valuation.

That would be a 10% earnings yield.

Now with P/B of 1.2, the implied P/E could be 10 with a ROE of 12% for example. But our ROE is at 15% already!

I explained how a company with credit card loans like Capital one, although not perfect, is an interesting financial peer, because we see that in a stress scenario, ROE drops to 0% but for a very small time period before recovering.

I expect the ROE to touch the 29% goal and even go higher, but with temporary crisis in between when there are non performing loans.

Therefore, I view the long term conservative ROE for Pics (And Inter) conservatively in the 20% range and the corresponding valuation should be 20 times earnings and two times book value: This should be around 9$ for the price to book or 15$ for the price to earnings, since the price to earnings (Forwards) is at 6.5.

Looking at projected earnings growth, earnings per share are planned to grow 40% in 2026, then 30% in 2027. In 2025, it grew 45%.

I am not certain that the fast earnings growth we saw in the past five years can continue very far after this, due to the macro economic situation and the market saturation with fintechs, but this represent extra bonus upside.

Conclusion

I am not here to take some earnings slides that show 40% growth and be over enthusiastic and tracing a trendline into infinity.

I want to have a more nuanced view. After reviewing the company, I am only able to project that type of growth for 1-2 years due to market and credit saturation hitting soon, default risks, and a potential for a lower interest rate in Brazil.

However, I think that the risks are pretty priced in at low price to book, while the opportunities are completely ignored by the current price: future credit growth, customer expansion, international expansion and continued gain of marketshare versus legacy banks. A double in long term earnings totally possible, followed by lower growth.

At 6.5 times forward, it is a bargain.

Long Inter, Pics

Note: this is not financial advice, but my opinion on the company. Anyone should do their own due diligence to confirm a company thesis presented and form their independent opinion.

If you want all the EM stock reviews and the portfolio ideas, as well as some DM hidden champions, please upgrade now. Thank you.

Subscribe now

There will be no weekly update early this week like usual. Since my last analysis of PICS NV published on Saturday, I want to keep my mind focused on this industry. I don’t want any distraction from other company news.

I have started an analysis of competitor Inter&Co, which will be my next article published within a week maximum.

It should be ve…

Read more

• Pics: NASDAQ

• Price 10.82

• Market cap 1.42 Billion USD.

It has already halved in value since the IPO. The market has been brutal for Brazil, for Fintech, and for Brazil Fintech in the past months.

Now valued at 1.4 billion USD, PICS has 44 million active clients. Just looking at these two metrics, it seems like the deal of a lifetime. Is it really?

History

Pics was founded in 2012 in Vitoria, ES:

I also thought it’s Vitoria, España, but ES means Espirito Santo and it is here on the right above Rio:

Anderson Chamon, who is now also vice president of PicPay, was 31 years old at the time and had many annoyances when making payments on a daily basis. Each place had a different logic: some only accepted cash, others only credit and debit cards , and some even preferred checks (this was twelve years ago!).

The idea behind PicPay was simple: pay for everything with your cell phone just by scanning a code

Anderson (Chamon), que hoje é também vice-presidente do PicPay, tinha 31 anos na época e muitos incômodos na hora de fazer pagamentos no dia a dia. Em cada lugar, era uma lógica diferente: alguns só aceitavam dinheiro, outros só cartão de crédito e débito, e alguns preferiam até cheques (isso já faz doze anos!).

A ideia por trás do PicPay era simples: pagar tudo pelo celular só ao escanear um código

So, PicPay was created in Central Brazil by three entrepreneurs. It was a pioneer in P2P QR code payments, B2B money transfers.

And growth was slow. very slow. “Among the few people who already knew the app, Chamon knew everyone’s name.

Many of the users were customers of a café in Vitória, which was the first establishment in the city to accept PicPay as a form of payment: Café Bamboo.”

In 2017 it became part of J&F Group, after the first investments in 2015. The J&F Group is one of the largest business conglomerates in Brazil, present in many industries such as meatpacking, media, infrastructure, energy, etc.

In 2020 it got a payment institution licence that enabled it to launch credit cards and credit products.

Pix was launched on the same year. Pix is a Brazil based payment system, and it is not related to Picpay or Pics NV the company.

2021: Guiabolso acquisition: the company then ceased to be a standalone offer and the products were integrated into PicPay.

2022: PicPay became a financial institution, giving it more option to offer loans.

2023: BX blue acquisition: another fintech specialised in payroll loans, and integration of Banco Original retail customer base, as it was facing more and more non performing loans, as a move to streamline operations and cut costs.

2025: Kovr acquisition: a digital insurance company.

2026: IPO on the Nasdaq, raising 490 million $.

The revenue and income development has been a large success:

Business overview

So after avoiding Nubank, Inter&Co for years, am I excited or not? We will see in my review below:

Subscribe now

Read more

.

While I don’t look for new investments actively in Europe, I keep an eye on my watchlist and portfolio investments. I have included the 1x P/E serial acquirer: high risk high reward.

6-Ambra SA-Poland

Ambra is the leading wine and sparkling wines companies in Poland. It is a company I rebought in 2026. After lackluster performance, it showed good growth in 2025 December half earnings. December is the most important period for this company.

