CMStatistics 2016: Start Registration
View Submission - CFE
Title: Estimation of the discontinuous leverage effect: Evidence from the NASDAQ order book Authors:  Lars Winkelmann - Freie Universitaet Berlin (Germany) [presenting]
Abstract: A new nonparametric estimation approach is provided for studying the correlation of price and volatility jump sizes in high-frequency data. A spectral estimator is proposed which allows to asymptotically remove market microstructure noise from observable prices and to locate and estimate the size of price and volatility jumps in the efficient price process. The leverage effect is estimated by the scaled discontinuous part of the quadratic covariation. Based on five years of NASDAQ transaction data of about 300 firms, we are able to confirm evidence about an elusive relation of price and volatility jump sizes in terms of sign, magnitude and significance. We show that this inconsistency is due to different sources of jumps and a time-varying correlation. The filtering of market-wide events and important macroeconomic news releases gives uniform and highly significant firm specific correlations. Idiosyncratic price volatility jumps are identified as one disrupting component of the discontinuous leverage effect.