Forecasting of Nifty 50 Index Volatility Using Seasonal ARIMA Model

Traders who are at the beginner stage are more likely to make mistakes and suffer losses in the volatile market. The study focuses to develop an autoregressive integrated moving average model for forecasting the Nifty 50 index based on the historical dataset by using the advanced technique of time series analysis. The train-testing approach of machine learning and Akaike Information Criterion (AIC) has been used on the monthly data of eight years for the selection of the autoregressive model. Using the auto-correlation functions (ACF & PACF) and the seasonality component in the dataset, the seasonal ARIMA model of (0,1,1)(0,1,2)
12
has been selected to forecast the Nifty 50 index values. Based on this selected model, values have been predicted and compared with the actual values for validation purposes. The results show that the model could be utilized to forecast the Nifty 50 index. The model can be helpful for a researcher, and the forecasted values will help a trader to protect from severe losses in his beginner stage.