Ambra produces wines, sparkling wines, liquor, non alcoholic beverages. What I like about this company is that it is acquiring small companies, and growing the business organically with a strong marketing team constantly launching products.

The last quarter earnings were not great, but H1 2025 showed good recovery.

Ambra is going to benefit from productivity and savings capex done in the past two years. The rise in oil price is nevertheless temporarily impacting earnings costs.

Ambra trades at 4.5 times operating cash flow, and the investments are generally growth investments, towards non alcoholic and soft drinks investments.

A cheap boring cash machine at 8 times earnings with a 6% dividend.

5-Kontron AG-Austria

Kontron is an Internet of things company. It develops products (embedded computers and parts). It trades at 13.8 times earnings, but a more accurate 8 times EV/EBITDA, because earnings are affected by one offs, and less than 9 times operating cash flows.

This is a project related company with some recurring revenue.

In Q1, revenue and EBITDA were flat, up 1.7% year on year. (We had 7.7% of revenue divested also).

What is interesting about Kontron is that it is at the beginning of a bell curve of internet of things equipment:

Computers and devices working in autonomy and communicating safely over the internet. They require hardware, software and protocols. Kontron is also present in smart transportation, defense and green energy.

The green energy segment is seeing sales drop but they will likely recover over the next few years.

Divestments are hiding the picture of long term growth trend: “EUR 579 million of low margin biz divested/discontinued over last 3 years.” This represent more than 30% of the annual revenue.

Kontron finds the company undervalued and launched a buy back program in 2026. With low leverage, it can continue with acquisitions expected in the future.

https://cms.kontron.com/kontron/ir/reports/ec-2025_ir_presentation.pdf

4-SDI PLC- UK

A left for dead serial acquirer in scientific instruments and scientific/high tech manufacturing businesses at 11 times earnings.

It grew very fast and over earned during covid as it was supplying some cameras used covid testing inside of PCR instruments. This caused euphoria, and later despair as sales of came back down.

After that, management changed for a new CEO. It was feared that the business model was dead, but SDI resumed with value enhancing acquisitions, such as PRP Optoelectronics

Revenues and profits are trending positive again.

It is highly surprising that in this AI driven market, 10 times profit companies like SDI growing 5, 10 or even 15% interest no one.

The next ones are similar.

Subscribe now

Read more

Intrum is a debt collection company.

The company had recently stabilised the stock price and earnings, but decided to announce last month a large capital increase that sent the stock falling 50%.

Do I accept being diluted by a management that plans to keep destroying shareholder value and forget about this investment, or I refuse and join the capital increase, despite bad management alignment?

I will review the recent operations and the capital increase.

Subscribe now

Read more

The portfolio is up 6.5% this year, despite provisioning capital gains.

I did not publish recent company related earnings and news, I was busy analysing the whole emerging market tech landscape and making these moves.

This is now done, so catching up with earnings and updates will be the next step and one update is coming soon..

In terms of sales, I got a take over offer at delivery hero with forced a brutal re-rating and it made me execute a sell at 37.65, with my average cost from last year at 25, because it now it is more of a buyout speculation.

After my article on emerging market tech landscape I found some opportunities to reallocate capital in some names, which I will detail below.

I also decided to prune the portfolio a bit with Egyptian company IDHC with 72% gains ex dividends in a short time. It’s a good business but its probably the worst macro situation in the world.

I bought with IDHC proceeds a serial acquirer at 1 times earnings. You read it right. No, it’s not a net cash situation, it is that cheap. I have never seen this. The market thinks that there is a risk of fraud, or bad capital allocation, and it is very misunderstood. So far, there are no real signs of fraud, but it could happen. I decided that it was worth 1% of the portfolio. If it goes to zero, so be it. It could reasonably be a 10x if it’s not taken private like many small caps.

I will also reveal what I thought of Coupang long term potential with a specific big picture approach.

I will also detail my thoughts on Brazilian Fintech. While not having done a deep dive yet, I provide a balanced review of all the metrics discussed earlier (price to sale, active users, TAM, cyclicality).

Subscribe now

Read more

I did a summary of the emerging market online companies, mostly in E-commerce and Fintech.

This did not include Tencent or Baidu in China as I mostly focused on e-commerce and fintech. I included MakeMyTrip in India because we don’t have any other Indian Tech ADR.

In this list I wanted to look at global commerce companies with another lens than the classic price to earnings ratio, which I used in the past to find good bets in the sector. (I point you to my analysis and portfolio reviews).

This list was tailored by me, pre-populated by claude AI, corrected manually for some datapoints and cleaned, before I asked claude AI to generate simple charts instead of doing them myself in sheets (same result but better looks).

I believe that we have to work with AI now and not pretend we do not use it and get along with the times.

If you liked this article please share it. Please note that not all data in the charts will be exact to the last decimal, but good enough to see the big picture. I am not S&P Global. Thank you, and let’s go.

Also the article was prepared before delivery hero is entered into a bidding war and is rising another 10% on Monday.

1-overview of the group of companies and Market cap ranking.

