Alphabet Q2 Preview: Brace For A Blowout EPS That Won't Mean What You Think

The market risk here is that Alphabet’s reported earnings quality gets re-rated lower even if the headline number looks strong. If a meaningful share of profit is coming from private-markup gains, then the stock should trade more like a sum of operating businesses plus a volatile venture book, not a clean cash-earnings compounder. That matters because the multiple on GOOGL is being paid for durable search/cloud cash flow; any perception that EPS is being inflated by mark-to-market gains can cap upside and widen the discount to peers with cleaner recurring monetization.

Near term, the next 1-3 months are about whether analysts and investors strip out the non-operating noise. A larger Anthropic mark would mechanically support EPS, but it also raises the odds of a sharper giveback later if private AI valuations cool or if the market starts normalizing earnings ex-investments. Second-order, this is mildly supportive for the broader AI complex and compute suppliers only insofar as it signals continued capital inflow to frontier AI, but it is negative for software multiples if investors conclude the sector is being propped up by paper gains rather than operating conversion.

Contrarian view: consensus may be focused on dilution and capex while missing the bigger issue — volatility in other income can make operating momentum look better than it is. The thesis is falsified if Alphabet can show strong ex-investment operating income, cloud margin expansion, and minimal dependence on other gains. If that happens, the markup debate is noise; if not, the stock should de-rate on a cleaner earnings basis over the next 1-2 quarters.