The way we measure 'market sentiment' is about as accurate as a broken compass.
April 16, 2026
Original Paper
Disagreement of Disagreement
SSRN · 6566844
The Takeaway
Economists use various 'proxies' to figure out when investors disagree, assuming these tools all point to the same truth. But a new analysis shows these major metrics have almost zero correlation with each other, rendering most current theories unreliable. If one tool says investors are divided, another might say they're in total agreement, meaning decades of financial advice might be based on data that is essentially noise. We’ve been trying to navigate the stock market using a map where the landmarks aren't even real. For regular investors, it means the 'expert' sentiment you hear on the news is often based on fundamentally flawed math.
From the abstract
We show that major investor disagreement proxies have near zero correlation – making disagreement pricing inferences challenging. We develop a frictionless model that generates overpricing from latent disagreement, jointly rationalizes major proxies, and generates new testable predictions. Equilibrium interrelationships motivate a new disagreement measure, DIS, which is more predictive of returns cross-sectionally than major proxies and more consistently predictive across macroeconomic regimes.