Historically when P/E of S&P 500 was 30 or higher it returned 0% annualized over next 10 years. This is also the second time ever that Shiller P/E went above 40.

The main question is are there still any value investments and how can we protect ourself for the risks, or said in the other words, what to do with our money..

This is always a relative decision and it depends on how much risk we want to take, but if we want to stay in this game for long time and compound over the long term, we need to think of optimal investing scenarios where we are investing with margin of safety.

So there is always simple example where we have to think if our investing is going to bring enough of returns for the risk we are taking compared to US bonds.

Let’s say short term T-bills 1-3 months will get us around 4% without almost any risk, and then if we compare that with which is projected to return around 4% per year with significantly higher risk, T-bills look like a no-brainer.

The same thing is with , at current valuation I don’t even see if it can give us any returns, but I can see huge risks.

In his book “Margin of Safety”, Seth Klarman emphasizes that investors should target risk first, rather than desired return rate. He doesn’t suggest holding money in T-bills, but encourages finding stocks that will deliver us returns worth the risk we are taking.

And without further ado, lets get to the stocks I think currently give us enough returns to justify the risk!

Read more

Reasons why a company could be undervalued are simple:

- Company is going under the market’s radar

- Its a boring company, nobody cares

- there is some kind of risk around the company or the sector

- bad news causing overreaction from the market

So for Mag 7 companies we can exclude that the company is going under the radar, that they are boring, and most of the time, unless there is big risk involved we wont see huge overreactions from minor events or for slightly missing earnings.

We are basically left with “there is some risk around them”, and I will try to breakdown what are currently the biggest risk of each Mag 7 company, and how is market perceiving them.. 1. Nvidia Recently we saw

1.

Making three moves which caught my eye, and made me think.. First move was Nvidia decided to sell $25 billion in high-grade corporate bonds in June 2026, marking its first debt issuance since 2021.

This was surprising to me as $25B doesnt seem necessary for a company that just committed spending $250B on a deal with OpenAI, and more with other companies involved in AI.

Second move came recently when Nvidia decided to cut the deal with OpenAI from $250B to $120B which is kinda leaving OpenAI in a bad spot as they committed over 1.4T in spending towards other companies, and this could be first sign of links breaking, and if one link breaks it could make the whole thing collapse.

Third move was just few days ago when Nvidia announced they are going to be raising prices of their products by 15%. Everything is leaning in the direction that something is not right, and we might be in for a shock in the upcoming earnings on August, 26th after market closes..

2.

I think the Apple has completely the opposite risks to ones Nvidia has. Apple has significantly lower capex, and the biggest risk is missing the AI integration and huge lag compared to other Mag 7 competitors.. The other risk might be lack of product innovation, and China exposure..

3.

Currently the worst performing Mag 7 over the last 5 years with only 53% returns while S&P 500 returned 87% not including dividends.. Meta is currently facing a lawsuit from 30 US states, for endangering safety of children and highly addictive product.

The lawsuit penalties could reach up to $1.4T which is almost the whole Meta market cap.. And ofc there is risk of all the capex could not give returns as they could lose a AI race to the competitors..

4.

Amazon has been increasing its capex with the data center infrastructure making their free cash flow negative. I already mentioned risks of data centers and AI infrastructure in my Oracle article, but basically its a big bet that the AI training data centers are still going to be used enough even after the current contracts expire and companies will be able to get new deals.

This is in danger cause AI is turning more to inference and this means different type of data centers are growing in demand, while the cost of old ones are going up. Another issue is that the price of tokens is going down rapidly and its a question if earnings will be able to catch up with cost..

Subscribe now

5.

Except the risk of high spending like all hyperscalers, Microsoft is also facing risk of its software tools being disrupted by AI, and how all of these factors will affect the company’s ROI.

ROI trended downward, sliding to roughly 23.51% in 2023, 22.78% in 2024, and dipping further as heavy investments in artificial intelligence and data infrastructure ramped up.

6.

From the company with high cash flow Google became a company without cash, and with increasingly growing capex..

There was some risks about the search business as it may get impacted by AI, and it faces similar risky as Microsoft, but for now its holding up better. To be real, this is not the same company it was few years ago, which was money producing machine with bulletproof business..

7.

To add some drama to this post, I dont even consider Tesla a Mag 7 company, and I think Visa, MA or even Walmart would be better fit, but thats just me.. Now to the risks.. Execution delays with everything.

