Signal or Noise? Uncertainty and Learning about Whether Other Traders are Informed

Abstract

We develop a model where some investors are uncertain whether others are trading on informative signals or noise. Uncertainty about others leads to a non-linear price that reacts asymmetrically to news. We incorporate this uncertainty into a dynamic setting where traders gradually learn about others and show that it generates empirically relevant return dynamics: expected returns are stochastic but predictable, and volatility exhibits clustering and the “leverage” effect. The model nests both the rational expectations (RE) and differences of opinions (DO) approaches and highlights a link between disagreement about fundamentals and uncertainty about other traders.

Keywords: Learning, Asymmetric Information, Rational Expectations, Noise Trading, Sentiment, Difference of opinions, Volatility clustering, Leverage effect

JEL Classification: G12, G14