A tale that doesn’t seem to add up on the surface is circulating in the financial community. Strong revenue growth is reported by Alphabet. Tesla outperforms on a number of metrics. And both stocks suffer severe losses in a matter of hours. In a single session, Alphabet lost almost $293 billion in market value, making it its worst day since going public in 2004. Tesla experienced a 15% decline. Both of the records. Not the kind that anyone desired, though.
What precisely is the market responding to, then? Not, it turns out, the revenue figures. The top line is no longer of interest to investors. They are looking at free cash flow, which is more profound and, in some respects, more unsettling.
The biggest names in technology have had an almost unfair advantage for the majority of the last ten years. They made huge profits and still had money to pay dividends, buy back stock, and invest in their next big venture all at once. They appeared nearly impenetrable with that combination. The so-called Magnificent Seven traded at high valuations for a considerable amount of time because of this. They weren’t merely expanding. In the process, they were making money.
These days, that dynamic is shifting in ways that seem more structural than transient. For the first time since its initial public offering (IPO) more than two years ago, Alphabet reported negative free cash flow last quarter. It wasn’t presented by management as an isolated incident. They said there would be more pressure. The full-year capital expenditure forecast was increased to $195 billion to $205 billion, with an even greater increase anticipated in 2027. Similar results were reported by Tesla, whose spending on self-driving technology and new manufacturing infrastructure caused its free cash flow to fall into negative territory for the first time in over two years.
Those two are not the only ones. Meta is anticipated to come next. Earlier this year, Amazon had already entered negative territory. Capital expenditure estimates for the year across the hyperscaler group currently surpass $750 billion, and that amount continues to rise with each earnings cycle. There is a perception in the market that investors are no longer questioning the existence of artificial intelligence. They’re posing a more difficult query: when and by whom does this expenditure truly pay off?
The structural change taking place beneath the headline numbers contributes to the understanding of the anxiety. In order to finance AI infrastructure, these businesses are taking on debt on a scale never seen before. Simultaneously, less expensive open-source AI models continue to surface, subtly endangering the extremely computationally demanding systems being constructed at such high expense. In three years, the economy might look very different. They might not, too. Right now, that uncertainty is being priced in.

Additionally, the suffering hasn’t been limited to a few megacap names. The semiconductor index is officially in bear market territory since it has dropped more than 20% from its most recent highs. Taiwan’s and South Korea’s chip-heavy markets have been particularly hard hit. A portion of that pressure stems from factors that have nothing to do with AI. For example, growing tensions in the Middle East have driven up energy prices, which has raised concerns about inflation and raised bond yields. Growth stocks, whose valuations heavily rely on future earnings, are especially negatively impacted by higher yields.
However, it’s difficult to ignore the fact that the S&P 500 as a whole hasn’t always followed in lockstep on days when chips have had their worst sessions this year. That is an important detail. It implies that rather than spreading like a virus throughout the entire market, the anxiety is concentrated in a particular area. In fact, the same week that the tech selloff was making headlines, an equally weighted version of the S&P 500 reached an all-time high. The market is not losing money. It’s shifting within it.
Quietly, Microsoft has emerged as the standard by which others are judged. It is anticipated to produce more than $16 billion in free cash flow even though it intends to double its own capital expenditures this year. Investors are aware of the difference between Microsoft and its competitors. The market notices when one company continues to print cash comfortably while everyone else’s cash generation declines.
It’s not really a rejection of AI. It’s a more developed, and possibly long overdue, kind of examination. The formula that virtually automatically rewarded hyperscalers for investing more money in AI is no longer applicable to all situations. Investors are beginning to distinguish between businesses that are likely to generate a genuine return and those that might not. Two years ago, that distinction was hardly noticeable, but now it is crucial. And it probably makes sense to exercise some caution until it’s clearer who ends up in which column.
