Volatility measures and the variance risk premium

• Applicant: Dr. Karin Stürmer
• Subject Area: Statistics and Econometrics
• Term: from 2016 to 2018
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 317199074

The 2007/08 financial crisis and the subsequent Great Recession have revealed that the interplay between financial market risks and the broad macroeconomy have not yet been fully understood. In particular, the focus on short-term horizons in determining risk measures turned out to be a serious flaw in existing risk management. Motivated by these observations, this research project deals with new concepts on modeling financial market volatility and the variance risk premium (VRP), which explicitly incorporate the macroeconomic environment.In order to obtain improved volatility forecasts, the first part of the project focuses on the persistence of the conditional variance. First, a new GARCH volatility model with time-varying persistence that is determined by the dynamics of an exogeneous variable will be developed. By using MIDAS techniques, high-frequency financial data will be directly combined with low-frequency macroeconomic data. It can be expected that a flexible persistence will lead to improved volatility forecasts (compared to standard models with constant persistence). By means of extensive simulations, the entire distribution of returns will be predicted over the short-, medium- and long-term horizon. Then, an in-depth evaluation of the forecasting ability of the new model will be carried out in a Value-at-Risk exercise. Furthermore, the direct inclusion of macroeconomic variables will allow to run a scenario-based systemic risk analysis. In summary, the first study will enhance our understanding of financial risks over the relevant time horizon. This is crucial both for the risk management of individual financial institutions as well as from a regulatory point of view, since an improved quantification of systemic risk measures will be provided.The second part of the project will deal with the VRP, i.e. the difference between the implied volatility and the statistical expectation of volatility. The VRP can be interpreted as the premium that market participants are willing to pay in order to ensure themselves against extreme changes in volatility. The study aims at characterizing the macroeconomic determinants of the VRP. While the VRP is usually considered only over a one-month horizon, the study will focus on estimating the entire term structure of the VRP and its reaction to macroeconomic announcements. On the one hand, enhancing our understanding of the VRP is of particular interest from the perspective of financial economics due to its role in predicting market returns. On the other hand, the VRP may be interpreted as a measure of how much risks financial market participants are willing to take and is therefore also highly relevant from a regulatory point of view.

• DFG Programme: Research Fellowships
• International Connection: USA
• Host: Professor Torben G. Andersen, Ph.D.