For one day, the market went green without AI leading it.

That should not be remarkable. Yet in the market we have been trading, it was.

On October 20, the Dow Jones Industrial Average gained about 1%, the S&P 500 added 0.3%, while the Nasdaq Composite was roughly flat. Healthcare was among the strongest parts of the market, and Eli Lilly closed at $1,283.66, up about 3.5%.

At the same time, parts of the mega-cap trade were softer during the regular session. The AP closing report highlighted declines in Apple and Meta as a drag on the Nasdaq.

For a few hours, the market looked broader.

Then the closing bell rang, new information arrived, and AI came straight back.

Super Micro Computer said preliminary fiscal first-quarter revenue was expected at $18.5 billion to $19.5 billion, above its own previous guidance of $14.5 billion to $15.5 billion and the $15.23 billion LSEG consensus quoted by Reuters that day. The company described the figures as preliminary and unaudited, subject to revision. Its shares rose after hours, and Dell and HPE were also reported higher later in the extended session.

SpaceX then reported third-quarter revenue of $14.6 billion, up 161% year over year, with net income of $4.4 billion. Its Super Intelligence business generated approximately $7.6 billion of revenue, making AI infrastructure an increasingly important part of a company still associated in many investors' minds with rockets and Starlink.

The interesting part was not that AI "came back."

It was that both halves of the session could be rational at the same time.

The market can question valuation and still reward exceptional growth.

That tension is becoming one of the defining investment questions of this cycle.

The rotation was not an anti-AI vote

It is tempting to describe any session in which healthcare, industrials, financials or smaller companies outperform technology as investors "leaving AI."

I think that interpretation is too simple.

A one-day rotation does not tell us that the AI investment cycle is over. It tells us that capital still has alternatives.

When a small group of companies has carried an unusually large share of index returns, even a modest change in expectations can create visible rotation. Investors do not need to conclude that AI is a bad business. They only need to decide that, at a particular price, the next dollar of capital may have a better risk-adjusted home somewhere else.

Healthcare is a useful example.

The attraction is not that healthcare somehow produces "real value" while technology does not. The distinction is between different combinations of growth, visibility, valuation, cash-flow duration and expectations.

A pharmaceutical company with powerful product economics can offer genuine growth without requiring an investor to underwrite the same assumptions embedded in a high-duration technology valuation. Eli Lilly's strength around Mounjaro and Zepbound illustrates why investors can find growth outside the traditional AI leadership group.

That is market breadth, not a rejection of innovation.

Great businesses can still be difficult investments at the wrong price

One of the easiest mistakes in investing is to confuse business quality with investment quality.

They are related, but they are not the same thing.

A company can have exceptional technology, dominant market share, expanding margins and a huge addressable market. None of those facts automatically answer the question that matters to the investor:

What return is available from the price I am paying today?

Valuation is the bridge between a company's future and an investor's future return.

When expectations become extremely high, the company does not merely need to perform well. It needs to perform well relative to what the price already assumes.

That is why high-multiple markets can produce a strange result: very good news can lead to a weak share-price reaction, while a small disappointment can cause an outsized decline.

The underlying business may still be excellent. The problem is that excellence was already in the price.

This is also why the phrase "AI bubble" is often less useful than it sounds. There is no single AI valuation and no single AI business model. Semiconductor companies, cloud infrastructure providers, application businesses, data-center operators and companies monetising AI through existing platforms have very different economics.

The right question is not whether AI is overvalued.

The right question is where expected returns still compensate investors for the assumptions they are being asked to make.

Then earnings reminded investors why AI became the story

Super Micro's update was a good example of the other side of the argument.

The company did not simply offer another narrative about future AI demand. It raised its preliminary quarterly revenue expectation to $18.5 billion to $19.5 billion and indicated gross margins of 12.0% to 12.5%, both materially above its previous guidance.

That matters because one of the central concerns around the AI buildout is whether extraordinary capital expenditure will translate into extraordinary economic output.