We can clearly see that Jumia in Africa is the smallest while Chinese giants Alibaba and PDD dominate the market cap ranking.

Two Brazilian Fintech, PicPay and Inter&Co, are among the smallest market capitalisations in the ranking. This could highlight an opportunity.

Another point I notice is that the vast majority of these companies are under the 100 billion USD market capitalisation figure, which is quite stunning giving the size of the opportunity, and many are under 50 billions.

2-Companies per revenue

In terms of revenue we can see again the Chinese e-commerce giants being heads and shoulders above the rest.

JD.com is first, but revenue is not the best way to compare JD, Alibaba and PDD. JD operates on a first party business model and sells the merchandise itself, while the others let third party merchants and only take a fee on the sale. GMV would be the best comparative between the three.

We have a group of companies under 10 billion in revenue, and these are companies early in the development cycle.

3-Companies per active users:

(Ecommerce branch for Sea limited).

Again, China is number one. This is truly another scale.

Sea limited (E-commerce branch) comes as second behind China in South East Asia.

Latin American countries are behind, and last are Eastern Europe and Africa.

4-Total Addressable market

Next is the population total addressable and reached:

Note that here for PDD and Alibaba, I put the almost the whole world at 6 Billion people, due to them being offensive with Temu, Aliexpress, and for Alibaba local acquisitions in Turkyie and South East Asia.

We see the scale of China. What shows up is that JD and Sea limited are much closer to their total adressable markets than the others. It’s difficult to see Coupang, but it’s well penetrated too.

This leads us to our next chart: the percentage penetrated: note that kids are not ordering on e-commerce platforms with their own accounts, so 70-80% of the population is probably the ceiling.

Sea limited is at a high percent, but I did not include Brazil here in the definition of TAM. Nevertheless it is very well penetrated in terms of TAM.

Mercado Libre is very under penetrated, which kind of surprised me as it has a large user base and market capitalisation, but Latin America has over 700 million people.

Delivery Hero is small in terms of penetration.

Makemytrip too, and the winner is Jumia, which has penetrated a tiny tiny part of the African population.

Pure fintech like Nu, Picspay and Inter&Co may never reach 80% of the population as people don’t use fintech as much as they use dominant e-commerce websites. They stuck to one or a few online banks.

5-Price to TAM

When looking at the price to TAM or the total addressable market, it tells us how much we pay for the total addressable market. This is more of a ratio to compare companies than a number that can be exploited.

Coupang is the most expensive because Korea and Taiwan are quite small countries, and the value of the market cap is quite high related to this population of 75 millions.

We find that Mercado Libre or Kaspi and Allegro are not cheap on these metrics. The cheapest are the un penetrated African and Indian consumer businesses Jumia and MakeMyTrip.

The brazilian fintech are also cheap here, PicPay and Inter&Co. But pure fintech is not e-commerce and we cannot expect all the companies to reach a 80% penetration. In e-commerce it is more common to have several memberships versus in fintech, but it could happen.

6-Price per active user

The market cap per active user is showing the cheapest stocks per active users.

We can see that Picpay is the cheapest on price per active user, and other Brazil Fintech Inter&Co is also low. This is reflected in the lower monetization of users when a product or service is of lower usage and average revenue per user is lower.

On the other hand, Mercado Libre (price has fallen a bit since then) and Coupang are the most expensive per user. This means that the market is willing to pay a higher price for users who are very active and bringing a lot of revenue.

This leads to our next chart:

7-Revenue per user

Coupang has a lot of revenue per user due to its dominance in South Korea and first party commerce sales too.

Mercado Libre and Kaspi are doing very well here.

Nu is impressive for a fintech.

We have then a group of companies, Jumia, Grab, Inter, PDD, with different business model and low revenue per user.

Makemytrip, at 14$ seems very low, and it’s either an error, or something due to some customers just using it for bus and train tickets, or cheap hotels in a country with a very low GDP per capita.

8-Valuation for price to sales

First party companies like JD, Coupang and Delivery hero are the cheapest.

On the other hand, less proven business models like Grab and Jumia and MakeMyTrip carry a growth potential premium.

Claude reviewed my internal comments and left the following ones on valuation

9-Combining price to sales and low penetration

Here, ideally you want the company at the bottom left corner: low price to sales and low population penetration. But it’s not easy as price to sales could mean a first party retail model.

Delivery hero is a food delivery company and now quick commerce also business that may not be able to penetrate the whole population. Delivery hero is subject to take over rumours from Uber and others at the moment. It was even cheaper when I built my position (see my article from last summer) and last follow up with better understanding of the business.

On the other side, Sea limited is very penetrated in South East Asia, but has buyers in Brazil that are impacting the penetration on the chart.

10-Conclusions and next steps for research.

What companies that I don’t own interest me the most?

What do I want to see in each company ranked attractively to make it a compelling investment?

Plus, I also charted the companies by a subjective “Market dominance ratio” from 1 to 5, versus price to sales. This next part is for premium subscribers only, but I hope that the above free research can already give some hints and tracks to explore for every reader.

This is in the premium section of the article. You will also be able to access analysis of some companies in the list such as Jumia and Allegro.

Subscribe now

Read more