<![CDATA[Ideal stock price prediction model should capture the random and fluctuating nature of stock price movements to produce more accurate forecasts. Since stock prices are highly dynamic and subject to unexpected changes, predicting future prices remains challenging due to market uncertainty and volatility. Nevertheless, many previous studies have relied on deterministic approaches, such as linear regression and ARIMA, which often fail to adequately represent stochastic market behavior. Therefore, this study aims to develop a mathematical model for predicting future stock prices using the Geometric Brownian Motion (GBM) model, particularly to support investors in selecting companies within the insurance sector. This research employs descriptive and predictive quantitative approaches. The descriptive approach examines the historical characteristics of stock price data, while the predictive approach applies the GBM model to represent asset price movements as a stochastic process influenced by return and volatility parameters. The study focuses on PT Asuransi Multi Artha Guna Tbk (AMAG) during the 2024–2026 period, using approximately 500 historical data observations collected from Investing.com. The findings indicate that the GBM model achieved a Mean Absolute Percentage Error (MAPE) of 8.09%, which is below the 10% threshold and demonstrates a very high level of predictive accuracy. These results suggest that the GBM model effectively captures the price dynamics of AMAG stock and can serve as a reliable forecasting tool for investment analysis. Future studies are recommended to incorporate external economic factors or compare GBM with other stochastic models to further improve prediction accuracy and reduce forecasting errors.]]>
Copyrights © 2026
