The impact of jumps on the stylized facts of returns and volatility

We investigate, using a number of econometric tools, the impact of jumps on the stylised facts of returns and volatility. We look at 100 stocks, representing 10 sectors over the period 2000-2010. We find that jumps had little or no impact on the stylised facts of returns and volatility. The volatility- volume (trades) relation and leverage and feedback effects work only via the continuous component of realised variance. Jumps have no memory and stocks with higher jump activity are less persistent. Aggregational and conditional normality are jump independent such that low or high levels of jump activity do not affect returns convergence towards normality nor the normality of returns conditional on volatility. Forecasting in the presence or absence of jumps is not significantly different.