We should have already been on Mars, robo taxies should already be in half of the world, Tesla cars should already be the number 1 EV cars in the world, but non of these happened.. Promises and marketing campaigns keep ramping up but there is no significant improvement in fundamentals of the business.

Revenue hasnt moved since 2023, EPS is down, shareholders are getting diluted while CEO is collecting compensation.. Current profit margin of Tesla is 3.67% which is in range of Walmart, but the revenue of Walmart is 7x higher, and it is growing faster..

We dont even have to talk about valuation, right? New marketing scheme coming soon when Tesla joins forces with

I also made an article that companies like , , will do very well and explained why..

is up 46.5%, is up 20 %, is up 15% since I posted that article. This was an obvious prediction since we had biggest sports event in the history of the planet coming in few months.. (FIFA World Cup that was extended by 34.48% and number of games was increased by 62% from the previous one).

Share Deep Value Investing

That meant more money will be spent and overall US economy will prosper, and mostly the companies I picked..

I checked the hotels in areas where the games of most popular teams will be, and everything was booked, so the alternatives were? You guessed it AirBNB and Booking..

On top of World Cup we also had events like World Series of Poker 2026 at the same period to keep more tourists in US and some of the people surely extended their stay in US.. (this is a minor one, but still something).

Even tho results of income are still not in the earnings reports and it will be visible in Q3 and Q4, the money has already moved into these stocks and we were one step ahead..

That being said, we can go into my new insight that is coming to disrupt the market, and give us another like golden opportunity..

Read more

This one came from Charlie Munger who said: “Investing is like fishing, you have to fish where the fish are. The two rules of fishing are: Rule number one, fish where the fish are. Rule number two, never forget rule number one.”

In my head, I was like “yeah, makes sense” but I just continued listening without really realizing that was exactly what I was looking for.

There is no point for me to be looking at overvalued US companies when I can just look elsewhere and find significantly cheaper companies, and if I do a good research they shouldn’t be lower in quality..

I know there is skepticism about other markets within the finance industry, but if we realize every investment carries risk, its just game of numbers..

If we have two exact same companies, and company A is at 30 P/E but it is in US, and company B is in China at 10 P/E, which one has more margin of safety, and which one carries higher risk?

is the living proof that risk is acceptable as long as the quality is top notch..

Read more

One of them was Li Lu increasing his PDD position by 133%, which I understand as PDD looks like one of the most undervalued major cap businesses in the world.

But what was the most shocking was Warren Buffett (or Greg Abel) buying Google at $357/share. I think its Greg Abel even tho Buffett went to give an interview and said he bought Google.

Before you start telling me how Google is cheap or best business or whatever, let me first lay down some facts..

Positives about Google:

- an AI leader

- most profitable company in the world

- best company in the world

- extremely well managed

- huge moat

- diversified with amazing businesses

Negatives about Google:

- questionable SpaceX position

- adding 40B into Anthropic investment

- growing capex (91B in 2025, 190B in 2026)

- valuation (screening at 18, but closer to 40 if we discount for Anthropic and SpaceX earnings)

And I understand all of this is fine, and Google is in good position to succed, but if we compare it to previous big “bets” Buffett and Munger did, for example Apple..

Five of the biggest buys of Apple came at these valuations:

- Q2 2016 - 5.4 million shares - 10-12 P/E, (S&P 500 P/E 25)

- Q4 2026 - 42.1 million shares - 11-14 P/E, (S&P 500 P/E 22.9)

- Q1 2017 - 70 million shares - 13.5-16 P/E, (S&P 500 P/E 23.5)

- Q3 2017 - 3.9 million shares - 16-18 P/E, (S&P 500 P/E 24.1)

- Q1 2018 - 75 million shares - 14-17.5 P/E, (S&P 500 P/E 22.18)

So as we can see Berkshire was buying Apple BELOW the market value, which is not the case with which is trading around market valuation.

The S&P 500’s normalized P/E, known as the Shiller CAPE (Cyclically Adjusted P/E) ratio is approximately 41 to 42. This valuation is the second-highest in U.S. market history, trailing only the extreme peak of the dot-com bubble (which hit 44.2 in late 1999).

I also understand that BRK had to deploy cash somewhere, but weren’t: at $200-220, at $460, at $500, at $350, 350-400 better options?

