Implementation of Holt’s Linear Exponential Smoothing in Predicting Stock Prices

Abstract

Historical data of the daily closing price of Bank Tabungan Negara (BBTN) shares for the 2019-2025 period was processed using Holt's Linear Exponential Smoothing method to generate short-term trend projections. The modeling process was carried out by calculating the level component, trend component, and predicted values based on predetermined smoothing parameters α and β. Model performance evaluation was conducted using the Mean Absolute Percentage Error (MAPE), yielding a value of 4.46%, which falls into the excellent category (<10%). This prediction system was implemented as an interactive web application allowing users to upload data, configure calculation parameters, and display prediction results in tables and graphs. The research results show that the used method can provide accurate stock price estimates and can be used to support trend analysis and short-term investment decision-making.