Is Capital Rotation a Myth While AI Keeps Delivering?

HSBC Asset Management shared a note on February 23 in which they noted a “violent rotation” out of tech and artificial-intelligence into value and cyclicals sectors. According to the report, sectors such as industrials, staples, materials, and energy had posted double-digit year to date growth while tech and AI stocks had barely moved the dial. Even the broader S&P 500 was effectively flat year to date.

But as headlines scream “capital rotation”, artificial-intelligence infrastructure spending is at wartime levels. Enterprises are also adopting the technology at a faster pace.

So, if AI is genuinely delivering, why is capital running for the door from a sector that built the bull market of the last two years? Walk with us in this week’s newsletter to get a sense of this situation.

As of this writing, February 25, the S&P 500 Materials, which tracks the performance of basic materials stocks, was up 16.78% year to date. The S&P 500 Industrials, for the industrials sector, was up 13.03%, and the S&P 500 Energy, for the energy sector, was the biggest gainer at 21.10%.

Contrarily, the tech-heavy NASDAQ Composite was down 0.74% in the same period. And because tech companies have in recent years dominated the S&P 500, the result is that this broad market index has gained a paltry 1.17% year to date.

This reality confirms HSBC’s assessment that investors are increasingly moving towards less expensive sectors.

Experts already have a term for the phenomenon, they’re calling it the HALO effect. This refers to the shift towards “Heavy Assets, Low Obsolescence” stocks, or HALO stocks, as investors seek stability outside speculative software.

In fact, Barron’s Martin Baccardax advised investors on February 12 that they should think HALO if they want stocks with AI immunity. “You won’t get a burger from a chatbot, receive a package from an artificial intelligence-coded program, drink an icy cold Coke with the help of a data center, or plow a field with a large language model,” Baccardax wrote.

The interesting part is that all this is happening when the four largest hyperscalers are committing more than half a trillion dollars in artificial-intelligence capex. This is for 2026 alone, and this sum approaches or exceeds each company’s combined spending over the prior three years. Even some believe that without AI infrastructure spending, US corporate investment in equipment would be negative.

AI capex aside, the fundamentals are firmly on the side of the AI trade. So far, hyperscaler revenues remain robust. Even Nvidia, widely seen as a bellwether of the AI trade, has posted blowout results over the past several quarters. And for the coming quarter, Q4 FY2026, analysts expect the earnings per share to climb 71.9% year over year and revenue to grow 67% year over year.

In fact, Wall Street has revised Nvidia’s results upwards three times in the last 30 days. The analysts point to demand that remains supply-constrained rather than demand-constrained.

Such an upbeat environment shouldn’t witness the kind of second-guessing the AI trade is facing. To some, the issue is a crisis of confidence regarding future margins. Goldman Sachs has, for instance, framed the concern this way: the issue is not earnings today, but uncertainty around margins tomorrow. But since analysts have not slashed estimates, this suggests the market fears a future margin compression from massive capex and not immediate revenue shortfalls.

In other words, investors are thinking, well, AI is delivering in terms of earnings, but when will all this spending generate returns?

According to Bank of America, hyperscalers will spend about 90% of operating cash flow on capex in 2026. This will squeeze dividends and buybacks. And Morgan Stanley expects borrowing to top $400 billion this year, more than double 2025’s $165 billion.

This could be the biggest reason capital is gravitating toward Energy and Materials. It is not because those sectors are more innovative, but because they offer cleaner, shorter paths from investment to cash flow. This is precisely what some investors feel tech currently lacks.

In the HSBC report cited earlier, the analysts noted that “the investment clock seems to have gone haywire. Like some jet-lagged business executive, it can’t work out what time zone it is in.” For instance, cyclicals and defensives are both rallying simultaneously, which is a contradiction in a normal cycle. To the analysts, the signal here is that the rotation is less about conviction in value sectors and more about discomfort with AI’s timeline to returns.

There is also the sense that investors are shifting from builders of the artificial-intelligence technology to adopters and enablers. And if you look at this rotation from this lens, the conclusion is that capital isn’t leaving AI but only repositioning within the AI story.

For example, when the Caterpillar and GE Vernova stocks rise, the former is up 34.19% year to date and the latter 35.10%, the story is that it is because they are directly benefiting from the AI buildout. Caterpillar’s generators power data centers, and GE Vernova supplies electrical infrastructure.

So, when investors buy these stocks, they are not betting against AI but rather betting that the physical infrastructure required to sustain the AI supercycle is undervalued relative to software.

Capital rotation is happening, and it is happening aggressively. But so is AI delivering. What is happening is that AI is simply evolving from a software story to an infrastructure reality. This is to say that the market is not rejecting the AI trade but merely repricing it.