Bernstein maintains Amazon at Outperform and raises the target price to 300 USD

Overview

This is a Bernstein earnings preview/valuation update report on Amazon.Com Inc. Ahead of the earnings releases of large AI capex companies in the first quarter of 2026, the report focuses on whether Amazon can break free from its repeated share-price surges and pullbacks of the past few years and achieve a relative breakout among large-cap internet and cloud computing stocks. The conclusion is positive: Amazon is maintained at Outperform, with the target price raised to 300 USD.

Core views

The core view is that Amazon can simultaneously satisfy the market's four requirements for large AI capex companies: core revenue beats, accelerating AI revenue, unchanged 2026 capex, and operating leverage released through pricing, automation, layoffs, and cost advantages. The report argues that AWS remains the absolute core of investor focus, and AWS's base cloud demand, AI revenue, Trainium/Graviton, proprietary data center capacity, and data gravity together form the bullish thesis. On the e-commerce side, price, selection, and fulfillment form a three-layer moat against the potential impact of agentic commerce on consumer entry points.

Analysis framework

The report uses a combination of earnings estimate revisions, peer comparison of cloud businesses, AI computing capacity and revenue/GW analysis, margin bridge analysis, sum-of-the-parts valuation, and DCF. For valuation, it uses a 50/50 SOTP and DCF approach: in SOTP, retail is valued at 3x 2027E EV/Sales and AWS at 25x 2027E EV/EBIT; DCF uses 9% WACC and a 3.5% terminal growth rate.

Methodology notes

• Valuation methodSOTP + DCF

  Sum-of-the-parts valuation and discounted cash flow each account for 50%

  The 300 USD target price comes from a 50/50 combination: SOTP raises valuation multiples for retail and AWS, while DCF assumes 9% WACC and a 3.5% terminal growth rate.

• Operating analysisAI cloud capacity monetization

  AI revenue/GW and capacity utilization

  The report uses AWS's online AI capacity, future Trainium capacity commitments, and room for improvement in revenue/GW to judge whether AI cloud revenue can continue to accelerate sustainably.

• Competitive analysisData gravity

  Data gravity

  A large amount of enterprise data resides on AWS, causing inference workloads to tend to run close to the data, reducing willingness to migrate across clouds and strengthening AWS's strategic value in model distribution partnerships such as with OpenAI.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• Amazon.Com Inc / AMZN

  Core covered target

  Strengths

  AWS leadership, accelerating AI revenue, cost advantages from Trainium/Graviton, data gravity, retail fulfillment network, and robotic automation.

  Weaknesses

  High capex intensity, relatively high AWS dependence on IaaS and frontier lab customers, and continued margin pressure in international business.

  Comparison

  Compared with strong tech stocks in the same period such as Meta, Google, Microsoft, and NVIDIA, the report believes Amazon has the potential to break out from the basket; in cloud business, AWS still maintains scale advantages, but Google Cloud and Azure pose competitive pressure in the AI era.

  Risks

  Regulation, continued capex increases, AI cloud commoditization, customer concentration, rising BOM costs, depreciation pressure, and e-commerce competition.

• AWS

  Core of Amazon's valuation and controversy

  Strengths

  Solid core cloud demand, rapid growth in AI workloads, 3.9GW of added capacity in 2025, and increased stickiness from self-developed chips and data gravity.

  Weaknesses

  About two-thirds of the business still leans toward IaaS; if inference workloads can be freely routed to the lowest-cost compute, it may face price competition.

  Comparison

  Azure has moved upmarket earlier by relying on enterprise software relationships and historic ties with OpenAI; Google has strengthened its presence in AI cloud through vertical integration and TPU/Gemini; AWS is filling the gap through Bedrock and OpenAI Frontier partnerships.

  Risks

  Overbuilding or insufficient capacity, reliance on Trainium discounts, and reversals in the funding cycle of frontier model labs.

• Trainium / Graviton

  AWS self-developed chips and margin leverage

  Strengths

  Trainium has received future capacity commitments of 7GW from OpenAI and Anthropic, Graviton has already been validated, and both help reduce BOM and improve profitability per GW.

