How AI Turned into a Market Headwind (Note: this is long and for some, if you want the bottom line, move down to the “Overreaching Point”)
The biggest “event” of 2026 so far has been the fact that AI market-related news has transitioned from being a consistent positive for stocks to suddenly becoming a direct headwind, as concerns about the ROI of AI spending by major tech firms and potential AI disruption of various sectors pressured stocks last week (and really since November).
Given the focus on AI and the recent market volatility, I want to take some time to tell the “story” of what’s happening with AI in plain English—because I know investors I talk to want to know what’s happening. I believe this story will help you 1) Explain why tech is weak, 2) Why the market is volatile, and 3) What’s next for AI.
Before we explain what’s changed, however, we have to establish important context. For the past three years (even since the AI bull market started), investors have viewed AI as universally positive for stocks, specifically that AI was going to 1) Boost productivity and 2) Increase corporate profits and be positive for corporate earnings (for all companies). That’s why the simple mention of any AI build-out/integration was a positive catalyst for virtually any stock in the market since 2023.
Additionally (and importantly), investors had the view that no amount of money spent on AI was “too much” because it was all viewed as money that would grow earnings. So, the more a company spent to build out AI, the more they would grow and the faster the stock would appreciate!
The peak of this thinking came in September of last year, highlighted by a huge 30% daily rally in Oracle (ORCL) after it revealed a massive backlog of orders. However, in October (and immediately following the ORCL results), sentiment towards AI began to change.
Problem 1: What if OpenAI Doesn’t Actually Spend All This Money? After further review, the ORCL backlog was almost totally attributable to one company: OpenAI. OpenAI, which created ChatGPT, had committed to spending more than $1 trillion with various AI infrastructure tech companies like ORCL, NVDA, AVGO, MSFT, and others. Expectations of these payments and the massive earnings growth they would provide were key in fueling the huge rallies in many major tech stocks (which, coincidentally, powered the markets higher over the past three years).
However, that meant part of this tech-driven bull market was almost entirely attributable to the commitments of just one company. So, what would happen if OpenAI couldn’t obtain financing to make these payments in the future? Or what would happen if another competitor began to take market share and hurt ChatGPT. In that case, the gains in many tech stocks would be premature, and those stocks would see their multiples contract as expected earnings growth wouldn’t materialize.
Problem 2: Gemini. In November, parts of those fears were realized. Alphabet (GOOGL) released an update to its AI LLM named Gemini. The update was so good that, based on many testing metrics, Gemini now outperforms ChatGPT. This was the initial catalyst that shook the positive AI mantra, and here’s why...
First, if Gemini takes market share from ChatGPT, then OpenAI may not have the money to fulfill the $1 trillion in obligations made to major tech companies. That means lower multiples and declines for the AI hyper scalers and infrastructure names (MSFT/NVDA/ORCL, etc.).
Second, Gemini is being built on Google’s proprietary semiconductors. This is a big deal because the reason that Nvidia, Broadcom, Taiwan Semi, and others have exploded in recent years was because of insatiable demand for their semiconductor chips, as they are necessary to build out LLMs. Google making its own chips implies demand for chips from NVDA, AVGO, and TSMC may be less than expected. That means less earnings growth and a lower multiple for semiconductor stocks.
Finally, if Google can make Gemini as good as ChatGPT on its own chips, then others likely can as well. The fear is that AI becomes commoditized, making trillions of dollars in AI infrastructure investment foolish.
Put plainly, Gemini broke the idea that all money spent on AI was “good” money that would result in earnings growth. Instead, it ushered in scrutiny of AI capex spending, and that altered the paradigm AI/tech stocks in which it existed. Practically, that means it’s no longer the case that the company that spends the most on AI infrastructure “wins,” and we can see that in the market reaction to the collapse of mega-cap free cash flow.
Problem 3: Cannibalization. This concern has emerged more recently, as AI advancements have created worries that AI will disrupt not just white-collar office jobs but also entire industries that are integral to the markets. That fear is why Claude Cowork and Claude Legal crushed software stocks two weeks ago and why the brokers and transports were hit so hard last week, as fears erupted that AI will disrupt the entire industry. Those AI anxiety fears spreading beyond tech are negatively impacting the rotation trade that’s supported stocks since October, as ostensibly no sector is safe from fears that AI could disrupt it.
Here’s the overarching point we need to understand: The bull market of the past three years has been driven by two factors: 1) Increased earnings expectations (as tech companies massively grew profits amidst surging AI demand) and 2) Multiple expansion, as investors viewed AI as a productivity boosting machine for all industries, leading investors to pencil in better future earnings growth. Gemini, cannibalization, over-reliance on OpenAI, and other recent AI headlines are eroding those two beliefs.
For this to stop, we need a proof element to appear that shows AI capex is generating a positive ROI and that AI will not destroy entrenched industries, but instead make them more efficient, productive, and boost earnings. Until we get that, we can expect mixed sentiment and elevated volatility.
Source: Sevens Report 2/17/16