Analysis of Google Stock Prices from 2020 to 2023 using the GARCH Method

Abstract

This research focuses on Google's share price movements, considering their significant impact on the financial market, using Google's share price data from 2020 to 2023. The aim is to analyze error variance and forecast and provide valuable information to stockbrokers and investors. The ARMA model has shortcomings in dealing with volatility, so the GARCH model is used to overcome it. Research methods include financial data analysis, preprocessing, and modeling with GARCH. The rolling forecast method describes changes in price patterns over time. Evaluation using MAPE validates the prediction accuracy of the ARIMA model. The best model chosen with the most negligible AIC value criteria was the ARIMA(3,0,2)GARCH(1,1) model. The forecasting results show accurate stock price predictions with an average MAPE value of 20.7%. This research provides an essential basis for brokers and investors in making investment decisions based on a deep understanding of the dynamics of Google's share price movements in the above time frame.

Keywords

• garch
• stocks
• stock prices
• time series