Prediction for Some Tech Stock Prices of U.S. Stock Market based on ARIMA Model

Abstract

Predicting the future price of stocks is a common practice in the financial industry. Individuals and organizations engage in stock price forecasting for investment decision-making, risk management, portfolio optimization and asset allocation. However, the predicting stock prices is challenging due to the complexity of financial markets, the influence of numerous variables, and the presence of random and unpredictable events, which all make it harder to predict. This paper uses predicted future stock price trends for technology companies in The United States based on ARIMA and breaks it down using APPLE as a specific example for a more in-depth discussion of ARIMA. The raw data is collected from 2001/5/11 to 2021/4/22. Then it predicted the stock prices for the next two weeks (working days) by ARIMA and made a fitted trend curve, and finally compared it with the real data from 2021/4/23 to 2021/5/12 to see if ARIMA's predictions were accurate. The result shows that prediction model is in the form of ARIMA (1,1,1), and the prediction of APPL from 2021/4/23 to 2021/5/12 is trending down, which matches the real trend.