AMDJuly 22, 2026 at 12:00 AM UTCSemiconductors & Semiconductor Equipment

AMD: Anthropic Speculation Adds Upside, But Execution Risk Remains

Read source article

What happened

A Seeking Alpha article speculates that AMD may announce a partnership with Anthropic, potentially unlocking additional AI chip revenue and pushing EPS to $20-30 by 2028, well above consensus of $19. However, the latest DeepValue master report maintains a WAIT rating at $500.9, citing a P/E of 162.6 and no margin of safety. The report emphasizes that the current valuation already discounts significant AI wins, and key risks include Helios rack-scale platform delays, Meta/OpenAI deployment timing, and custom silicon competition from Amazon, Google, and Meta. While an Anthropic deal would be a positive catalyst, it does not address the near-term execution hurdles that must be cleared in the next 6-9 months. Until the company provides concrete proof of revenue conversion from its existing multi-gigawatt agreements, the risk-reward remains unattractive.

Implication

The Anthropic speculation does not change the fundamental risk-reward calculus. The stock is priced for perfection at 162.6x earnings, and the DeepValue report's $430 attractive entry point offers a 14% downside from current levels. Investors should wait until the July 23 Advancing AI event and Q2 earnings provide hard evidence of MI450/Helios production timing, or until the stock pulls back to a level that builds in a margin of safety. In the long run, if AMD secures an Anthropic deal and delivers on Meta/OpenAI deployments, the $20-30 EPS scenario supports a $1,000+ price, but this is a 2-3 year outlook with significant execution risk.

Thesis delta

The speculative Anthropic deal adds a potential upside catalyst to AMD's bull case, but does not alter the core thesis that the stock is overvalued relative to near-term execution risks. The DeepValue report's WAIT rating remains appropriate because the article provides no concrete evidence of a deal, and the existing warrants with Meta/OpenAI already imply significant future dilution. The thesis shifts from 'wait for deployment proof' to 'wait for deployment proof or a better price' as the potential upside from Anthropic is too distant to justify current valuation without execution milestones.

Confidence

Medium