Amazon CEO sees AI doubling his prior AWS sales projections to $600 billion by 2036

Andy Jassy said AWS could reach $600 billion in annual sales—double his prior $300B estimate—citing AI-driven upside. AWS booked $128.7 billion in 2025 sales, up 19% year-over-year, and Jassy's projection implies roughly a 17% annual growth rate over the next decade. Amazon shares were up about 1% to $213.87 on the commentary.

Analysis

Cloud incumbents and the upstream hardware vendors that supply capacity (servers, networking, power/cooling) are the primary asymmetric beneficiaries of an AI-driven lift in demand for external compute — the incremental dollars flow disproportionately to whoever controls the rack-level procurement cadence and long-term contracts, not the model vendors. That amplifies the case for names that can scale procurement, logistics and margin capture quickly (server OEMs, data center REITs, high-density networking) while compressing returns for small cloud players who must compete on price and data transfer economics.

A plumbing-intensive build cycle creates a multi-stage investment horizon: an initial 0–6 month re-rating as guidance and bookings beat, a 6–18 month capex surge that tightens supply chains (memory, GPUs, PSUs, racks) and a 12–36 month risk of inventory digestion and margin normalization. Key reversal triggers are a sharp macro slowdown that freezes enterprise AI projects, rapid price erosion from hyperscalers using scale to lower unit economics, or a sudden acceleration in on-prem / private AI deployments that bypass public-cloud spend.

The consensus underweights second-order supply effects: an intensive GPU/server buildout will inflate OEM orderbooks then flip to cyclical oversupply when hyperscalers standardize on fewer validated SKUs, benefiting large-volume integrators at the expense of smaller boutique OEMs. Concurrently, software monetization plays (mobile adtech, app monetization) are exposed to slower-than-expected re-acceleration of ad budgets; they are exposed to the long tail of developer adoption rather than the immediate capex-driven tailwind enjoyed by infrastructure suppliers.

Key Decisions for Investors

• Long AMZN equity or synthetic long via a 9-month call spread: buy 1x 9m ATM-to-+25% call spread (risk = premium ~1x), target asymmetric pay-off of 2x+ if cloud guidance/AI bookings re-accelerate; haircut position by 30% of size on a 12% drawdown and reassess on next earnings cycle.
• Tactical long SMCI (server OEM) for 3–9 months: initiate 2–3% portfolio weight with a stop-loss at -18% and a price target of +40–60% reflecting near-term orderbook visibility; trim to half at +25% since OEMs are first to benefit but face 12–18 month oversupply risk.
• Speculative long APP via 6-month 20% OTM calls (small size, <0.5% portfolio risk): thesis is optionality on faster mobile monetization and AI-driven app engagement; max loss limited to premium, target 3x if ad RPMs re-accelerate and user LTV expands.
• Hedge / cash-flow layer: sell 1–2 month covered calls on AMZN (near-term strikes ~5–7% OTM) to monetize theta while retaining directional exposure; if shares gap >12% on positive catalyst, buy back calls and re-establish longer-dated exposure.
• Risk-take hedge: allocate 0.5–1% to buying puts on cyclical small-cap OEMs/hosts (not core hyperscalers) for 6–9 months to protect against a fast pivot to on-prem or a macro-led pause in enterprise AI spend — these names suffer first in demand drawdowns.