Parameter Estimation and Pairs Trading for Some Lévy-driven Ornstein-Uhlenbeck Processes

Thurs. March 2 | 5:00-6:00 PM | Babbio Center Room 219

We discuss parameter estimation using maximum likelihood and Fourier inversion for Lévy-driven Ornstein-Uhlenbeck processes, where the stationary distribution or background driving Lévy process is a weak variance alpha-gamma distribution, a multivariate generalization of the variance gamma distribution. These processes allow for the modeling of possibly infinite activity mean reverting price processes with jumps, and we then study how to perform pairs trading in this framework in the univariate case. Specifically, we use simulation methods to demonstrate how to find the optimal level of the process to enter and exit trades, with control variate as a variance reduction technique.

Join virtually on Zoom: https://stevens.zoom.us/j/96283727815

Bio: 
Kevin Lu's research is in the area of stochastic processes, probability theory, and its applications to mathematical finance. Focusing on Lévy processes, he has worked on the subordination of Lévy processes to create time-change models of multivariate asset prices, and statistical estimation for multivariate Lévy-driven Ornstein-Uhlenbeck processes.