Forecasting the direction of daily changes in the India VIX index using deep learning

Abstract

The VIX index is an indicator of the market's perception of risk, and an accurate forecast of the movements in VIX can be very useful for investment risk management. So, the aim of this study is to predict the day-to-day movement of the India VIX using six deep learning architectures. All six architectures performed well and achieved a higher level of accuracy with minor differences than in previous studies. The findings of the study are of great relevance for assessing short-term risk as well as long-term strategies for hedgers, risk-averse investors, volatility traders, investors, and financial researchers.

Keywords

• Volatility
• VIX
• LSTM
• GRU
• RNN
• CNN