Think Again : Volatility Asymmetry and Volatility Persistence

We use quantile regressions to demonstrate that volatility persistence and the asymmetric "leverage" effect are high volatility phenomena. More specifically, we find that (i) low volatility is not persistent, but high volatility all the more, even featuring properties of explosive processes; (ii) both positive and negative shocks increase volatility but negative shocks display a stronger effect; and (iii) jumps do neither drive nor destroy the persistence of volatility. The analysis illustrates that quantile regression can provide information that is hidden in commonly used GARCH or realized volatility models. The quantile regression results also explain the weak empirical evidence of the leverage effect and the volatility feedback effect