A Model of Momentum and Market States : Theory and Evidence

I develop a model that connects market states and momentum. The model analyzes asset pricing implications of two well-known psychological biases, overconfidence and self-attribution bias, in a setting of multiple risky assets whose payoffs contain a common factor. Due to self-attribution bias, overconfidence varies asymmetrically between winners and losers, resulting in asymmetric return behaviors between them. The model generates a set of implications regarding the relation between market states and momentum and long-run reversals. I find empirical evidence consistent with these implications