<![CDATA[This study aims to analyse the performance of a stock price forecasting model based on Geometric Brownian Motion (GBM) modified with the Kalman Filter (KF) approach. The GBM model is used to represent the basic behaviour of stock price movements, which are stochastic in nature, while the Kalman Filter plays a role in estimating model parameters based on actual observation data. This study uses closing price data for PT. Aneka Tambang Tbk. (ANTM) shares. The results of this study show the accuracy level using Mean Absolute Error Percentage (MAPE) in the GBM-KF hybrid model and are in the best fitting condition. For ANTM shares, it is 5.72% (GBM) and 3.83% (Hybrid GBM-KF). The GBM-KF hybrid model has proven to be effective in minimising prediction errors and capturing changes in market trends and volatility that cannot be explained by the classic GBM model. Furthermore, this study highlights that integrating Kalman Filter into GBM improves the model’s adaptability to dynamic market conditions, allowing for real-time parameter estimation and enhanced predictive stability. The findings suggest that the GBM-KF framework can serve as a robust tool for financial forecasting, particularly in volatile markets where traditional models tend to underperform.]]>
Copyrights © 2025
