Mercredi | 2014-02-12
B103

Denisa BANULESCU-RADU – Zhuo HUANG – Marius² MATEI – Peter REINHARD HANSEN

We study the financial volatility during the global fi nancial crisis and use the largest volatility shocks to identify the major events during the crisis. First, we propose an improved version of the Realized GARCH which accounts for the asymmetry and the impact of outliers on volatility. This model takes advantage of the realized measures of volatility, in the modeling of daily returns. The time series of daily realized volatility measures that are computed from high frequency data spaning the period from 1997 to 2009 that includes several international crises and major events such as 9/11. This adds perspective to the magnitude of the global financial crisis. We use this timeseries to zoom in on the events during the 2007-2009 period, and the model produce a daily series of volatility shocks. Interestingly the largest volatility shock is found to coincide with a computer glitch in the trading system.