You can hold these 5 companies for next 10 years without worrying they will disappear, and I would bet they outperform Google in next 3 years which gives you enough time to see the situation and rotate if needed..

Conclusion:

This looks nothing like old Berkshire and it feels to me like Greg Abel is feeling the pressure as the new CEO to deploy that huge pile of cash (400B --> 365B).

His other picks in Q2 were , , and ..

So my question is what is the idea of Berkshire now? Is it value investing, quality investing or FOMO?

I think Charlie Munger is turning in his grave.. RIP GOAT..

Who Did the Best Job in Q2?

But one of the investors played this Q2 how it is supposed to be played, and bought the right stocks.

Read more

I don’t see many things in the market that look like amazing buys, except the ones that I already own and that’s why I decided to double down on some of them..

If you follow my X, you might be surprised that I own some of the stocks which I criticized, but just because I criticize a certain stock at specific price/valuation, it doesn’t mean I think its a bad company..

For us value investors its all about getting the best quality at best VALUATION, and this is what matters the most.

Yes, if I find and average quality company at ridiculously low price, I might pick it up, but my main focus is on high quality not at decent price, but at amazing price. This will mean that a lot of the time I have nothing to buy and just hold what I already own, and this will also besides fundamentals and price dictate if I sell something..

If there is no better replacement, Im not selling even if the price caught up fundamentals (intrinsic value)

So without a further delay, let’s get into the breakdown of my portfolio where I will try to answer all the questions why I sold or why I hold something..

Read more

“Insiders might sell their shares for any number of reasons, but they buy them for only one: they think the price will rise.”

When I see big insider buys, it’s a big sign which tells me to turn my focus to these stocks and redo my research to see if I missed something or, if I already own these stocks, to check the valuations and evaluate if it’s time to buy more.

Here are 4 stocks: (I currently don’t have a position in any of these stocks)

(The valuation numbers are from Finviz and might deviate slightly from other screeners)”

Subscribe now

Fiserv Inc - FISV - (fwd P/E - 5.68, PEG - 0.92)

Chief Financial Officer -Todd Paul M: $1,560,952

- Jun 17 ‘26: $499,982

- Dec 01 ‘25: $1,060,970

Chief Admin & Legal Off - Rosman Adam L: $999,901

- Jun 16 ‘26: $500,700

- Dec 02 ‘25: $499,201

4 other Directors: $722,719

- Mamilli Wafaa: $149,746

- Yarkoni Charlotte: $100,118

- DiSimone Harry: $101,080

- Nixon Gordon M.: $371,775

Elevance Health - ELV - (fwd P/E - 12.62, P/S - 0.4)

President & CEO - Boudreaux Gail: $1,002,230

- Jul 17 ‘26: $1,002,230

Director - PERU RAMIRO G: $366,050

- Jul 17 ‘26: $366,050

Lululemon Athletica - LULU - (fwd P/E 10.26, P/FCF 10.8)

Director - Bergh Charles V: $1,500,362

- Jun 15 ‘26: $500,384

- Mar 20 ‘26: $999,978

Pres, CCO & Interim Co-CEO - Maestrini Andre: $494,591

- Apr 01 ‘26: $494,591

MercadoLibre - MELI (PEG - 1.26, P/FCF - 10.7)

SVP - Chief Accounting Officer - Marcelo Melamud: $300,079

- June 11, ‘26: $200,000

- Feb 27, ‘26: $100,079

Director - Alejandro Nicolas Aguzin: $993,556

- May 22, ‘26: $993,556

Conclusion

All of these companies had some issues in the business, but they still look like they are down more then they should be and I would say there has been overreaction from the market..

When I have this kind of feeling, and on top of that I see insiders are buying big amounts of shares, it gives me more conviction..

NFA, DYOR!

This could also be additional risk and sign of bubble, all of the companies are just competing even if it doesn’t make perfect sense or if its considered great capital allocation..

I expect once OpenAI can’t pay what they owe to the companies, they will offer them stake in the company.

OpenAI, the Unprofitable Company

OpenAI currently generates $25B/year, and it is not a profitable company..

Yesterday I also saw that there is only 22% the will IPO this year, so most likely 2027, this is where they could get part of the money.

They are expected to IPO at $1T valuation and companies usually sell 15% to 25% of the shares which would give them 150-250B..

This is type of the investment I would never consider regardless of how fast it is growing..

What Are OpenAI’s Commitments and to Who?

