New insights in idiosyncratic risk

Résumé

This thesis comprises three essays on the idiosyncratic risk anomaly. The first essay argues that the anomaly is not pervasive over investor time horizon and that it is only observed for short term investors. The empirical results suggest that features changing with the investment horizon such as risk aversion are relevant to explain the anomaly. The second essay links the anomaly to managerial decisions related to investment. It shows that including controls for investment and profitability in the cross-section of stock returns is sufficient to account for the anomaly. The empirical results suggest that the idiosyncratic volatility anomaly reflects the negative impact that corporate investment has over expected returns and that is not captured correctly by the Fama and French (1993) model. Also, that that this effect arises in part from investor mispricing and in part from risk exposure. The third essay highlights the relevance of economic regimes on the anomaly and, shows that the flight to liquidity evidenced by Acharya et al., (2012) might be the reason why the anomaly is not observed after recession periods. The main contributions of this thesis can be summarized as follows: the contributions in the second chapter are threefold. On the one hand, it proposes the co-existence of heterogeneous market players as the source of the idiosyncratic volatility anomaly. On the other hand, the paper proposes a methodology resulting in the estimation of one particular idiosyncratic risk measure for different group of investors defined according to their investment time horizon. Finally, it highlights the necessity finance discipline has of considering more complex mathematical methodologies that offer more realistic approximations to the complexity of financial markets. The major limitation of the paper is that no time horizon shorter than 2 days can be addressed given the daily character of the data used in the analysis. The third chapter offers an innovative approach based on the idea that the anomaly should be linked to managerial decision making and not necessarily iv to investors. The empirical results show that the anomaly is fully accounted for when both investment and profitability controls are considered in the cross-section of stock returns. The results cast doubt on the generalized idea that the anomaly is related investor mispricing. They suggest that the anomaly is also constituted by a component of risk. The major limitation of the analysis is its inability to disentangle how much influence each component has in the anomaly. The third essay proves that the idiosyncratic volatility anomaly is conditional to the state of the economy and is not observed after recessions. The study stresses that during recessions investors move away from high firm specific risk stocks to cover their liquidity needs. It also shows that the effect of this flight to liquidity is larger than the one of the idiosyncratic volatility. Its main limitation is that the feature treated is not general enough to explain the anomaly across all economic regimes