Axelrod Research — GOOGL: A $351 Price Still Needs an Earnings Thesis
The live question is not whether Alphabet owns important AI infrastructure; it is whether the evidence in hand supports paying today’s price for it. It does not: the underlying research establishes only a partly verified personnel connection to Google’s TPU programme, not a quantified earnings bridge, so the disciplined call is AVOID until primary evidence closes that gap.[1][2]
GOOGL — AVOID
Why now. GOOGL closed at $351.16 on 22 September 2026, after trading between $350.22 and $364.17 that session.[3] At that price, a memo whose central claim is merely that Amir Salek led Google’s TPU programme before joining Anthropic is too thin to support action; the memo itself classifies that claim as partly verified or unsupported and says it is contextual rather than an Alphabet thesis.[1][2]
The evidence. The 28 August 2026 analysis identifies advertising as Alphabet’s largest revenue source, with Cloud, subscriptions/devices and Other Bets also contributing, but it contains no complete segment breakdown, current valuation, cash-flow analysis or verified link from the personnel claim to revenue or profit.[1][2] That absence matters more than the narrative: no TPU-related revenue, margin, contract value or earnings contribution was established as of 28 August 2026.[1][2] The only fresh market fact needed for today’s decision is the adjusted 22 September 2026 close of $351.16; price has become precise while the thesis has not.[3]
Levels & triggers. There is no entry level yet because the present evidence cannot support a defensible valuation range.[1][2] The call can move to HOLD only after an issuer filing or investor-relations release quantifies the relevant AI/Cloud economics—revenue, operating margin and capital intensity for a named period—and supports a valuation at the then-current price.[1][2] It can move to ACCUMULATE only if that evidence also shows durable incremental free cash flow at a price offering a clear margin of safety. The thesis is killed, rather than merely delayed, if subsequent filings show AI infrastructure spending rising without corresponding segment revenue and margin conversion.
Horizon. This AVOID is valid through the next results and filing cycle; it is an evidence call, not a short-term price forecast.[1][2]
The bear case to this call. The strongest reason restraint is wrong is that TPU capability may already be strengthening Cloud economics faster than reported disclosures reveal. If the next primary-source results quantify that conversion, waiting will have sacrificed some upside—but that is preferable to treating a personnel biography as an earnings model.
What I’m watching: Alphabet’s next results package for a dated Cloud revenue and operating-margin print, plus any quantified disclosure connecting AI infrastructure investment to incremental free cash flow.[1]
Sources
Independent equity analysis, for information only, not investment advice.