($350B)

OpenAI has signed a deal with Broadcom to design OpenAI accelerators and new chip infrastructure.. The deal stretches over 4 years, but the later hardware deployment estimates are over 10 years..

($300B)

This 5-year cloud computing deal targets to heavily expand data center capacity and its tied to “Stargate” data center initiative..

You can find about more about Oracle and if it is a good investment in the article below:

($250B)

Ongoing compute commitments for Azure cloud infrastructure and inference costs.. The deal is signed over 6 years..

($100B)

Built on an initial $38B compute agreement OpenAI extended the deal with AWS for additional $100B that is supposed to be paid out over 8 years..

($100B could go up to $600B)

Mass purchasing and leasing of the top GPUs, massively intertwined with a big strategic investing round.. There are talks between two companies that would dramatically increase the size of the deal up to $600B total.. Nvidia would finance OpenAI’s 350B purchase of AI Nvidia chips..

($90B)

Multi year deals to purchase AMD instinct accelerators over 4-5 years.. Gradual scaling toward the 6-gigawatt ceiling across future generations of AMD hardware, projected through approximately 2030. AMD also has option to get 10% of OpenAI stake..

($22B)

Specialized data center space and GPU cloud rental to assist model training. The deal is signed for the next 5 years..

Do you think OpenAI can find a way to pay out all of this commitments or the link is going to break at some point?

It all started on September 10, 2025, when Oracle signed a monumental cloud computing contract worth $300 billion with OpenAI, causing Oracle’s stock to skyrocket by roughly 36% to 40% in a single day!

Even before signing that deal, Oracle stock price had big jumps after announcing deals with and OpenAI which also moved stock around +100% over the year..

All of these jumps were very unusual to me and served as a clear sign that something unnatural was happening, causing my sixth sense to activate and scream “WARNING.”

I didn’t even have to wait long to confirm my concerns. The stock started dropping aggressively from its all-time high, eventually hitting a 52-week low and returning to price levels not seen since early 2024.

So basically we were back where we were before any of the exciting deals were announced. Do I even have to mention that I was calling it out on my X and suggesting to people that they should take profits and reduce the risk, same like Im doing now with , , , .

As you might expect, the reactions were exactly the same: people told me I was just “late to the party,” or that I was crying because I didn’t own shares.

But the truth is, this never even crosses my mind. I do not evaluate whether an investment is good or bad based on how much the stock price went up or down; I evaluate it based on the data available at the moment I chose to buy or pass on it. Clearly, in the previously mentioned cases, I did not think they were good investments, and I still hold that exact same opinion today.

For example should have been bought when Mohnish Pabrai was calling it one of the best tech play, and when the stock was below $100. The same Mohnish who later got criticized for his fund performance and called “scammer” or a “clown investor”.

Back to the Oracle..

Oracle is a major enterprise software, database, and AI infrastructure company undergoing a high-stakes transformation driven by AI demand.

Currently, cloud infrastructure is the largest and fastest-growing segment of the business. It generates 27% of total revenue and is growing at a rate of 77% year-over-year. This places it ahead of the cloud applications segment, which accounts for roughly 24% of total revenue and is growing at a significantly slower rate of 11% year-over-year. Together, these two segments make up 51% of the company’s total revenue.

This shows us a clear direction in which the company is moving and highlights the previously mentioned “high-stakes” transformation.

Software used to be the main source of income, accounting for 36% of total revenue, but it is now declining at a rate of 1% year-over-year.

The other two business segments are services and hardware. Services contribute 8.5% of total revenue and are growing at 10% year-over-year, while hardware contributes 4.6% and is growing at 5% year-over-year.

Read more

After I wrote last two articles I had just one more good idea about where I could potentially put my money before I would just look for shorts and/or consider putting remaining cash in T-bills until something great pops up, and that’s exactly what happened..

I was avoiding this stock, and even some friends who’s money Im managing were suggesting that the stock went down a bit and it might be interesting..

Ofc I saw it and knew that it could continue going down and saw that its earnings are coming soon, so it was a perfect moment to wait it out and see if it becomes great deal or I stay away..

After earnings the stock dropped significantly, and it looked attractive, but I still wasn’t fully on board, as I am always hoping for deepest possible value.

As the Buffett and Munger were suggesting, at investing we don’t have to swing at every pitch, but we have time to wait for the perfect ones, and this was starting to look like its time to swing..

