|
The Market Learned to Tell the Spenders Apart
For two years, owning the artificial-intelligence build meant owning all of it. When one hyperscaler raised its spending, the group moved together, and the debate was simply whether the whole trade had gotten too expensive. Last night that changed in the space of a few minutes. Microsoft and Meta, the two heaviest spenders in the group, reported almost simultaneously, and the market pulled them in opposite directions.
Microsoft came in strong. Fiscal fourth-quarter revenue reached $90.0 billion against an $87.6 billion estimate, up roughly 18 percent, and Azure, the cloud engine that is supposed to justify the whole outlay, accelerated. After a year in which the stock lost close to a quarter of its value on fears the spending would never pay off, that acceleration bought relief. Meta told a harder story. Revenue grew 28 percent to $60.8 billion, a figure that in another era would have been celebrated, and the shares fell anyway.
What went wrong for Meta was on the bottom line. Net income dropped 14 percent to $15.85 billion. Earnings per share fell to $6.18 from $7.14 a year earlier. The operating margin collapsed from 43 percent to 31 percent as the cost of the AI buildout landed on the income statement, and management nudged the floor of its spending range higher. Revenue up 28 percent, profit down 14 percent: that gap is the sound of capital spending outrunning the return on it.
The tell is in the history. Three months ago Microsoft fell about 10 percent after a quarter that also beat, punished for the same heavy spending that drew applause last night. The company spent just as heavily this quarter and showed, through Azure, that the money is converting to revenue. That distinction is the whole story, and it is a more demanding standard than the one that governed the past two years. The buildout will still be financed, selectively, for the names that can show the return arriving.
The same discipline runs at the macro level. A day earlier the Federal Reserve held rates and split three ways, with three officials voting to raise, the most hawkish dissent since 2016. This morning the government releases second-quarter growth and the June reading of the Fed’s preferred inflation gauge, with core prices expected to ease toward 3.3 percent from May’s three-year high. June captured a brief lull in energy costs before oil turned back up about 5 percent this week on fresh strikes near Iraq, so even a soft reading is a backward look the committee’s hawks can set aside. Money is getting more discerning from the central bank down to the single stock.
Where that leaves you. Apple and Amazon close out the group after the bell tonight, and they report into the standard the market just set: growth alone will not carry a stock, and spending alone will not sink one, so long as the spending can be shown to work. For a portfolio, the lesson is quieter than any single day’s move. The AI trade has stopped being one decision to own or avoid and become a series of separate judgments about which balance sheets are earning their ambitions. That is harder work. It is also healthier.
Last night gave the market its cleanest read in months. One company showed its spending turning into revenue and was rewarded. Another showed its spending turning into thinner margins and was not. Tonight two more report into that same test, and the sorting goes on.
|