Stock Price Forecasting of PT. Bank Rakyat Indonesia (Persero) Tbk. Using Long Short-Term Memory (LSTM) Method

Abstract

Stock price forecasting is a major challenge in financial market analysis due to the volatility and unpredictability of price movements. The limitations of traditional statistical methods in capturing nonlinear patterns and long-term temporal dependencies have encouraged the adoption of deep learning–based approaches. This research aims to predict the stock price of PT Bank Rakyat Indonesia (Persero) Tbk. (BBRI) using the Long Short-Term Memory (LSTM) method, which is effective at handling problems with fading information and identifying long-term trends in time series data. The dataset comprises historical BBRI share prices from April 16, 2015, to April 16, 2025, with 80% of the data used for training and 20% for testing. LSTM’s model was trained for 10 epochs with a batch size of 32 using the Adam optimizer. The results prove that the LSTM model can effectively capture stock price movement patterns, achieving a mean absolute error (MAE) of 8.42 and a mean absolute percentage error (MAPE) of 1.50%, indicating a high level of accuracy. The visualization of the prediction results reveals a trend that closely aligns with the actual values. These findings reinforce LSTM’s position as a reliable approach to stock price forecasting and highlight its potential as a strategic tool for investors and policymakers in managing market risk.