The Effect of Heteroscedasticity in Apple Stock Price Towards Predicting its Future Value

Abstract

The purpose of this study is to examine the effect of heteroscedasticity on the accuracy of predicting Apple Inc.'s stock price using an ARIMA (Autoregressive Integrated Moving Average) and GARCH (Generalized Autoregressive Conditional Heteroscedasticity) model. Heteroscedasticity is the phenomenon of varying levels of volatility over time. The study uses advanced statistical techniques to find and measure patterns of changing volatility in past Apple stock data. By looking at how these changes affect predictions, the research aims to help us better understand the potential errors introduced to forecasting models. A study of historical Apple Inc. stock price data from January 2008 through June 2022 was conducted. Based on model accuracy measures such as Mean Absolute Error (MAE), Root Mean Squared Error (RMSE) and Mean Absolute Percent Error (MAPE), ARIMA (8,1,8) was found to be the best model in this study. Next, among the statistical methods ARIMA and GARCH, ARIMA is the most effective method based on the model accuracy measures. ARIMA is also easy to use and interpret, making it the preferred choice for many investors while GARCH is more to captures volatility. GARCH models explicitly model the volatility of a time series and designed to model and forecast the conditional variance of a time series, specifically focusing on volatility clustering and time-varying volatility.