“The trick in investing is just to sit there and watch pitch after pitch go by and wait for the one right in your sweet spot. And if people are yelling, 'Swing, you bum!,' ignore them.”

- Warren Buffett

I was waiting to see if it keeps dropping when the market opens, and it did a bit.. While I was waiting I was researching everything I could about the stock and things were looking good and it felt like an overreaction.

So I decided to buy a big position in the stock, but still leave some room for it to drop more so I can do what I did with , and kinda go all-in..

The stock was already screening as one of the best businesses in the world after the drop, and if it potentially dropped more, I would be so happy to put bunch of money into it and still sleep well.

Read more

My plan was to write more about the global market and start thinking more about what are the best short opportunities, and hope market makes some mistakes during the upcoming earnings season. I was prepared to spend more time researching and waiting than buying or selling stocks..

But as things usually go in the stock market, we are presented with unexpected opportunities, and the best ones are ones that come from the sectors or directly the stocks you previously deeply researched.

This is exactly what happened this time, I wasn’t expecting for this stock to go down, but I expected it to continue going up, as its peers are, and as the whole sector is..

I feel this stock is a great buy relative to the current market and what else is available, but I don’t think it has insane margin of safety, but I do think its a very favorable bet with amazing long term upside.

Read more

I still hear people asking me, “Why do you think the market is expensive?”, often following up with, “Micron is trading at just a 5.6 forward P/E.”

When I hear that, I just take a deep breath, and if I’m in the mood to pull up a chart that explains it, I post it. Most of the time, this turns into a discussion about PEG, forward P/E, and expected returns.

Here are the charts comparing the Shiller P/E to the Dot Com Bubble, alongside the current S&P 500 P/S ratio hitting historical high..

Now that we got this out of the way, I want to highlight how much more valuable it is to find a true value pick like the one I just discovered.

The valuation is amazing, with strong fundamentals that are undergoing rapid improvements. This is already reflected in higher operating margins and solid revenue growth. With plenty of cash on hand, no debt, and a great dividend, this is a true value GEM!

I bought the stock yesterday and would gladly increase my position if the price drops. Since I don’t have an infinite number of amazing ideas in the current market (I don’t think anyone does), this one really stands out..

Read more

When I was in the early stages of my investing career, fresh off my big crypto score where I 12x’d my starting capital in a single year, I decided to transition into stocks and value investing. One of my first high-conviction investments was .

Not yet aware of the China-specific risks, I saw the stock trading -50% from its all-time high as a buying opportunity.

Very quickly, I learned that things don’t work the way they do in your head. I lost 40% on a single stock that represented 20% of my portfolio. I did use it for tax-loss harvesting, selling Alibaba at a loss to offset gains elsewhere.

So from my experience, investing in Alibaba has been anything but a boring ride. I stayed away from the stock for a long time because it never felt like an amazing opportunity, especially considering the added geopolitical risk, until now. While we’re in the most expensive market environment in history and most international markets are hitting all-time highs, China remains reasonably priced, and its AI giant, Alibaba, stands out.I’ve been allocating capital to for a while now.

After poring over the latest filings, earnings trends, and industry shifts as of mid-2026, I wanted to share my updated thoughts with you. This isn’t some quick take. I’ve built out a full framework here, including fresh numbers on the stock, ongoing capital returns, the AI/cloud story in Asia, and a complete discounted cash flow model I put together myself.

The stock is up from the low $90s it touched earlier this year, with a 52-week range roughly between $92 and $193. The market cap sits near $262 billion with about 2.4 billion shares outstanding (down meaningfully from prior years thanks to consistent share repurchases).

What drew me back is a combination of things, interesting valuation, buybacks, Alibaba’s market position, strong core business, and more..

So let’s make a deep breakdown and see if is a banger investment ATM..

Read more

I’m not sure what the exact price of TSM was at that point, and I’m too lazy to check the dates, but I have been buying the stock since it was at $66. I think I started posting about it when the price was in the $100 to $120 range.

Back then, people kept warning me about ‘China risk’ and claiming that TSLA, CELH, or SOFI were better picks. Looking back, I still can’t believe what I was reading.

As of July 8, 2026, Taiwan Semiconductor Manufacturing Company (TSM) stock is up 606.86% since its lowest point of 2022.

