Suitability of Volatility Models for Forecasting Stock Market Returns: A Study on the Indian National Stock Exchange

Abstract

Problem statement: Measuring volatility is an important issue for stock market traders. Also, volatility has been used as a proxy for riskiness associated with the asset. This study aims to compare the different volatility models based on how well they model the volatility of the India NSE. Approach: The study has made use of five models which are Historical/Rolling Window Moving Average Estimator, (ii) Exponentially Weighted Moving Average (EWMA), (iii) GARCH models, (iv) Extreme Value Indicators (EVI) and (v) Volatility Index (VIX). The data includes the daily closing, high, low and open values of the NSE returns from 2005-2008. The model comparison was done on how well the models explained the ex-post volatility. Wald’s constant’s test was used to test which method best suited the requirements. Results: It was concluded that the AGARCH and VIX models proved to be the best methods. At the same time Extreme Value models fail to perform because of the low frequency data being used. Conclusions: As other research suggests these models perform best when they are applied to high frequency data such as the daily or intraday data. EVIs give the best forecasting performance followed by the GARCH and VIX models.