Megacap Earnings Collide with the Fed This Week

There are weeks that confirm the existing narrative, and weeks that rewrite it. Next week is the second kind. Three of the largest companies on Earth report earnings in a 48-hour window, directly overlapping with the Federal Reserve's July policy meeting. Either the AI capital expenditure story is delivering revenue or it is not. Either the Fed holds steady or it surprises. The combination of those two answers will set the direction for the second half of 2026.

The market is not walking into this blind. Alphabet reported earlier this week and the stock sank, not because results were terrible, but because management commentary pointed to accelerating AI infrastructure spending with uncertain near-term revenue. The market heard "we are going to spend more" and "we are not sure when it pays back" in the same sentence. That combination, at a 5% risk-free rate, is uncomfortable.

The AI Capex Question

The single most important question for megacap tech right now is not top-line revenue growth. It is the relationship between AI capital expenditure and AI revenue. Every hyperscaler has committed tens of billions to GPU clusters, data centers, and inference infrastructure. The investment thesis depends on those dollars converting into revenue within a reasonable time horizon.

Alphabet's stumble set the template for what the market does not want to hear: "We are spending more than we told you we would, and the revenue line is still developing." Amazon, Meta, and Microsoft each have their own version of this story, and the market will parse every line of guidance for the same signal.

• Amazon (AWS): The most pure-play cloud and AI infrastructure story. AWS revenue growth and margin commentary will be dissected for evidence that enterprise AI workloads are converting from experimentation to production spend.
• Meta: The AI spending story here is less about cloud revenue and more about whether AI-driven ad targeting and content recommendation improvements are translating into higher ad revenue per user. Meta is spending at hyperscaler scale but monetizing through a different channel.
• Microsoft (Azure + Copilot): Azure growth rate is the headline number. Microsoft also has the Copilot monetization story, which is the closest thing to a direct "AI revenue from enterprise seats" metric among the three.

The Fed Overlay

Earnings alone would make this a heavy week. The Fed meeting makes it volatile. The 10-year Treasury yield is pressing against 5%, a level that changes the discount rate for every growth stock on the planet. If the Fed signals any hawkish surprise, duration-sensitive assets (which includes every megacap tech stock) reprice immediately.

The consensus expectation is a hold with mildly dovish language. But consensus expectations have been wrong before, and the combination of persistent inflation and a labor market that refuses to soften gives the Fed room to surprise if it wants to. The timing is the problem: if hawkish Fed commentary lands on Thursday morning and big tech misses on Thursday evening, the compounding effect on sentiment could be severe.

Position Sizing, Not Direction

The temptation going into a week like this is to make a directional call. That is usually the wrong instinct when the distribution of outcomes is this wide. The right question is not "up or down" but "how much exposure do I want to a binary event cluster where I cannot see the outcome in advance?"

For concentrated megacap positions, the risk-management answer is straightforward: reduce to the size you can hold through a two-standard-deviation move in either direction without emotional decision-making. If the earnings beat and the Fed cooperates, you participate on your remaining position. If both disappoint, you keep your capital.

The Trump Trade unwind (covered in today's companion piece) is adding a secondary rotation current. Capital leaving policy-bet trades has been flowing into megacap tech and quality names. If megacap earnings disappoint, that rotation has nowhere obvious to go, which is when broad-market risk rises.

Watch the spread between the 10-year yield and hyperscaler forward P/E ratios on Wednesday and Thursday. That spread is the market's real-time answer to whether AI spending at these levels is justified at this cost of capital.