Stock Prices Prediction Based on ARMA Model

Abstract

The financial time series contain some information that indicates the operation law of the system. Researchers can use classic models of time series to study previous stock prices and predict a short-term trend of the volatility of the prices. In this article, we choose the adjusted closing prices of Apple Inc from 2018 to the end of 2019. Then we perform the first difference on the original data to make the sequence stationary to apply the ARMA model to predict the adjusted closing prices of Apple Inc in the next five days. The time series, which we predict, is compared to the actual value. And it turns out that the data's error rates are low, indicating that the ARMA model is suitable for the short-term prediction of the prices and further. Meanwhile, it further proves that the time series model serves as a positive catalyst in the study of finance.