Stock Price Bubble Dissection Requires a look at Fundamentals and Speculation

Stock-market valuations contain a substantial speculative component that can affect both the economy and monetary policy, but identifying a bubble requires separating expectations about fundamentals from non-fundamental price increases.

• The study combines a present-value model with survey expectations about future dividends and stock prices, decomposing valuations into expected dividends, expected returns, and a rational bubble.
• Using monthly U.S. data from January 1971 to December 2024, the estimated bubble explains 61% of the variation in the S&P 500 price-dividend ratio, while expected returns explain 42%; expected dividend growth contributes little over the full sample, and the components are not additive.
• The model estimates bubble components of about 10% during the early-1970s Nifty Fifty boom, 27% in January 2001, 19% in November 2021, and nearly 25% in December 2024, potentially reflecting enthusiasm about artificial intelligence.
• Removing the estimated bubble reduces excess valuation volatility and eliminates explosive dynamics in the unadjusted price-dividend ratio, although the December 2024 estimate is not proof that AI stocks were overvalued or that a correction was inevitable.
• In a structural vector autoregression, positive bubble shocks are followed by lower unemployment, higher industrial production, more working hours, and a policy-rate increase peaking at about 48 basis points eight months after a one-standard-deviation shock; inflation rises only modestly and is generally not statistically significant.

Takeaway: Investors should distinguish high valuations from speculative bubbles, avoid treating the estimate as a market-timing signal, and use bubble exposure in scenario analysis and portfolio-risk assessment because bubbles can persist and influence employment, production, credit conditions, and monetary policy.