Forecast Evaluation in Large Cross-Sections of Realized Volatility
This work addresses forecast evaluation for financial volatility modeling, which is incremental as it builds on existing HAR models by incorporating cross-sectional and jump components.
The paper tackles the problem of evaluating forecasts for realized volatility measures in large cross-sections with dependence, using equal predictive accuracy tests. It finds that an augmented HAR model with LASSO shrinkage and corrections for measurement error and jumps improves predictive accuracy compared to a standard HAR benchmark, as demonstrated through numerical implementations.
In this paper, we consider the forecast evaluation of realized volatility measures under cross-section dependence using equal predictive accuracy testing procedures. We evaluate the predictive accuracy of the model based on the augmented cross-section when forecasting Realized Volatility. Under the null hypothesis of equal predictive accuracy the benchmark model employed is a standard HAR model while under the alternative of non-equal predictive accuracy the forecast model is an augmented HAR model estimated via the LASSO shrinkage. We study the sensitivity of forecasts to the model specification by incorporating a measurement error correction as well as cross-sectional jump component measures. The out-of-sample forecast evaluation of the models is assessed with numerical implementations.