This Is How You Know When the Stock Market Is in a Bubble

Episode Summary

Executive Summary: The episode examines what makes a true stock-market bubble, arguing that the term is overused and should be reserved for a rapid price run-up followed by a crash. Harvard's Robin Greenwood explains that bubbles are hard to call in real time, but research finds recurring features—especially acceleration in prices, heavy issuance, and exposure to new firms/industries—that help distinguish bubbles from ordinary bull markets.

Main Topics: Why the word “bubble” is overused (Priority: 5/5): Joe and Tracy open by criticizing how often journalists and investors label markets as bubbles, noting the term is usually deployed too casually and retrospectively. Defining a bubble rigorously (Priority: 5/5): Robin Greenwood explains that his research uses a narrow definition: a very rapid price run-up followed by a crash, rather than broader notions of mispricing. Empirical bubble patterns across history (Priority: 5/5): Greenwood describes findings from U.S. industry data and cross-country data showing that only about half of extreme run-ups end in crashes. Predictors of crash-prone booms (Priority: 5/5): The discussion centers on three recurring characteristics—price acceleration, issuance, and novelty/new firms—that help forecast which booms are more likely to end badly. Career risk and timing difficulty (Priority: 4/5): The guests stress that correctly identifying a bubble can still be financially and professionally damaging because prices often keep rising for months after a bubble signal. Behavioral finance and human nature (Priority: 4/5): The episode links bubbles to FOMO, belief formation, and generational differences in investing behavior, especially during the dot-com era. Bubbles beyond equities and future research (Priority: 3/5): Greenwood notes similar issuance signals in credit markets and says commodities and press-language analysis are possible future research areas.

Key Arguments: Calling something a bubble before the crash is much harder and more meaningful than labeling it after a collapse. A narrow definition—rapid run-up followed by crash—helps make bubbles empirically studyable. In historical samples, only about half of extreme price run-ups actually crash, so many are just prolonged bull runs. Even when a bubble is correctly identified, prices may continue rising for months, making shorting or avoiding the market costly. Acceleration in recent price gains is a useful warning sign because stronger near-term momentum often precedes crashes. Issuance matters because firms exploit buoyant markets to sell new equity, and heavy issuance combined with big run-ups predicts poor future returns. New industries and new firms are central to many bubbles because investors chase new stories rather than established businesses. The dot-com bubble showed strong generational and behavioral effects: younger managers participated earlier, while older managers were slower to buy in. Bubble dynamics are likely permanent because human nature and new narratives keep recurring, even if the specific form changes over time.

Data Points: Stock Movers report length: five minutes or less - Bloomberg promo describing short audio reports Bloomberg global reporting staff: 3,000 journalists and analysts - Promo highlighting the reporting behind Stock Movers Price-run-up threshold studied: 100% price run-ups - Greenwood describes the historical episodes his team identified U.S. bubble episodes found: 40 - Episodes since the 1920s in U.S. industry data Share of U.S. episodes that crashed: roughly half - About half of the 40 U.S. episodes ended in a crash International sample size: 34 countries - Cross-country study back to the 1980s International episodes identified: 107 - Episodes found across 34 countries Share of international episodes that crashed: roughly half - About half of the 107 international episodes ended in a crash Average timing error: 5 months - Even correctly identified bubbles kept rising after the signal Average additional gain after signal: 30% - Prices rose further after the bubble warning in historical episodes Tech bubble shorting example: March 1999 to April 2000 - Illustrates how early a bubble signal could have been during the dot-com era 1920s electrification household penetration: 35% to 70% - Used as an example of a new-industry bubble in the 1920s Dot-com naming premium: 70% - Adding “.com” to a company name reportedly boosted stock price by 70% Mutual fund manager age cutoff: over 45 - Older managers were slower to enter dot-com stocks in Greenwood’s prior research

Pivotal Quotes: "there's really no such thing as a bubble" — Robin Greenwood: Greenwood describing the skepticism many economists have toward the term "a very rapid price run up followed by a crash" — Robin Greenwood: His paper’s narrow operational definition of a bubble "a bubble is an asset that I get fired for not owning" — Joe Weisenthal: Discussion of career incentives that push investors to join speculative manias

Implications: Listeners should treat bubble calls cautiously: the best signals are empirical, not rhetorical. For investors, the danger is missing a long run-up or getting in too late. For markets, bubbles likely remain inevitable because new stories and human behavior repeat.