My deep understanding of TSM, and my high conviction regarding Micron and Intel, stems from studying the semiconductor industry for the last six years. Furthermore, I prioritize analyzing valuations and staying away from risky stocks, even if it means missing out on some good returns.

Michael Burry Opened Micron Short

Recently, when Burry closed his short position, I realized my investing style is very similar to his. For instance, I predicted on X that Palantir would drop roughly 50% from its highs, though I never pulled the trigger on the shorts because I don't believe shorting is the optimal strategy.

By combining logic with my observations of Michael Burry's moves, I shorted Micron, Intel, Tesla, and SpaceX. This strategy secured me a profit of roughly $18k in less than three weeks

Here is the key to shorting..

Read more

The annual inflation rate in the US was around 4.2%, which means their money is worth 4.2% less after one year. If we add a compounding effect to it, things get even worse.

For example:

If you had $10,000 sitting in your savings account for 5 years with a 4.2% inflation rate, you would lose $1,859.31 in purchasing power, leaving you with $8,140.69 in today’s money.

That would be a loss of 18.59% of your purchasing power.

Year-by-Year Purchasing Power Breakdown

Starting Balance: $10,000

Year 1: Value drops to $9,596.93 (You lose $403.07 in value)

Year 2: Value drops to $9,210.10 (You lose another $386.83 in value)

Year 3: Value drops to $8,838.87 (You lose another $371.23 in value)

Year 4: Value drops to $8,482.60 (You lose another $356.27 in value)

Year 5: Value drops to $8,140.69 (You lose another $341.91 in value)

This is the main reason why holding cash is the opposite of safe and secure. The safest thing is actually putting your money to work, and the main way I do it is through stocks. This specific stock is what they call a “mega compounder,” while also being as safe as a Swiss vault..

Read more

These initiatives are likely losing $8 billion or more annually in total. Once you back those losses out, the underlying operating profit of the main businesses probably sits closer to $95 billion. After a realistic 20% cash tax rate, that translates to roughly $76 billion in after-tax core earnings.

Even after netting out cash and a major AI-related investment stake, the enterprise value comes to about $2.08 trillion. In simple terms, you’re getting two world-class operations, one in retail and one in cloud computing, both positioned in fast-growing markets and riding strong AI tailwinds, at a mid-20s multiple on normalized earnings. That setup has the potential to deliver solid long-term results.

The Strengthening Retail Advantage

Amazon’s online retail business has built one of the widest and fastest-widening competitive edges in the market. While many traditional advantages are coming under pressure from new technology, Amazon’s position keeps getting stronger as it scales. The company already handles more than 40% of all U.S. online shopping, and the gap to the next player is huge.

Greater sales volume lets Amazon stock products in more warehouses located nearer to customers. That proximity improves product selection, speeds up delivery, reduces miles traveled per package, and cuts down on handling. All of these improvements drive costs lower. Lower costs allow for better prices and quicker shipping, which brings in even more shoppers and orders.

The cycle feeds itself. Amazon has also moved early and aggressively into warehouse automation and robotics, which now support the majority of its worldwide shipments.Even with its dominant U.S. position, e-commerce still makes up only about 16-17% of total U.S. retail spending, and far less overseas. As more shopping shifts online and delivery gets faster, the opportunity keeps expanding. Despite heavy spending on logistics, the retail side generates strong returns on the capital it invests. Suppliers help a lot here by essentially financing part of the growth through favorable payment terms.

Between 2018 and 2025, North American retail revenue more than tripled, while the company invested heavily in facilities and equipment. After accounting for supplier support through working capital, the effective capital deployed delivered incremental returns in the mid-to-high 20% range. A big hidden gem inside retail is the advertising business, which now brings in nearly $69 billion a year and is growing around 22%.

Amazon has a unique edge because it sees not just what people buy, but what they search for, click on, and decide against. That depth of customer insight is extremely powerful for targeted ads.

The company is also starting to leverage this data advantage outside its own sites by helping advertisers reach audiences across other major platforms.

Subscribe now

AWS: Cloud Leadership with AI Upside

In 2025, AWS generated $129 billion in revenue, up 20% from the prior year, and delivered $46 billion in operating income. Growth picked up to 24% in the final quarter, the strongest in more than three years. Its own custom chips for AI and general computing are already contributing over $10 billion annually and expanding at triple-digit rates.

The outlook is straightforward: enterprises keep moving to the cloud, and AI is creating an entirely new layer of demand for computing power and infrastructure. There’s valid worry that today’s massive buildout of data centers could eventually create oversupply.