Strong orders and revenue from infrastructure suppliers do not settle that debate, but they demonstrate that the spending cycle remains powerful.

SpaceX's results made the point even more dramatically. Third-quarter revenue reached $14.6 billion, 161% above the prior year, while its Super Intelligence segment generated about $7.6 billion. The company also spent heavily on AI infrastructure.

This creates a fascinating valuation problem.

If the earnings base is expanding fast enough, a valuation that looks extreme on trailing numbers can compress rapidly. But if the growth disappoints, or if the capital required to sustain that growth becomes larger than expected, the same valuation can become unforgiving.

That is why investors need to separate three questions:

  1. Is the technology important?
  2. Is the company likely to capture meaningful economics from it?
  3. Does today's price leave enough return for the investor after those expectations are accounted for?

The first two can be answered "yes" while the third is still "not at this price."

Market breadth matters more than one day's leaderboard

For portfolio managers, the most useful part of the October 20 session was not which sector won the day.

It was the reminder that a healthy market does not need one narrative to explain every return.

Broader participation matters because concentration creates hidden portfolio risk. An investor may believe they own a diversified equity portfolio because they hold an index, several ETFs and a number of individual stocks. But if those holdings are all economically dependent on the same AI capital-expenditure cycle, interest-rate assumptions or mega-cap leadership, the diversification may be weaker than it appears.

This is particularly relevant for multi-asset investors and wealth portfolios.

Diversification should not mean mechanically selling winners. Nor should it mean avoiding structural growth themes. It means understanding what actually drives each position and ensuring the portfolio is not dependent on one outcome.

A portfolio can participate in AI while also owning healthcare, financials, quality fixed income, selected GCC exposures, real assets or other return drivers whose economics are different.

The objective is not to predict which theme wins every session.

The objective is to build a portfolio that does not require one theme to win every session.

What I would take from the session

I would not call October 20 a regime change.

One trading day is not evidence of a durable rotation, and after-hours trading immediately demonstrated how quickly strong company-specific fundamentals can pull capital back toward AI.

But I would take three lessons from it.

First, valuation discipline is not dead. When leadership becomes crowded, investors will look elsewhere when the expected return no longer appears attractive enough.

Second, earnings still matter more than narratives. Super Micro and SpaceX did not revive interest in AI by repeating the long-term story. They delivered numbers that forced investors to update their assumptions.

Third, breadth is valuable. A market in which healthcare, industrials, financials, smaller companies and technology can each lead for different fundamental reasons is more interesting to an allocator than a market where every decision reduces to owning or avoiding one theme.

The AI investment cycle may still have a long way to run. The infrastructure spending is real, the revenue growth in parts of the ecosystem is real, and the productivity opportunity may prove enormous.

None of that removes the need to ask what is already priced in.

That is the discipline investors sometimes forget during powerful themes.

For one session, the market remembered it.

Then strong earnings reminded everyone why the theme became powerful in the first place.

A great company and a great investment are not always the same thing. Price is what connects the two.

Sources and data notes

This is a dated market commentary using the 20 October 2026 close and company updates released that day. Super Micro's estimates were preliminary and unaudited; SpaceX figures are management-reported results. The portfolio interpretation is the author's own analysis. After-hours prices are time-sensitive.

About the Author

Ahmed Eisa is a UAE CMA-Accredited Investment Portfolio Manager with more than 16 years of experience across regulated banking, wealth management, multi-asset portfolios, investment products, Islamic finance, digital wealth platforms and digital assets. Based in Abu Dhabi, he focuses on portfolio strategy, GCC and global markets, regulated digital-asset investment, and the intersection of traditional investment management with emerging financial infrastructure. He is also the founder of EisaX Intelligence.

Ahmed Eisa
UAE CMA-Accredited Investment Portfolio Manager
Abu Dhabi, UAE
21 October 2026

This article reflects the author's personal views and is provided for informational and educational purposes only. It does not constitute investment advice, a recommendation, an offer, or a solicitation to buy or sell any security or financial product. Market information is time-sensitive and may change after publication.