Analyst reactions across the megacap wave formed a reaction ledger distinct from the individual prints. MSM covered each company's results separately; the paper reads them as a batch to determine which AI-capex stories the market believed and which it didn't [1].

Amazon posted $200.6 billion in revenue with AWS growing 37%, but its trailing twelve-month free cash flow turned negative at $7.6 billion [2]. Apple delivered record iPhone growth but guided next-quarter revenue below expectations, with Tim Cook warning of a "hundred-year flood" in memory pricing [1]. The divergence between demand acceleration and cash-flow deterioration is the batch's central tension.

Reddit and Coinbase reported the retail-investor segment's health, while Mastercard's results reflected consumer spending resilience despite war-driven inflation [1]. Each company is a different lens on the same question: can revenue growth outrun the cost of building for AI.

The synthesis matters because individual earnings stories lose context when separated from the batch. Amazon's negative FCF reads differently next to Microsoft's $678 billion order book and Meta's positive but shrinking cash generation [2]. The market priced the batch, not the companies — and the batch's verdict was that the AI buildout's cost is now visible in the numbers, even when the revenue is too.

UBS analysts estimated that operating cash flow at hyperscalers will be overtaken by their cash capex, raising the risk of AI spending in 2028 coming in below 2027 levels [1]. That projection is the analytical thread connecting the individual prints into a single narrative about the sustainability of the current investment cycle.

-- HENDRIK VAN DER BERG, Brussels