In the classical asset pricing theory,the representative investors are considered as families or individuals.However,in modern economies,a significant share of fi-nancial wealth is delegated to professional institutional investors rather than managed directly by the owners.This separation results in agency problems.Since traditional behavioral asset pricing theory takes a single representative agent to model institutional investors as a group,it largely ignores the competing behaviors that result from fund manager’s relative performance concern.To fill the gap,I propose a new model with heterogeneous belief agents who try to maximize their relative performance under a Nash equilibrium framework.In the model,two agents with heterogeneous beliefs exist and each represents for an infinite number of fund managers having the identical investment behaviors.At the end of the period,each agent will receive a bonus or penalty according to her relative performance to the opponent.In the equilibrium,the presence of heterogeneous beliefs affects the stock price and increases volatility.Heterogeneous beliefs enlarge the difference of agents’ perceptions of market prices of risk,rendering them a counter-cyclical pattern.The optimistic agent will allocate more fraction of wealth into the risky asset than her opponent so as to be consistent with her belief.In order to finance this additional demand,the optimistic agent always employs leverage and borrows money from the opponent.I consider two extreme situations to discuss the effect of relative performance concern.One scenario is where one agent temporarily beats the opponent.The presence of relative performance concern will reduce the impact of the winner and give the loser opportunities to reenter the market.The other scenario emerges when the intensity of agent’s concern on relative performance becomes infinitely strong.On this occasion,both agents would be extremely risk averse and apply similar trading strategies.I study the short-term effects of investors’ heterogeneous beliefs on Chinese stock price using weekly trading data.Based on the portfolio analysis and the Fama-MacBeth cross-sectional regression analysis,I find the heterogeneity of investors’ beliefs will raise the current stock price under the short selling restrictions,and the stock with suf-ficiently high heterogeneous beliefs will present return reversal in the following week.If short selling is allowed,the divergence of investors’opinions still generates modest upward pressure on the current price.But its impact on the next week’s stock return is no longer significant.Using data of quarterly equity holdings,I analyze institutional investors’ herd be-havior and its impact on the Chinese market.A significant herding is found between 2006 and 2015,during which time the intensity of the buy-side herding is stronger than the sell-side.Moreover,I find the herding intensity declines first and then increases with the heterogeneous beliefs increasing.The buy-side herding destabilizes the price of stock with high heterogeneous beliefs,presenting the pattern of stock price up fol-lowed by a down.In the case of stocks with low heterogeneous beliefs,there is no significant reversal after the buy-side herding.Besides,the sell-side herding generates a permanent impact on the stock price without a temporary price adjustment. |