ARIMA Model Application in Predicting NVIDIAs Stock Price

The volatility of stock prices in the technology sector has drawn significant attention from both investors and researchers, particularly in the rapidly evolving GPU market. This study aims to address the challenges of accurately forecasting stock prices by applying the ARIMA model to NVIDIA's stock, a prominent player in the GPU industry. Using monthly data from July 2014 to July 2024 obtained from Yahoo Finance, the data is preprocessed through first and second-order differencing to ensure stationarity. Several ARIMA models are constructed, with ARIMA(2,2,1), the time series value is determined by the combination of the first two autoregressive terms (after two differencing) and the previous error term, and ARIMA(0,2,1), the model depends only on the previous error term, demonstrating the best fit based on AIC and BIC criteria. The predictive accuracy of these models is compared by calculating key statistical indicators, such as RMSE, MAE, MAPE, and Symmetric MAPE. The findings suggest that while both models offer valuable insights, ARIMA(021) provides more precise forecasts. However, the limitations of ARIMA models in capturing non-linear and sudden market movements are acknowledged, emphasizing the need for incorporating additional real-world factors in future analyses. This study provides practical recommendations for investors and researchers to understanding and predicting stock market trends.