Do Artificial Neural Networks Provide Improved Volatility Forecasts : Evidence from Asian Markets

Forecasts of stock market volatility is an important input for market participants in measuring and managing investment risks. Thus, understanding the most appropriate methods to generate accurate is key. This paper examines the ability of Machine Learning methods, and specifically Artificial Neural Network (ANN) models to forecast volatility. The ANN models are compared against traditional econometric models for ten Asian markets across 24 years of daily data. The empirical results for ANN models are promising. Out-of-sample forecast evaluation reveals that ANN models are superior for each index compared to benchmark GARCH and EGARCH models. In addition to standard statistics forecast metrics, we consider risk management measures including the value-at-risk (VaR) average failure rate, the Kupiec LR test, the Christoffersen independence test, the expected shortfall (ES) and the dynamic quantile test. The findings again provide general support for the ANN and suggest that this may be a fruitfull approach for risk management