  Weaknesses

  Trainium's early market acceptance was once weak, and it still needs to prove that revenue/GW can improve rather than relying on deep discounts.

  Comparison

  NVIDIA remains the gold standard, and Google TPU has been validated over multiple generations; the report believes Trainium is becoming a clear third choice, and combined with Graviton, Amazon's advantage in an integrated CPU/XPU stack is widening.

  Risks

  Chip performance, customer adoption, supply constraints, and shortened server lifespans due to AI hardware iteration.

Key data

• RatingOutperformBernstein maintains Amazon at Outperform.
• Target price300.00 USDRaised from 265.00 USD.
• Closing price255.36 USDAs of 2026-04-22.
• Implied upside17%Based on the Upside/(Downside) in the report table.
• 1Q26 / 2026 AWS revenue growth forecast27.5%The report raised its 1Q26 and FY26 AWS growth forecasts to about 27.5%.
• 2026 revenue forecast817.544B USDRaised from 814.543B USD.
• 2027 revenue forecast926.939B USDRaised from 920.302B USD.
• 1Q26 EBIT forecast21.838B USDSlightly above the high end of company guidance of 16.5B-21.5B USD.
• 2026 capex guidance200B USDJassy reiterated this figure in the shareholder letter.
• AWS added capacity in 20253.9GWThe report says AWS added 3.9GW of data center capacity in 2025 and expects it to double again by 2027.
• AWS AI revenue run-rate15B+ USDAnnualized run rate of AI revenue in 1Q26.
• AWS chip business run-rate20B+ USDAnnualized run rate in 1Q26, a significant increase from about 10B USD in 4Q25.

Impact & implications

If the report's judgment plays out, Amazon's investment thesis will shift from being 'long recommended but lacking relative returns' to a story of upward earnings revisions driven jointly by AWS AI cloud, custom chips, and retail efficiency. For portfolios, Amazon is not just part of the large internet basket, but may become a composite asset benefiting simultaneously from AI infrastructure, cloud platform migration upward, inference cost optimization, and e-commerce fulfillment advantages.

Risks

• AWS's leading position in the AI cloud era may be weakened by Google, Microsoft, or other vertically integrated competitors.
• If IaaS is abstracted by frontier AI labs into low-cost compute that can be freely routed, commoditization and price competition may emerge.
• Frontier lab customers such as OpenAI and Anthropic are highly important, but their cash burn depends on continued financing, creating customer concentration and funding-cycle risk.
• If capex continues to rise in step functions in 2027 and beyond, investors may find it hard to bear a spending cycle with no clear end point.
• Rising BOM costs, shorter server useful lives, and increased depreciation may pressure AWS margins.
• Regulatory investigations in antitrust and privacy may affect Amazon's financials and valuation.
• New investments in groceries, healthcare, AI, and other areas may raise capital intensity, compress operating margins, and reduce free cash flow.
• Share loss to new e-commerce entrants and major AWS competitors could lead to lower valuation multiples for the core business.

What to watch

• Whether AWS revenue growth in the first quarter of 2026 approaches or exceeds 27.5%.
• Whether Amazon maintains its 2026 capex guidance of 200B USD and whether it signals continued large capex increases in 2027.
• AWS AI revenue run-rate, revenue/GW, capacity utilization, and the pace of Trainium shipments.
• Commercialization progress of partnerships with OpenAI and Anthropic, especially the delivery of Bedrock, OpenAI Frontier, and the future 7GW Trainium capacity commitment.
• Whether AWS margins remain in the low-to-mid 30% range, and the pressure of depreciation and BOM costs on margins.
• Whether North American retail's 10% operating margin can be sustained, and whether robotics and fulfillment efficiency continue to reduce cost-to-serve.
• The trajectory of the advertising business, especially whether agentic commerce weakens Amazon's consumer entry point and first-party data advantage.
• 2Q26 operating income guidance, investment pressure in international business, Prime Day, and progress in cost capitalization for Project Leo.