Stochastic Volatility: Univariate and Multivariate Extensions
Eric Jacquier, Cornell University
Nicholas G. Polson, University of Chicago
Peter Rossi, University of Chicago

This projects applies Jacquier, Polson and Rossi’s innovative stochastic volatility estimation methods to models of fat-tailed and skewed conditional distributions that are appropriate for modeling leverage effects, stochastic factor structures and stochastic discounts. The results show strong evidence for non-normal conditional returns in stock returns and exchange rates. These characteristics suggest strong implications for option pricing models and other models that depend on stochastic volatility. (Accepted Spring 1996.)