After a brutal July selloff that erased trillions from global tech valuations, U.S. and Asian equity markets have staged a decisive rebound in early August, with AI stocks reclaiming bullish momentum. The turning point came as three major cloud providers reported quarterly earnings and capital expenditure plans, dispelling the market’s worst fears about an AI spending slowdown. What was widely dismissed as an overvalued bubble just weeks ago now looks like a valuation overshoot to the downside — setting the stage for a two-tier recovery: AI hardware names reclaiming fair value, while AI software and application stocks push sustainably higher.
Global Markets Snap Back as AI Sentiment Reverses
The recovery has been broad-based across major tech hubs. On Wall Street, the Nasdaq Composite has rallied roughly 9% from its July lows, led by mega-cap cloud and semiconductor names. The Philadelphia Semiconductor Index rebounded 12% in five trading sessions, clawing back roughly half of its July losses. In Asia, the move has been even more dramatic. South Korea’s KOSPI index surged nearly 5% in a single session on August 5, triggering exchange circuit breakers as SK Hynix and Samsung Electronics led chip stocks higher. Japan’s Nikkei 225 climbed 3.7%, with SoftBank and chip equipment maker Advantest posting double-digit gains. China’s STAR Market followed suit, with the STAR 50 Index jumping 4.8% as domestic semiconductor and AI hardware names swept higher. This synchronized global rebound is more than a technical bounce. It reflects a fundamental repricing of AI demand expectations, driven by hard numbers from the hyperscalers themselves.
Cloud Capex Results: The AI Compute Race Is Still Accelerating
Heading into earnings season, the dominant market narrative was that AI capital spending had peaked and would soon decelerate, crushing hardware revenue forecasts. That thesis has been largely dismantled by the Q2 results from Amazon, Microsoft, and Alphabet. Amazon raised its full-year 2026 capex guidance from $200 billion to $220 billion, citing relentless demand for AI infrastructure from enterprise customers. AWS revenue grew 37% year-over-year, the fastest pace in 18 quarters, with AI-related annualized revenue surpassing $25 billion. Alphabet’s Google Cloud revenue exploded 82% higher, expanding operating margins to 35.6%, while Microsoft’s Azure grew 43% as Copilot adoption accelerated across enterprise accounts. Meta also lifted its capex floor from $125 billion to $130 billion. Combined, the four largest U.S. hyperscalers now plan to spend roughly $730 billion in 2026, up nearly 79% year-over-year. What changed the market’s mood is not just the sheer size of spending, but the visibility of returns. Cloud backlogs now exceed $2.3 trillion globally, and AI workloads are translating directly into revenue growth and margin expansion rather than just costs. For the first time, investors can see a clear path from infrastructure investment to profitable monetization — which changes the entire valuation framework for the AI supply chain.
The Path Forward: Hardware Valuation Reset, Software Sustained Growth
Against this backdrop, the market is likely entering a two-speed AI recovery. First, AI hardware stocks — semiconductor equipment, memory chips, optical modules and server makers — are poised for a valuation mean-reversion trade. The July selloff pushed many of these names down 30–40% on fears of a capex cliff that now appears unlikely. With cloud spending plans reaffirmed and even upgraded, much of that pessimism looks overdone. This does not mean a return to the 2025-style parabolic beta rally, but rather a gradual repair of excessively depressed multiples as earnings visibility improves. Second, AI software and application names are positioned for more sustained upward momentum. The market is increasingly differentiating between companies that sell shovels and those that sell finished productivity tools. Microsoft’s Copilot milestone — surpassing 30 million seats with 75% sequential growth in enterprise deployments — proved that enterprise AI software can scale rapidly and generate predictable recurring revenue. As the market shifts its focus from infrastructure buildout to monetization and ROI, software companies with proven adoption, pricing power and clear use cases deserve premium valuations that may continue to expand.
Risks to Monitor
This recovery thesis is not without risks. Yen intervention dynamics remain a wild card for global liquidity, and any re-escalation of U.S.–China trade tensions could hit semiconductor supply chains. Valuation rebounds will also be uneven — companies without genuine earnings leverage will likely lag behind. Still, the cloud capex numbers have reset the baseline: the AI compute cycle is far from over, and the July selloff looks increasingly like a valuation overshoot rather than a fundamental top.