Predicting Apple Stock Price Based on ARIMA Model

ARIMA stands for Auto-Regressive Integrated Moving Average and is a popular statistical model to forecast time-series. It has since been used to predict the behaviour of time series based on historical data, and a wide range of applications in which it is useful like stock market forecasting, economic growth analysis, etc. While the ARIMA model has significant benefits, it is a model with limitations, particularly when the underlying data are polluted by noise or present non-linear patterns. Long-term forecasting with ARIMA is less accurate because it is a linear model which does not always accurately reflect the complexity or immediate changes found in most real-world financial data. Our focus will be on Apple stock price data from 2014 to 2024 (working days for simplicity), which we analyze using ARIMA model, and we compare with other analysis models. In this analysis, we would like to point out two things why ARIMA models are not suitable for longer-term market predictions and demonstrate how poorly it fits on nonlinear data.