Predicting Financial Risks

Jour Fixe talk by Roxana Halbleib on July 26, 2017

Roxana Halbleib is a Research Fellow of the Zukunftskolleg affiliated with the Department of Economics. 

During and following the financial crisis from 2007/2008 numerous financial institutions experienced extreme losses or even went bankrupt (e.g., Lehman Brothers): between 2007 and 2010, the losses of US and European banks amounted to 2.6 trillion USD, which is as large as the Gross Domestic Product (GDP) of France in 2012. These financial losses have revealed serious pitfalls in the existing financial risk measures. Although the theoretical research and empirical work on these measures have been continuously developing, their performance during the periods when they are needed most, such as during financial crises, has proven to be quite poor. Roxana’s presentation focused on which are the pitfalls of these measures and how one can improve them by developing new methodologies for measuring and forecasting risks by exploiting the rich information content of high-frequency financial data sampled in the classical calendar time, but also in an alternative time dimension that accounts best for the market’s activity intensity and riskiness. The research for this project has a particularly practical relevance in the nowadays context of dramatic increase in the trading activity on financial markets and in the context of the occurrence of huge losses by financial institutions during and after turbulent times, by which the whole economy, but also each individual, as a tax payer, are strongly affected.