Microsoft and Amazon Earnings Ignite a Rally — Bounce or Reversal?

On July 30, as Microsoft's and Amazon's earnings rolled in, sentiment got a real jolt ,a substantial chunk of both companies' profits still came from their data center businesses, which handed Wall Street tech investors a reassuring answer to the question that had been nagging them: is all this AI spending actually going anywhere? The semiconductor index rallied on heavy volume in response. But today's session turned into a classic gap-up, fade-down.

That whiplash is exactly what I want to think through carefully here: is this rally just an emotional bounce, or is it the start of a genuine trend reversal? I want to walk through my own reasoning across two dimensions , the fundamentals, and the price action.

1. The Fundamentals: Great Profits, But Free Cash Flow Is Exposing a Real Split

Start with the fundamentals. Across this earnings season, overall profit margins at the major players have looked genuinely strong both Microsoft and Amazon beat expectations comfortably. But switch the lens to free cash flow, and a real divergence starts to show up between companies, and that gap is widening.

The specifics: Amazon raised its 2026 capex guidance to $220 billion in one move, up more than 70% year over year, while Google raised its full-year capex guidance from a prior $180–190 billion range to $195–205 billion. Raising guidance isn't the problem by itself , the problem is that the gap between spending and the return on that investment is visibly widening. Amazon's trailing-twelve-month free cash flow has now turned into a net outflow, and Google just posted its first-ever negative quarterly free cash flow since going public. Microsoft, by contrast, is one of the few among this group still running positive free cash flow , and it didn't raise its capex guidance this round.

The biggest takeaway from this divergence, for me, is this: the market's focus within the AI narrative may be quietly shifting from "whose capex is growing the fastest" toward "who can actually turn that capital into real returns, and how efficiently." In other words, the pure "stack more compute" story is starting to lose its edge on its own. The question investors actually need answered from here is when, and how, these massive investments start paying off.

Beyond the company-level split, there are two macro variables worth keeping on the watchlist:

• The Fed held rates steady at its July 30 meeting, in line with expectations but there was internal disagreement over the inflation outlook, a reminder that "rates unchanged" is only today's conclusion, not a guarantee for what comes next. The September FOMC meeting is the most direct signal to track from here.
• Oil price volatility, and how hardware makers' capacity expansion plans reshape the supply-demand balance for AI hardware, will keep affecting both the cost side and the overall health of the AI hardware chain also worth keeping on the list.

2. The Price Action: The Sellers Are Mostly Done, But a Reversal Still Needs a Catalyst

Now the chart. The semiconductor index has moved off its prior downtrend and settled into a range-bound, consolidating pattern the money that was going to sell has largely already sold, and the capital sitting on the sidelines is waiting for a clear enough signal before committing.

That raises the obvious question: is this already a level worth buying?

My own read is: follow the market, don't jump ahead of it. The prior selloff looks more like the market rationally correcting an overly optimistic prior view on the AI sector, rather than a full trend reversal in its own right. Confirming an actual reversal still requires a stronger catalyst to drive a genuinely new uptrend not just a couple of days of emotional lift from earnings headlines.

In terms of volume-price logic, I've set myself a fairly simple rule to watch: if the market keeps building volume with genuinely new money coming in, the odds of a real reversal rise meaningfully; but if volume stays flat and range-bound, the more likely path is continued consolidation, or even a slow, grinding drift lower. In other words, rather than trying to guess which day marks the bottom, I'd rather use "is volume genuinely expanding" as the key anchor for whether a reversal is actually confirmed.

3. Closing Note

This round of market reaction points to the same conclusion from both angles: the direction isn't settled yet, and it needs more confirming signals. Earnings delivered on profit, but also exposed the growing tension between capex and free cash flow. The chart has stabilized into a range, but it's still missing the catalyst that would bring in genuinely new money. From here, I'll be watching two things closely , whether the free cash flow divergence between companies keeps widening, and whether trading volume can genuinely expand. Until those two questions have clearer answers, I'm not treating this bounce as a confirmed reversal.

This post records my own real-time thinking and reflects a personal view it is not investment advice. Market conditions move fast, and any actual decisions should be grounded in your own ongoing research and independent judgment.