Alphabet's Stock Is a Must Buy Before July 22

The real variable is not the size of the quarter, but whether GCP is proving that Alphabet can monetize AI infrastructure before the market re-rates the capex. If cloud growth stays above a high-40s/50%-plus range and margins hold, the stock can graduate from a 'cheap mega-cap' to a durable AI compounder, which is the setup for multiple expansion rather than just an earnings beat. That would also be read as positive for NVDA demand persistence, even if custom TPUs eventually take some share in inference workloads.

The first-order loser is AMZN on relative cloud momentum: even a modest acceleration gap in GCP versus AWS can compress Amazon's cloud multiple because investors will infer that smaller share gains are being bought with the same or higher level of AI spend. The harder question is whether TPU sales are incremental or merely internal substitution; if the latter, the market may overestimate the profit benefit and underwrite a rerating that never fully arrives. Over 1-3 months, analyst revisions should track cloud backlog, operating margin, and capex intensity more than headline revenue.

Contrarian risk: the market may already be paying for AI optionality and underpricing the possibility that higher spend keeps free cash flow conversion mediocre, capping upside even with a strong print. The thesis breaks if cloud growth decelerates meaningfully, ad growth reverts, or guidance implies another step-up in data-center spend without commensurate monetization. In that case, the right response is not a growth long but a valuation reset, especially if the stock stops working above the recent forward P/E band.