Search for content and authors
 

Influence of robust estimation on volatility forecast

Ewa Ratuszny 

Warsaw School of Economics, Al. Niepodległości 162, Warsaw 02-554, Poland

Abstract

The research concerns modelling of several financial time series. Due to presence of different type of outliers in the data the robust estimation of ARMA-GARCH-type models is required. A Monte Carlo simulation is used to measure sensitivity of volatility forecast on different fraction of outliers in the time series. For comparison purposes the QML estimation results were uses as a benchmark. The analysis is based on several type of instruments  from FX, capital, and commodity market. To illustrate the behavior of robust estimate, the VaR forecast is simulated, and for comparison purposes we use tests based on the decision-making aspect, i.e. Binary Loss, Regulatory Loss, Firm Loss and the Expected Shortfall.

 

Legal notice
  • Legal notice:
 

Related papers

Presentation: Oral at Current Economic and Social Topics CEST2013, Symposium on Financial Market Analysis, by Ewa Ratuszny
See On-line Journal of Current Economic and Social Topics CEST2013

Submitted: 2013-04-25 09:24
Revised:   2013-04-25 09:25