Serenity: Don’t lump all AI infrastructure into a bubble; the current situation is deleveraging, not a collapse in demand.

Serenity has released a statement arguing that media’s repeated use of the term “AI bubble” is inappropriate. For infrastructure assets, storage firms from Samsung to ESMT carry forward price-to-earnings ratios of just 1 to 3 times, while HBM manufacturers have signed multi-year price floor agreements with hyperscale cloud providers. Optical companies such as POET even trade at a market capitalization roughly $97 million below their net cash, equivalent to a negative valuation for their operating businesses. Meanwhile, AI-related companies still commanding premiums continue to lock in demand for the next 3 to 5 years—for example, a roughly $20 billion long-term deal with Coherent and a $5 billion agreement with Lumentum. Serenity notes it is difficult to directly categorize current high-volatility optical stocks with negative enterprise value, as well as low-valued storage stocks that have locked in multi-year demand and price floors while remaining highly profitable today, as a bubble. There is clear divergence within the sector: private valuations of some startups like Instinct are bubbly, over-leveraged Neocloud platforms with improper financing, and the valuations and promises of some AI labs require careful scrutiny. Markets often use “leverage” or “debt” issues to intimidate investors. Serenity believes that although valuations may compress temporarily due to macroeconomic conditions, regulatory uncertainty, or capex cuts, for companies truly supporting AI infrastructure development, this is more like a “great deleveraging” in 2026 rather than a collapse in fundamental demand. With subsequent earnings releases and free cash flow accumulation, Serenity expects valuation recovery in bottleneck segments, and predicts the market will eventually distinguish between sustainable growth and genuine bubble valuations.