.jpg){kind=link}
The most revealing part of Wednesday’s earnings evening was how quickly one technology story became two company stories. Microsoft and Alphabet both reported expanding businesses. The market did not reward them in the same way. I do not think the useful response is to decide which stock has settled the argument about AI for everyone else.
The useful response is to ask what each report changes in its own valuation. Revenue growth can demonstrate demand. Operating profit can show something about the cost of serving that demand. Neither, without the rest of the statements, supplies the return on all the capital that management still plans to commit.
I would begin with the business before the comparison. Microsoft’s software distribution and cloud offering are connected, but that does not make every new cloud workload equivalent to another software subscription. Alphabet’s advertising operation and cloud business have their own relationship. They finance and reinforce parts of the same company while retaining different commercial characteristics.
Headline earnings per share can be affected by investment gains and losses. Operating results can be presented under different segment structures. Revenue growth may include several kinds of products sold to customers with different spending patterns. If I ignore those distinctions, I can produce a neat comparison that answers the wrong question.
This is also why I would not take an extended-hours share move as a ranking of business quality. A stock responds to what investors expected and what they were already paying for. A company can improve and still disappoint a demanding starting price. Another can deliver less absolute growth and create a better surprise.
The fair comparison requires a common research framework, not identical assumptions. I want to know how customers pay, what service is delivered, how much infrastructure is needed and how much cash remains after the investment. The answers may justify different valuations. They should not be forced into one convenient sector multiple.
I would treat the split as a prompt to revisit the assumptions that can now be tested. Which business line supplied the improvement? Did the improvement reach operating profit? What remains contingent on future deployment or a sustained pace of customer spending? Those questions are narrower than asking whether the market still believes in AI, and that is their advantage.
There is a reasonable bullish case here: the technology is producing paid usage and growth that public companies can report. There is also a reasonable demand for discipline: the cost of expanding that usage has to be earned back. I do not need to deny the first point to insist on the second.
For me, the next revision to a valuation should follow the newly disclosed evidence, not the color of the after-hours quote. The strongest company-specific argument will survive being written without a sentence about what the rest of the AI trade did that evening.