Still, the current stock price doesn’t seem to bake in overly optimistic assumptions for AWS, especially when paired with the retail performance. There’s also a realistic path where AI needs stay so strong that AWS can keep growing in the high teens to low 20s percent range with healthy margins for years to come.

Historically, AWS has shown solid efficiency in turning capital into profit. From 2018 through 2025, operating income increased dramatically while invested capital grew substantially. Adjusting for projects still under construction, the returns on new capital look to have been in the mid-20% area.

Bottom Line

I’m not running a complex spreadsheet here because the story doesn’t require one. The market’s main worry is the jump in capital spending toward $200 billion a year. To me, the bigger opportunity lies in how effectively Amazon can put that money to work at high returns.

That potential upside feels much more likely than a sharp slowdown or major competitive threat in cloud if AI enthusiasm cools off. Overall, paying around28 times core earnings for a company with this combination of quality, proven economics, and multi-decade runway feels like a favorable setup for patient investors.

Key Drivers Going Forward

Continued improvement in retail profitability (helped by growing ad revenue) and steady expansion at AWS. In the end, time and execution should do most of the work..

The narrative is seductive: artificial intelligence is reshaping the world, chips are the new oil, and anyone who doubts the sector is missing the opportunity of a lifetime. But in my opinion, this collective enthusiasm masks a harsher truth.

Semiconductor valuations have reached levels so detached from fundamentals that the stocks are, for most investors, uninvestable right now. The very “opportunities” everyone chases are precisely what make them dangerous.

I still don’t understand how people think that is normal for a large/mega cap stock to go up 570% in 1 year..

Subscribe now

The Seductive Power of the AI Narrative

The hype is understandable. AI data-center spending has exploded, driving explosive demand for high-end GPUs, high-bandwidth memory, and advanced foundry capacity.

dominates training and inference workloads. is the unchallenged king of leading-edge manufacturing. , , and others ride the coattails of hyperscaler capex. Revenue growth has been spectacular, margins are fat, and forward guidance from companies like TSM, raised to over 30% for 2026, only fuels the fire.

Analysts and pundits speak of a “supercycle,” a structural break from the industry’s traditional boom-bust rhythm. In this story, semiconductors are no longer cyclical; they are perpetual growth machines. But markets price expectations, not current reality. And the expectations embedded in today’s prices are delusional.

Valuations That Demand Perfection

Look at the numbers. As of early May 2026, the SOX index trades at roughly 60 times earnings, the richest valuation since the dot-com peak in 2000, when the index last staged a comparable short-term surge.

NVIDIA’s trailing P/E sits around 40.5x, with its forward P/E near 24x. On a price-to-book basis, the sector averages 10x. These are not modest premiums; they assume flawless execution, endless AI capex, and no meaningful competition or cyclical downturn for years, and what I see here is RISK, RISK, and more RISK!

Historical Warnings Ignored

Compare that to history. During the dot-com mania, the semiconductor sector’s multiples expanded dramatically on the promise of the internet. When reality intruded, overcapacity, slowing PC demand, and a recession, the SOX cratered.

Today’s setup echoes that era: massive capital expenditure chasing AI infrastructure, hyperscalers building out data centers at breakneck speed, and valuations that bake in multi-year growth rates few industries have ever sustained. Even optimistic forecasts from firms like IDC project the global semiconductor market hitting $1.29 trillion in 2026, heavily driven by AI.

But unit volumes outside AI remain soft, inventories in non-AI segments are still elevated, and average selling prices for many chips are propped up by scarcity rather than structural pricing power.

The TSM Premium Under Scrutiny

TSM, often viewed as the “safer” play, is not immune. Its forward P/E hovers in the low-20s to mid-20s depending on estimates, with a price-to-sales multiple around 14x in some analyses. That only looks reasonable if you believe 20-30% revenue growth and 50%+ net margins are sustainable indefinitely. The company is pouring $52-56 billion into capex in 2026 to chase 2nm and beyond, yet any slowdown in AI orders would leave expensive fabs underutilized.

Geopolitical risk compounds the issue: Taiwan’s centrality to the supply chain is a known vulnerability that markets conveniently ignore until headlines force a rethink.

