The Best Warren Buffett Stocks to Buy With $900 Right Now

AWS revenue was $128.7B in 2025 (+20% YoY) with AWS operating income of $45.6B, while Amazon’s e-commerce generated $588.19B in sales but only $36.36B in income. Berkshire’s notable holdings include American Express (22.1% stake) with AmEx Q4 net interest income of $4.5B (+12% YoY), Moody’s (13.9% stake) with 2025 revenue $7.71B (+~9% YoY), net income $2.45B (+19%) and EPS $13.67 (+21%), and The New York Times (≈3.2% stake) which added 450k digital-only subscribers to 12.78M and saw digital ad revenue +25% YoY. Berkshire’s five Japanese trading-house positions have risen from an initial combined value of ~$6.3B to about $41B (each stake now >8%), underscoring continued long-term allocation to diversified, cash-generative assets.

Analysis

Amazon's cloud dominance creates an option-like exposure to AI infrastructure spend for long-duration investors; the non-obvious lever is that incremental AWS margin accrues disproportionately to software and interconnect vendors (NICs, FPGA/ASIC suppliers) and to regions with favorable data‑center economics, not to unit e‑commerce throughput. That implies AWS share gains will amplify value upstream (chip cycle, power & cooling suppliers) and downstream (SaaS vendors reducing on‑prem costs), concentrating winner-take-most dynamics over 12–36 months.

Financial incumbents tied to fee-bearing, higher‑income customers (AmEx, Moody's) are positioned to convert macro stagnation into higher per-customer revenues, but they're simultaneously exposed to episodic credit cycles and regulatory shifts; a 2–4 quarter deterioration in delinquencies or increased regulatory scrutiny (pricing, ratings competition from AI) would materially compress multiples. Media plays with subscription-first models face a bifurcation: scale buys pricing power in digital ad markets, while regional ad weakness and programmatic displacement can erode ARPU within 3–8 quarters.

Second-order winners include data‑center real estate and power-infrastructure contractors (long cycle capex) and boutique analytics vendors that sell next‑gen credit/risk tooling to replace legacy ratings workflows. The contrarian corner to watch: market positioning assumes continued consolidation around incumbent clouds; a rapid shift to open-source LLMs on cheaper commodity GPUs or an aggressive multi‑cloud procurement cycle could cap AWS pricing power and re‑rate the sector within 6–18 months.

Key Decisions for Investors

• Long AWS/AI capture: buy AMZN Jan-2028 1,200/1,700 call spread (debit) sized to 1–2% portfolio exposure. Rationale: levered, multi-year exposure to cloud AI TAM while capping premium. Risk control: cap loss to premium; take 50% profits if AMZN rallies 40% in 12 months.
• Play fintech resilience vs payments: pair trade — long AXP (40% allocation of the pair) and short MA (60% allocation) for 6–12 months. Rationale: express yield/loan optionality in AmEx vs pure network exposure; target +20–30% relative outperformance. Stop: tighten if AXP underperforms MA by >15% in 3 months.
• Credit ratings convexity: long MCO equity with protective Jul-2026 puts (buy MCO stock + Jul-2026 230 puts) to capture upside from stable fee growth while limiting downside from cyclical hits. Timeframe 9–15 months; target 25–35% upside, max draw limited to put cost plus delta exposure.
• Tactical AI supply chain: small tactical long NVDA Jan-2027 calls (out-of-the-money) to capture accelerated cloud GPU demand from AWS/NVDA co-investment cycles. Position size: under 1% portfolio. Risk: GPU inventory rationalization; take profits at 80% gain or cut at 50% loss.
• Event hedge on NYT ad cyclicality: buy NYT Sep-2026 straddle around next two quarterly prints if implied vol < realized 90‑day move historically. Use for 3–6 month tactical exposure to subscriber/advertising re-acceleration vs disappointment; cap max loss to premium paid.