Millennials Capital

Nebius didn’t suddenly appear on my radar because it reported strong results today.

I’ve been following the company for months as part of a much bigger investment thesis:

the AI infrastructure buildout.

When people think about AI, most immediately think of ChatGPT, applications and software.

I prefer to look at the entire value chain.

Chips.
Data centers.
Cloud infrastructure.
Energy.
Networking.

This is essentially the “AI Cake” framework I use as a guide to understand where the massive amounts of capital surrounding artificial intelligence are actually flowing.

And Nebius ($NBIS) sits right in the middle of that story.

In today’s results, the company reported:

• $582.3 million in revenue, +454% YoY
• $236.2 million in Adjusted EBITDA
• 4 major AI Cloud deals, with an average value of more than $1 billion each
• a 5 GW contracted power target
• more than $9 billion in customer prepayments for 2026

But for me, the most important part isn’t one strong quarter.

It’s that an investment thesis we’ve been following for quite some time is beginning to be validated by actual results.

At Millennials Capital, we don’t simply look at which stock went up today.

We try to understand:

where capital is flowing,
which industries are benefiting,
which companies are seeing real demand,
and which data points can either confirm or invalidate our original thesis.

NBIS is a great example of that process.

Not because we “picked a winning stock.”

But because the real work happens long before the result shows up on the chart.

Returns are the outcome.
The process is what I’m trying to build every single day.

This content is for informational and educational purposes only and does not constitute investment advice.

In other news, July inflation data came in exactly in line with expectations, with no negative surprises.

We are witnessing one of the biggest investment bets in history.

And that bet is AI.

Hyperscalers have committed roughly $2.6 trillion in future spending on data centers, power infrastructure, equipment, and services.

The most striking part?

A large portion of these obligations relates to agreements that have not even started yet and are disclosed mainly in the footnotes of financial statements.

Google stands out with around $896 billion in total commitments.

Oracle, on the other hand, has approximately $292 billion.

And this is, in my opinion, where the really important part begins.

The world’s largest technology companies are not simply spending a few billion dollars to experiment with AI.

They are committing capital and infrastructure for years.

Data centers.
Energy.
Chips.
Networking.
Cloud infrastructure.

The AI buildout is becoming one of the largest long-term investment cycles we have ever seen.

And as this infrastructure continues to expand, one question becomes increasingly important for investors:

Which companies will ultimately manage to turn these trillions of dollars in investment into real profits?

NBIS up nearly 10% and breaking above $210 ahead of earnings

Meta unveiled a new AI model, while Mark Zuckerberg called for fewer restrictions on open-source and open-weight models in the U.S., arguing that this would help American companies compete more effectively with China.

The most important takeaway is not the model itself, but the broader trend: competition between Meta, Google, OpenAI, and Chinese AI companies continues to intensify.

This means even more investment in computing power—and therefore greater demand for AI infrastructure.

Tomorrow brings the most important economic data release of the week: July’s U.S. inflation report.

Economists expect approximately:

• CPI: 3.4% year-over-year

• Core CPI: 2.5% year-over-year

The announcement has become even more important following last week’s weak employment data. At the moment, the market sees roughly a 50/50 chance of an interest rate hike in September.

• Lower CPI → relief for the markets.

• Higher CPI → increased odds of another rate hike and potentially greater volatility.

Taiwan Semiconductor (TSM) announced that its July sales increased by approximately 45% year over year, reaching $14.5 billion. The main driver was the enormous demand for chips used in AI.

TSMC is arguably the most important chip manufacturer in the world. Companies such as Nvidia and Apple design their own chips, but rely heavily on TSMC to manufacture them.

During the first seven months of 2026, the company generated approximately $89.1 billion in revenue, representing a 37% increase compared with the same period last year.

Most importantly, the company does not appear to expect a slowdown. Instead, it raised its 2026 investment plan to a record $60–64 billion. In other words, TSMC is investing enormous amounts in new factories, equipment and production capacity because it expects demand to remain strong.

For the full year of 2026, the company expects sales growth of slightly more than 40%, while the market currently estimates growth of approximately 47% for the current quarter.

Another interesting point is that TSMC’s shares in Taiwan have risen approximately 50% since the beginning of the year, compared with around 20% for Nvidia.

The key takeaway is that, so far, there are no signs that the AI infrastructure boom is coming to an end. On the contrary, the company that essentially “makes the picks and shovels” for a large part of the industry continues to see tremendous growth in orders.

Before the sell-off, too many positions had piled into the same popular stocks, mainly those linked to AI. Once the selling began, many investors rushed to reduce their exposure to risk. Retail investors became net sellers, AI stocks came under heavy pressure, and more than $60 billion flowed out of the market.

Simply put, this was a major market clean-up. A large part of the excess and excitement that had built up was flushed out.

“Smart money” appears to be making a strong return to technology stocks.

During the week ending July 30, hedge funds recorded their largest net purchases of U.S. technology stocks since December 2022.

It was also the third-strongest week of buying in at least the past five years.

The most important point is that the activity was driven mainly by new long positions. In other words, funds were not simply closing bearish bets. They were deploying fresh capital in anticipation of higher prices.

The strongest demand was seen in software, semiconductors, and semiconductor equipment companies. Technology hardware stocks followed.

At the same time, hedge funds bought shares of the Magnificent Seven for four consecutive days.

Of course, this does not mean that risk has disappeared or that the market will move higher every day.

However, it does suggest that, following the recent period of intense volatility, large investors are beginning to increase their exposure to technology once again.

In simple terms, big money appears to have started rebuilding positions in the sector.

PLTR reported impressive second-quarter results, clearly beating both analysts’ estimates and the company’s previous guidance. Revenue reached $1.94 billion, up 93% year over year, compared with expectations of approximately $1.80–$1.81 billion. Adjusted earnings per share came in at $0.41, versus the $0.35 expected by the market.

The most important highlight was once again the U.S. commercial business. Revenue from U.S. companies increased by 149% to $764 million, while the remaining deal value for this segment reached $6.2 billion, up 124%. This shows that Palantir’s growth is no longer driven only by government and defense contracts, but increasingly by businesses integrating the company’s AI platform into their operations.

Management also raised its outlook significantly. For the full year 2026, Palantir now expects revenue of approximately $8.15–$8.16 billion, compared with its previous guidance of around $7.65–$7.66 billion. For the third quarter, the company expects revenue of approximately $2.16 billion, well above analysts’ estimates of around $2 billion.

The stock is surging because Palantir did not simply report another strong quarter. The company showed that its growth is accelerating, demand for its AI products remains exceptionally strong, and management now expects significantly higher revenue and operating income for the remainder of the year. The post-earnings gain reached approximately 8%–14%, depending on the point in after-hours trading.