A False Discovery Rate Approach to Optimal Volatility Forecasting Model Selection

We propose a novel empirical strategy that enables the researcher to select buckets of volatility forecasting models with similar accuracy. Using our Multiple Hypothesis Testing framework with False Discovery Rate (FDR) approach, we are able to identify buckets of more accurate models relative to the literature’s benchmark volatility forecasting models of GARCH, GJR-GARCH, and HAR. We compare the FDR buckets with those of two Family-Wise Error Rate (FWER) testing frameworks and the evidence supports our proposed FDR approach to selecting optimal volatility forecasting models. We further investigate the role of financial stress on our buckets and comparative model selection frameworks