Efficient Likelihood Estimators for Multivariate Jump-Diffusions

SPRING 2015 RESEARCH INCUBATION AWARDEE 

Gustavo Schwenkler (Finance, Questrom School of Business) 

Financial markets regularly experience large shocks. From the great depression to the recent financial crisis – most financial market crises are characterized by large unexpected losses. These losses may affect corporations and governments in negative ways, reducing economic growth and increasing unemployment. This imposes real costs on society. It is therefore of primary importance to understand the dynamics of large losses in financial markets, and their impact on economic conditions. The statistical methods that François and I have developed empower researchers with tools that can estimate these dynamics from economic data. Our methods assist in deriving insights that can be used to make our financial markets safer, and our economy more stable.

In this project, joint work with François Guay, a PhD student from the Department of Economics, we derive statistical estimators for models of financial time series which can have small fluctuations on a day-to-day basis, but which can also experience large unexpected jumps.