Min-Jik Kim and Jaeho Cho
The asymmetric volatility phenomenon (‘the phenomenon’, henceforth), documented first by Black (1976), refers to the fact that the stock return and its conditional volatility are…
Abstract
The asymmetric volatility phenomenon (‘the phenomenon’, henceforth), documented first by Black (1976), refers to the fact that the stock return and its conditional volatility are negatively correlated. To explain ‘the phenomenon’, this paper presents an asymmetric information model under ambiguity, and provides an empirical test of its result as well. We assume that in the Grossman and Stiglitz (1980), uninformed liquidity traders face ambiguity about the distribution of asset payoffs, and that their attitudes toward ambiguity vary depending on the state of the economy. In model I, their utility functions exhibit ambiguity aversion in the bad state and ambiguity neutrality in the good state. In model II, liquidity traders are still ambiguity-averse in the bad state but ambiguity-seeking in the good state. We find that ‘the phenomenon’ appears in model II when the degree of ambiguity is not large. Furthermore, we show that the possibility of ‘the phenomenon’ is higher as the proportion of liquidity traders increases. To perform an empirical analysis, we measure the degree of ambiguity by the Kolmogorov-Smirnov statistic and show that this measure has a positive relationship with the difference between the volatilities in the good and bad states. In addition, we find that the risk factor constructed by the ambiguity measure has explanatory power about returns on 25 portfolios of the Fama-French type in the Korean market.