Why Andy Constan Says The AI Bubble is in Earnings, Not Price

Andy Constant, veteran macro trader and former Salomon Brothers and Bridgewater alum, argues that the current AI-driven market rally isn't a bubble in price—but in earnings expectations. While traditional valuation metrics like P/E ratios aren't extreme, he warns that corporate earnings forecasts—especially for AI and semiconductor stocks—are wildly unsustainable. The S&P 500 is projected to earn $400 billion next year, but only $1.5 trillion in new GDP is expected, meaning AI companies alone are forecast to capture more than half of all corporate earnings growth. This creates a 'K-shaped' economy at the corporate level: a few dominant firms feast on an expanding pie while the rest of the market gets squeezed. Constant draws parallels to past bubbles—not in price spikes, but in the structural imbalance of expectations. He cautions that even if AI delivers on its promise, the economy can't sustain such concentrated gains without a massive GDP explosion. The real risk isn't a sudden crash, but a prolonged, grinding correction as earnings fail to materialize. For investors, the key isn't predicting the top—but managing risk through disciplined exits and hedging strategies like selling deep out-of-the-money calls and buying low-cost puts on concentrated positions.