Tonight after the US market close (around 04:00 Beijing time on April 30th), $Alphabet(GOOGL.US) will release its Q1 earnings report.

To get straight to the point: beating earnings is highly likely, but whether the stock price surges significantly depends entirely on how management explains that $175–185 billion capital expenditure.

Why do I say that? Revenue is expected to be $106.89 billion, up 19% year-over-year, and cloud business revenue is expected to be $18.4 billion, up 50% year-over-year. These two figures are basically a given. Gemini's monthly active users have already reached 750 million, with paying users up 40% quarter-over-quarter. On the surface, everything looks great.

But EPS is expected to be down 6.4% year-over-year—that's the cost of the AI cycle. The market already knows Google is burning cash. The key is that the last earnings report directly raised 2026 capex from the market's assumed $120 billion to $175–185 billion, almost doubling it. Numbers at this scale can't be called "investment"; it's betting the company's future.

So the focus tonight isn't by how many points revenue beats, but whether management can clearly articulate the "cash burn → revenue" chain. Gemini API's throughput of 10 billion tokens per minute, the scale of the cloud backlog, the absorption pace of TPUs—these concrete numbers need to support the capex narrative. If explained clearly, the stock price can follow the Gemini story upward; if not, simply saying "we're winning in AI" won't hold up.

The options market has priced in an implied after-hours volatility of ±5.63%, nearly four times the average of 1.44% over the past four quarters. The long-short divergence is at an extreme, with neither side backing down. Trading earnings at this level is a gamble in itself.

My take: earnings can pass, but to bet on a big surge is to gamble on management telling a compelling capex story. If they just coldly read out numbers to meet obligations, even a beat won't save the stock price. For those without a position, it's not too late to act after watching the Q&A.