Evaluating the predictive accuracy of volatility models

✍ Scribed by Jose A. Lopez

• Publisher: John Wiley and Sons
• Year: 2001
• Tongue: English
• Weight: 207 KB
• Volume: 20
• Category: Article
• ISSN: 0277-6693
• DOI: 10.1002/1099-131x(200103)20:2<87::aid-for782>3.0.co;2-7

✦ Synopsis

Standard statistical loss functions, such as mean-squared error, are commonly used for evaluating ®nancial volatility forecasts. In this paper, an alternative evaluation framework, based on probability scoring rules that can be more closely tailored to a forecast user's decision problem, is proposed. According to the decision at hand, the user speci®es the economic events to be forecast, the scoring rule with which to evaluate these probability forecasts, and the subsets of the forecasts of particular interest. The volatility forecasts from a model are then transformed into probability forecasts of the relevant events and evaluated using the selected scoring rule and calibration tests. An empirical example using exchange rate data illustrates the framework and con®rms that the choice of loss function directly aects the forecast evaluation results.