Scholars in management and economics have shown increasing interest in isolating the behavioural dimension of market evolution. Indeed, by improving forecast accuracy and precision, this exercise would certainly help firms to anticipate economic fluctuations, thus leading to more profitable business and investment strategies. Yet, how to extract the behavioural component from real market data remains an open question. By using monthly data on the returns of the constituents of the S&P 500 index, we propose a Bayesian methodology to measure the extent to which market data conform to what is predicted by prospect theory (the behavioural perspective), relative to the (standard) subjective expected utility theory baseline. We document a significant behavioural component that reaches its peaks during recession periods and is correlated to measures of financial volatility, market sentiment and financial stress with expected sign. Moreover, the behavioural component decreases around macroeconomic corporate earnings news, while it reacts positively to the number of surprising announcements.
British Journal of Management, Vol. 30, Issue 3, pp. 712-729, 2019