|
The Market Changed Which Line It Reads
Alphabet reports Wednesday evening, and the setup looks friendly. Twenty-eight of thirty-three analysts carry a buy, the average target implies mid-teens upside, and the models all point at one figure: whether Google Cloud can hold its roughly 63% growth. But the tape already told you what it will actually grade. On July 16, TSMC delivered the best quarter in its history and the stock fell 4%, because it raised its capital spending. A record was not enough.
The tell moved from the top line to the cash line. For three years an AI name beat on revenue and was rewarded, and the spending behind the beat was treated as a moat being dug. That reflex has inverted. Now the beat is assumed, and the spending is the suspect. Alphabet can print 63% cloud growth on Wednesday and still fall if the capex guide climbs again and free cash flow keeps thinning. The question is no longer how fast the top line grows. It is what is left at the bottom after the build.
The math is starting to bite. Look at Alphabet’s own first quarter: capital spending up 107% to $35.7 billion, free cash flow down 47% in the same three months, and $80 billion raised in equity to keep the data centers coming. Pull back to the whole group and Epoch AI’s estimate lands with weight, that the combined free cash flow of the largest cloud builders reaches zero this summer. A company can outspend its cash flow for a while on the strength of a story. The market has now decided it wants to see the return instead of the story.
This is a concentration problem, not just a Google problem. The Magnificent Seven now make up 32.5% of the S&P 500. If investors re-rate AI spending from an asset to a liability, the same handful of names that carried the index to its highs are the ones that carry it back down. An index fund holder owns 500 companies in name and a fistful in weight, and that fistful is exactly the group being asked, this week, to prove the spending was worth it.
Where that leaves you. This is not a case for selling the giants into their prints. It is a case for not adding to them on the old assumption that a revenue beat settles the matter, because that stopped being true on July 16. Cash still pays close to 4% while the question resolves, and with no cut priced this year that yield is not going anywhere. When the release crosses Wednesday evening, read the capital spending line and the free cash flow line first. The cloud headline is the number the market already knows how to ignore.
For three years the market paid for the spending and called it a moat. On July 16 it started calling it a bill. Wednesday, Alphabet finds out which one investors think it is, and seven days later a Fed with no cut to give gets the same answer.
|