Concentration Risks and Emerging Competition

The delusion runs deeper when you examine the growth assumptions. NVIDIA’s forward multiples imply continued dominance and explosive earnings expansion. Yet hyperscalers are already developing custom ASICs (Google’s TPUs, Amazon’s Trainium, Microsoft’s Maia) to reduce reliance on third-party GPUs.

AMD is gaining share in servers. Chinese firms, despite export restrictions, are accelerating domestic alternatives. Supply is catching up to demand faster than bulls admit. Deloitte notes that high-value AI chips now drive nearly half of semiconductor revenue while representing a tiny fraction of unit volume, a concentration risk that screams vulnerability to any shift in AI investment priorities.

Here We Go Again..

History offers sobering parallels. The dot-com bubble taught us that transformative technology does not justify infinite multiples. The internet changed everything, yet Cisco, Intel, and scores of chip firms saw 80%+ drawdowns. Japanese asset bubbles in the 1980s showed how overcrowding capital into “strategic” sectors leads to malinvestment and painful resets. Even the 2022 semiconductor correction, triggered by post-COVID inventory gluts, demonstrated the industry’s cyclical DNA.

AI may be real and durable, but the current pace of capex is creating the classic capitalist problem: overproduction. When data-center ROI fails to materialize at promised levels (and early signs suggest many projects are still proving payback), spending will moderate. Earnings estimates will come down. Valuations will compress.

Chasing Trends..

Investors chasing these names today are not buying a business; they are buying a narrative at peak enthusiasm. Any earnings miss, guidance cut, or macroeconomic hiccup, higher interest rates, weaker corporate budgets, or geopolitical flare-up, will be punished severely. The SOX’s recent 30% surge in just 13 days in April 2026 was not a sign of strength; it was a classic momentum blow-off, the kind that precedes painful reversals.

Discipline Over FOMO

This does not mean semiconductors lack long-term promise. AI, edge computing, automotive electrification, and 5G/6G will drive genuine demand for decades. The industry will grow. But growth and investability are different things. At current prices, the margin of safety is nonexistent. Expected returns are mediocre at best and catastrophic at worst.

Better opportunities exist elsewhere: in undervalued industrials, energy, or even cash while waiting for a 30-50% correction that history suggests is probable.The delusion persists because recency bias is powerful. Three years of extraordinary returns have convinced many that trees grow to the sky. FOMO drowns out skepticism. Yet the data is clear: semiconductor stocks are priced for perfection in an industry defined by imperfection. The “opportunities” everyone sees are illusions created by stretched valuations.

When the music stops, and it always does, the hangover will be severe. True investors do not fight the tape in the short term, but they also refuse to pay any price for growth. In semiconductors today, the price is simply too high. The smart move is not to join the crowd in its delusion, but to step back, preserve capital, and prepare to buy quality names when sanity returns. The future of chips is bright. Their present valuations are anything but.

After that period, positive news started to emerge, and the stock finally rose, putting me up 50%.

Read more

As Mark Twain said, history doesn’t repeat itself, but it often rhymes. That’s why I’ve based some of my assumptions on past events. Of course, in today’s much more globalized world, exact repeats are harder, but clear parallels with previous periods can still be drawn.

Read more

One thing that caught my eye, and the eyes of many others, is their betting market: “Will Jesus Christ return before 2027?”

At first glance, this seems like a crazy bet where people might feel it’s free money for the taking. However, how it actually works is quite interesting.Let’s first address the event where Jesus Christ returns.

If there were an event of such huge magnitude that the media interpreted it as the return of Jesus Christ, then people betting against it would lose money, and those betting on it would obviously win. But this is not a realistic outcome, and Polymarket might not even factor this scenario into their calculations. If they do, it’s only a minor consideration.

Subscribe now

Most people betting on the “No” outcome will almost certainly make money, and this is considered one of the safest possible bets on the platform.

For example, if you put $100,000 on Jesus not returning, bettors would win $103,950.10. It seems like free money, but there’s a catch: Polymarket charges fees and uses your money to increase liquidity.

Another thing they could be doing is putting your money into short-term U.S. bonds or T-bills, which currently yield more than 3.95% (around 4.35–4.37%).

So the whole thing works out extremely well for Polymarket. They benefit from trading fees, increased liquidity, potential profits from investing in T-bills, and, most importantly completely free marketing.

And it’s not just any marketing: this betting event has been seen and discussed by countless people, which has significantly boosted Polymarket’s visibility and popularity.

There’s nothing else to say except: pure genius!