# Microsoft and Meta report tonight, the Fed decides at two: what the crowd is positioned for > The Fed decides at 2pm, Microsoft and Meta report after the close, and in the accounts we track both names walk in quiet and split: 34 and 49 mentions on the week, scores of 68 and 61, and one shared script, that both will beat and still fall on higher AI spending. The room's real conviction sits with the companies on the receiving end of that spending. And the Fed? About one mention in thirty. By Maya Koeva · 2026-07-29  Today is the densest day of the calendar. The Fed announces its rate decision at 2:00pm ET, with the current range at 3.5 to 3.75% and no new projections at this meeting, and the press conference follows at 2:30. Then, after the close, Microsoft and Meta report, with Apple and Amazon tomorrow, both confirmed after-market in our events data. On Monday we looked at [what the crowd was saying about all four going in](/blog/megacaps-round-two-going-in). Yesterday we [autopsied MaxLinear](/blog/signal-autopsy-maxlinear), a name that beat, raised, and fell 29% anyway. Today, the pre-print read on the two names reporting tonight, and the one thing nearly every account in the room seems to agree on. ## The read going in Here is the week into the prints, July 22 to 29, in the accounts we track: | Company | Reports | Mentions | Bullish | Bearish | Trusted | 7-day score | | --- | --- | ---: | ---: | ---: | ---: | ---: | | Microsoft (MSFT) | Tonight, after close | 34 | 50% | 38% | 15% | 68 | | Meta (META) | Tonight, after close | 49 | 41% | 43% | 20% | 61 | Both rooms are quiet, and neither has an edge worth calling. Meta is split almost exactly down the middle, 41% bullish to 43% bearish, which is about as undecided as a [sentiment read](/learn/how-to-read-a-sentiment-breakdown) gets, and the last-day burst of pre-print chatter split perfectly, six bullish posts against six bearish. Microsoft leans mildly bullish but with only 15% of its mentions coming from trusted voices; its score has firmed to 68 as the stock clawed back toward $400 this week. For contrast, Nebius drew 161 mentions over the same week and Alphabet 254. The two biggest reports of the night are, once again, far from the loudest conversations on the board. ## Most of their chatter is not even about them There is a second thing the raw counts hide. Most of the time Microsoft and Meta appear in our set, they are not the subject at all. They show up in passing, inside threads about other stocks, as the yardstick everything else gets measured against. The grey bars below are those passing references, and they dominate both charts. Loud names are not always [the signal](/learn/volume-vs-signal); sometimes they are just the scenery. *[Chart: Microsoft (MSFT) mentions in the accounts we track, July 15 to 29 (rolling 24h): a handful of direct calls, and a steady wall of passing references in other names' threads.]* *[Chart: Meta (META) mentions in the accounts we track, July 15 to 29 (rolling 24h): one mid-month burst, a quiet run-in, then a final-day surge that split perfectly, six bullish against six bearish.]* ## The script everyone is reciting Read through the actual posts and one script repeats, almost word for word, across the whole credibility spectrum. The most credible account talking about either name this week (1.00) put it like this: "Let me guess, MSFT META and AMZN beat on everything but increase capex and dump." A trusted voice on Microsoft: "Microsoft drops everytime they beat estimates, like clockwork." On Meta: "META will report explosive capex spending." Even the bull case gets told in the same grammar, one account joking that Microsoft lowering its spending guidance would pop the stock 14%. Capex is capital expenditure, the money these companies are pouring into AI data centers, and the room's fear is not the quarter, it is that spending line. The script has evidence behind it: Alphabet beat last Wednesday and fell, and as we covered yesterday, MaxLinear and Intel beat the day after and fell too. When every account expects the same reaction, that expectation is itself information, because it tells you what is already [priced in](/learn/what-does-priced-in-mean). Notably, the room's other 1.00-credibility voice pushed back on the doom half of the script: all the big cloud platforms will raise spending, "that doesn't mean they can't post beats. MSFT is more predictable with earnings." Last night the whole debate compressed into two back-to-back posts in a single thread. One: "They never stopped having strong numbers but whenever they say capex stock goes down." The reply: "Capex is priced in. Big pump incoming." That is the entire setup, in two sentences, unresolved. ## Where the conviction actually is Follow the money the script describes and you find where the crowd's real conviction sits. One account put numbers on the divide (0.49): they went cautious on the big spenders and optimistic on the companies those budgets get spent with, and counted the megacap spenders down 5.8% last week while the semiconductor index rose 1.2%. Our board says the same thing. The companies on the receiving end of AI capex are where the trusted rooms are loud and one-sided: Nebius at 161 mentions, 78% bullish, score 85. Micron at 85. MaxLinear, [as we wrote yesterday](/blog/signal-autopsy-maxlinear), holding a score of 80 through a 29% drop. Even Alphabet, after its beat-and-fall, has turned constructive at 53% bullish across 254 mentions, the biggest conversation on the board. So tonight's reports are not just about Microsoft and Meta. Their spending plans are the demand signal the entire AI-infrastructure complex trades on. The names our credible rooms are actually long need those checks to keep growing. ## The Fed at two And the rate decision that sits between now and the prints? In the accounts we track, it barely exists. Across the last seven days, about 137 of 4,481 signals mention the Fed, rates, or the FOMC in any form. That is one in thirty. Alphabet alone drew nearly twice as much conversation as the entire rate decision. This is worth reading correctly: our set skews toward stock-pickers' rooms, so it measures where stock-pickers put their attention, not whether [rates matter](/learn/how-interest-rates-move-stocks), and they matter plenty. But the crowd we track does not trade the announcement. It trades single names, and it will trade the reaction. ## What we will actually be watching Pre-print chatter measures expectation, not knowledge, the same [distinction we drew on Monday](/blog/megacaps-round-two-going-in), and nobody in these rooms knows tonight's capex number. So the useful signal arrives after the close, not before it: watch whether the trusted voices hold their read or fold it once the numbers land, the way MaxLinear's room held through a 29% drop. That reaction, from accounts with a [track record](/learn/how-a-track-record-is-graded), is the first honest verdict on the biggest night of the season. We will bring you the read tomorrow, when Apple and Amazon close out the gauntlet. --- *Quantral surfaces signals and context from public sources to support your own research. Nothing here is financial advice or a recommendation to buy or sell.*