AI fundamentals have become detached from their stock prices: Michael Green

Simplify Asset Management strategist Michael Green called it “totally absurd” for investors to assign a high earnings multiple to Google when earnings were boosted by marking up its position in Anthropic. He also warned that reflexivity dynamics are increasingly evident, and linked the popularity of levered vehicles to greater market volatility—potentially pressuring countries such as South Korea.

Analysis

The market risk here is not that AI optionality exists; it is that investors are paying operating-company multiples for earnings that may be partially inflated by a non-operating mark. That creates a fragile setup: if the next few prints show slower core revenue acceleration or any reversion in private-mark valuation marks, the stock can re-rate lower on both EPS quality and narrative compression, a bad mix for a name this large in the index.

The second-order issue is reflexivity. When a mega-cap becomes a consensus AI winner, passive and levered vehicles can turn that story into self-reinforcing flows; the same mechanism works in reverse once momentum stalls. That matters for QQQ/Nasdaq volatility and for global risk books with high retail leverage, including markets like South Korea where local leveraged products tend to magnify U.S. tech swings rather than absorb them.

Time horizon matters: the immediate catalyst is headline/earnings reaction, but the real vulnerability is over the next 1-3 months as investors distinguish cash earnings from portfolio marks and test whether AI monetization is actually lifting the core business. Over 6-18 months, the thesis is only falsified if cloud/search growth and monetization per user accelerate enough that the valuation can be justified on recurring operating power, not asset marks. Until then, the move looks more like a sentiment-driven multiple risk than a fundamental AI earnings inflection.

Key Decisions for Investors

• Initiate a small relative-value short: short GOOGL / long MSFT on any strength, targeting 5-8% relative underperformance over 1-3 months; cover if GOOGL shows two consecutive quarters of accelerating core revenue and gross margin expansion.
• Buy 1-2 month GOOGL put spreads into the next earnings window to express a quality-of-earnings view with defined risk; structure for roughly 2:1 to 3:1 upside if the market de-rates the stock on weaker recurring EPS.
• Use QQQ put spreads as a portfolio hedge against levered-vehicle-induced volatility if mega-cap tech breadth deteriorates; this is a tactical tail hedge, not a high-conviction directional short.
• Set an alert on any disclosure that isolates the contribution of non-operating marks to EPS; if that contribution remains material, treat rallies in GOOGL as fadeable rather than a sustainable multiple expansion signal.
• If QQQ and TQQQ volumes spike alongside widening intraday ranges, assume reflexive deleveraging risk and reduce gross exposure to crowded AI baskets rather than adding on dips.