Evaluating the Accuracy of Forecasted Stock Prices of Malaysian Banking Companies Using Geometric Brownian Motion Model with a Continuous Dividend Yield Parameter

Abstract

Accurate stock price forecasting is essential in Malaysia's banking industry for investors, analysts, and policymakers to make wise financial decisions. So, an accurate forecasting model is necessary for banking institutions to predict market behaviour during economic changes, as banks play an important role in the national financial system. To improve the forecast of price modelling and account for the effects of dividend distributions, this study reviews the forecasting accuracy of the geometric Brownian motion (GBM) model with a continuous dividend yield parameter. There are eight Malaysian banks that were selected and considered: Maybank, Public Bank, RHB Bank, CIMB Bank, Hong Leong Bank, AmBank, Affin Bank, and Bank Islam. The history of daily stock prices and dividend data from December 2022 to June 2024 was used to estimate the drift, volatility and dividend parameters for three historical periods of 80, 160, and 245 trading days. We used the mean absolute percentage error (MAPE) method to check the accuracy of forecasts made for trading days that were 5, 20, and 120 days ahead. Longer historical data improved the accuracy of predictions. Maybank, Public Bank, and RHB Bank had the lowest error rates, below 2%, which shows that the market is stable and efficient. These findings reflect the accuracy of the model in forecasting dividend-yielding banking stocks in Malaysia, as well as its ability to accurately capture stock price fluctuations. The study provides helpful information for investment analysis and policy development, improving our understanding of stochastic financial modelling in emerging markets.