Alphabet dropped almost 5% before the bell and slipped below its 100-day moving average — on a quarter that beat. This is worth sitting with, because it tells you what the market is now pricing.
The numbers as of July 23, 2026
| Metric | Reading |
|---|---|
| Q2 capital expenditure | $44.9B |
| Full-year capex outlook | raised to $195-205B |
| Q1 free cash flow | −47% to ~$10B |
| Capex as share of revenue | ~46% |
| Share reaction | −5%, below 100-DMA |
Why it moved
The raise was the problem. Alphabet has now lifted its capex guidance repeatedly through the year, and each raise lands on a market that has already stopped applauding. A year ago, a bigger AI budget read as conviction. Today it reads as a longer wait for the payoff and a deeper hole in free cash flow.
The 100-day moving average break matters less as a signal than as a description: the buyers who had been stepping in on every dip stopped stepping in.
The mechanism
Alphabet earns extraordinary operating profit from search and YouTube. That profit is now being routed into data centers, TPUs and power. Free cash flow — what's left after the building — fell 47% in Q1. The spend itself doesn't hit earnings immediately; it hits over the following years as depreciation, which means today's $200B budget is tomorrow's fixed cost regardless of what AI revenue does.
That's the asymmetry. The cash goes out now on management's timetable. The revenue comes in later on the customers' timetable.
The risk
The bear case here is fashionable and could easily be early. Alphabet is not a speculative name: search cash flow is enormous, Cloud is growing, and Gemini has real distribution through Android and Workspace that no rival can replicate. If AI revenue inflects in 2027, the current spend looks like the cheapest market share Alphabet ever bought, and anyone who sold at a 5% dip on a beat will feel foolish.
There's also a regulatory tail that has nothing to do with capex and could matter more than any of this.
My take
I don't think a capex raise is a sell signal for a business earning what Alphabet earns. But I've stopped treating the guidance raise as bullish, which is a change in my own framing this year. What I want to see before adding is the ratio moving the right way — Cloud revenue growth catching up to capex growth. Until then I'd hold rather than add, and I'd expect more days like this one, because the market now reacts to spending announcements the way it used to react to misses.
Bottom line: Alphabet fell on a beat because the market repriced the cost of the buildout, not the quality of the business. $44.9B in a quarter and a raise to $195-205B is a long wait for cash flow, and the tape is done waiting patiently.
Not investment advice.