Citizens reiterates Alphabet stock rating on TPU chip sales outlook

Citizens reiterated a Market Outperform on Alphabet with a $515 price target, citing TPU chip sales that could reach about $3 billion in 2026 and $25 billion in 2027. The firm raised its 2026 revenue estimate by roughly 1% and its 2027 estimate by about 5%, while increasing cloud revenue forecasts by 3% and 17%, respectively. Alphabet shares are already up about 133% over the past year and trade near $383.25, just below the 52-week high of $387.38.

Analysis

The market is starting to assign Alphabet a software-plus-hardware monetization mix rather than a pure ad/cloud multiple, and that matters for duration. If TPU revenue scales as expected, the stock can re-rate on improved visibility into a second profit engine with higher strategic value than incremental cloud share gains alone. The bigger second-order effect is competitive: custom silicon reduces Alphabet’s dependency on merchant GPU supply, which can compress the bargaining power of hyperscaler chip vendors and potentially tighten availability for smaller AI players.

The near-term risk is that the market is already paying for a lot of this optionality, so upside may come in lurches around capacity disclosures rather than linear multiple expansion. If TPU commercialization is slower to convert into margin than investors expect, the stock could de-rate even while revenue rises, because the market will treat hardware attach as lower-quality growth until it proves gross margin durability. Watch for any sign that AI capacity is being diverted from external cloud demand to internal model training, which would cap cloud acceleration and reduce the perceived scarcity premium.

For the broader AI complex, this is mildly negative for names whose thesis depends on constrained external supply and elevated inference pricing, because Alphabet is effectively self-funding its own compute stack. That is most relevant over a 6-18 month window: the more convincing Alphabet becomes as a vertically integrated AI platform, the more pressure builds on standalone AI enablers to defend growth with pricing or product differentiation rather than just scarcity. The contrarian angle is that the consensus may be underestimating how quickly TPU economics can normalize AI capex ROI across the industry, which would favor the largest balance sheets and hurt smaller, high-multiple beneficiaries of the same narrative.

Key Decisions for Investors

• Stay long GOOGL, but structure it as a 6-12 month core position rather than chasing here; upside exists if TPU monetization compounds, but entry should be on post-earnings volatility for better asymmetry.
• Buy GOOGL Jan-2027 call spreads to express the thesis that TPU revenue becomes material over the next 12-24 months; cap premium given elevated valuation and event risk.
• Short a basket of smaller AI hardware/compute beneficiaries against GOOGL on a 3-6 month horizon; the cleanest expression is long GOOGL / short a high-multiple AI infrastructure proxy, targeting multiple compression as hyperscalers internalize compute.
• If using options, buy downside protection on GOOGL into the print only if implied vol is cheap relative to realized; the main reversal risk is disappointment in margin quality, not revenue headline miss.