<![CDATA[Investors increasingly require systematic methods for sustainable stock screening, particularly for ESG-focused benchmarks like Inodnesia's SRI-KEHATI Index. This Study addresses a gap by developing and evaluating a stock screening framework using a Gaussian Naive Bayes (GNB) classifier to integrate both fundamental and technical analysis. The model utilized quarterly data from 25 SRI-KEHATI stocks from Q1 2024 to Q2 2025, training on 11 indicators to predict future quarterly returns, classifying stocks as "Investable" (Label 1) or "Non-investable" (Label 0). The model achieved an average training accuracy of 73.6%. Feature importance analysis revealed that technical indicators, such as Average Log Return, Average MACD, Average RSI, and key fundamental ratios, PBV and ROA, were the most influential predictors. Model predictions were evaluated through a simple equal-weighted portfolio simulation for Q3 2025. The simulation results showed the model-selected "Investable" portfolio generated a 29.9% return, substantially outperforming and the "Non-investable" portfolio (3.56%). These findings demonstrate that the GNB classifier is an effective framework for sustainable stock screening, successfully identifying ESG-compliant stocks that also deliver superior financial returns and providing a practical tool for responsible investing in the Indonesian capital market.]]>
Copyrights © 2026
