A0693
Title: Forecasting risk measures in high-dimensional portfolios
Authors: Carlos Trucios - University of Campinas (Brazil) [presenting]
Abstract: Forecasting risk measures accurately is of crucial interest from the regulatory policy point of view and for investors as well. However, estimation and forecasting in large panels is very difficult due to the curse of dimensionality. A new procedure is proposed to deal with risk measures estimation in portfolios with many assets. The new procedure is applied to US stocks, and the results suggest the proposal is quite competitive.
