Forecasting NEPSE Index: An ARIMA and GARCH Approach

Abstract

In this study, an attempt has been made to demonstrate the usefulness of univariate time series analysis as both an analytical and forecasting tool for Nepali stock Market. The data set covers the daily closing value of NEPSE index for two and half years starting from the middle of 2012 to end 2015. The forecasting analysis indicates the usefulness of the developed model in explaining the variations, trend and fluctuations in the values of the price index of Nepali stock exchange. Explanation of the fit of the model is described using the Correlogram, Unit Root tests and ARCH tests, which finally confirm that the ARIMA and EGARCH are good in forecasting and predicting daily stock index of Nepal. Furthermore, it is inferred that the daily stock price index contains an autoregressive, seasonal and moving average components; hence, one can predict stock returns through the identified models.