Volatility Forecasting and Value at Risk

P. Sadorsky (Canada)

Keywords

Risk Analysis, Forecasting, Value at Risk, Risk Management.

Abstract

Risk management techniques and financial engineering have developed extensively over the past ten years. Value at risk (VaR) is a popular approach to measuring market risk on a daily basis. Despite the widespread use of VaR there has been relatively little analytical work comparing the results from calculating VaR from different empirical models. This paper uses daily trading data for four widely traded financial assets, and one portfolio of these assets, and for each asset calculates VaR measures from five different models. For assets with fat tailed distributions, VaR values calculated from non-parametric models are more precise than VaR measures calculated from parametric models that assume normally distributed asset returns.

Important Links:



Go Back