So are we in an AI bubble? Here are clues to look for.

What It Covers

Planet Money examines whether the AI boom represents an economic bubble by exploring Harvard research on bubble detection indicators and analyzing current market conditions, including NVIDIA's valuation and the Magnificent Seven tech stocks.

Key Questions Answered

• •Bubble Detection Framework: Harvard researchers identified four predictive clues for bubbles: high price-to-earnings ratios, increased stock volatility, new stock issuance, and accelerating price growth. These indicators predict bubbles correctly about 60% of the time, only slightly better than chance.
• •Current AI Assessment: AI stocks show mixed bubble signals. NVIDIA's price-to-earnings ratio sits at 40 versus the S&P 500 average of 20, and volatility has increased. However, major AI companies aren't issuing new shares, and stock prices lack the exponential acceleration pattern typical of bubbles.
• •Lean Versus Clean Policy: Policymakers debate whether to actively deflate suspected bubbles or wait to address aftermath. The critical factor determining economic damage is borrowing levels. Housing bubble devastation stemmed from bank-backed mortgages, while AI companies primarily use private credit, potentially limiting systemic financial risk.
• •Productive Bubble Theory: Some economists argue bubbles can benefit society when they fund underprovided public goods like research and development. The dot-com bubble created excess fiber optic infrastructure that later enabled the broadband era, suggesting wasted investment sometimes generates unexpected long-term value for the economy.