Four companies, one earnings window, one question that drives all the search volume: does AI spending still pay off? Meta (META), Microsoft (MSFT), Amazon (AMZN), and Alphabet (GOOG) all report Q2 results between July 22 and 30. The data heading in doesn't tell one story. It tells four different ones.
What is happening right now
Combined 2026 AI capex across the four hyperscalers tops $650B. Amazon alone is tracking $200B. Alphabet guides $175B to $185B. Meta raised its full-year target to $125B to $145B. Microsoft is on pace for $145B in its fiscal year.
That wasn't supposed to be the problem. The problem is that capex is compounding faster than the market can track the payoff. Meta fell 6% and Microsoft slipped 2.5% after Q1 beats, even though both topped earnings estimates. Investors are now discounting top-line beats and repricing the spend.
Q2 is the second accountability test. The question isn't just whether they beat. It's whether they can prove the AI infrastructure bill has a short enough payback window to justify the position.
What history says about AI capex cycles
This isn't the first time hyperscaler spending spooked markets mid-cycle. In 2022, cloud infrastructure spending surged while revenue growth slowed, triggering sharp deratings across the group. The companies that held up were the ones with the strongest fundamental profiles: high cash flow, low leverage, durable margins.
That pattern matters now. Short-term momentum can reverse fast in earnings windows. Underlying fundamentals take longer to break. That's the gap the scores are designed to show.
What Stoxcraft scores show right now
Meta and Microsoft: elite fundamentals, weak short-term signals
Meta's health score sits at 9.6 out of 10. That places it among the strongest businesses in the Stoxcraft universe on fundamental quality. Its trend reading is Falling, the second-weakest signal on the six-tier scale. Free cash flow is exceptional. Short-term momentum is not. Those two facts both need to be true at the same time to understand the setup.
Microsoft's health score is 8.7 out of 10. Its performance score is 5.1, barely above the midpoint of the full universe. That gap reflects two straight quarters of Azure deceleration landing harder than expected. The business is strong. The market hasn't rewarded it for that yet.
Amazon leads the group heading into Q2
Amazon's performance score is 8.1 out of 10, highest of the four reporting this window. Its risk score is 3.9 out of 10, meaning lower risk than the majority of stocks in the Stoxcraft database. A higher risk score always means more risk, so 3.9 is the cleaner read here.
Its overall rating is the highest of the group, and its trend signal is Surging. That combination is the opposite of what Meta and Microsoft are showing.
What investors are getting wrong about hyperscaler earnings
The crowd is treating this as one trade. A single AI capex narrative covering all four companies equally. It's not. Meta's gap between fundamentals and trend is a market credibility problem. Microsoft's gap between health and performance is a momentum problem. Amazon's setup looks like neither.
Conflating them means buying or avoiding the wrong thing for the right reason. A Falling trend on a 9.6 health score is a different risk than a Falling trend on a 5.0 health score. The underlying business quality changes the stakes entirely.
Amazon's setup is the one to watch heading into July
The score that matters most this week isn't the capex headline. It's the health-to-trend alignment. Amazon has both pointing in the same direction. Meta has them pointing in opposite directions. That divergence is the actual story ahead of earnings.
Nvidia (NVDA) and Apple (AAPL) don't report in this window, but both carry Surging trend signals right now. Every dollar the four hyperscalers spend on AI infrastructure runs through Nvidia hardware. Confident Q2 guidance from all four acts as a direct read-through for the wider AI in global markets story.
