The discussion centers on whether the current AI-driven demand for semiconductors is a long-term secular growth story or a cyclical boom nearing its peak. Analysts present conflicting views, with some predicting the cycle could last until 2030 while others warn of unsustainable growth and margins.
The unprecedented capital expenditure by the largest cloud providers (hyperscalers) is identified as the primary engine for the semiconductor industry's growth. This spending, fueled by massive capital raises, is driving an infrastructure build-out but also siphoning liquidity from other market areas.
A key concern raised is that investors are mistaking the semiconductor industry's significant pricing power for permanent productivity gains, leading to inflated valuations. Analysts argue that current high margins are not sustainable and that stock valuation multiples have likely peaked.
Multiple analysts advise investors to look beyond the most popular mega-cap semiconductor stocks and diversify. Recommendations include spreading investments across the entire AI value chain (compute, memory, networking, software) and into non-correlated asset classes like real estate and commodities.
Keep pulling the thread on Bloomberg Surveillance.