Composite Behavioral–Fundamental Indexes via PCA for Predicting Indonesia Stock Exchange Excess Returns: An Integrated Anomaly Analysis

This study integrates traditional valuation metrics and behavioral sentiment indicators into a composite framework to predict excess returns on the Indonesia Stock Exchange (IDX).Seven proxies-price-to-book value, price-toearnings, turnover ratio, consumer confidence, global sentiment, Fear & Greed Index, and VIX-were standardized and subjected to Principal Component Analysis (PCA), yielding two orthogonal factors: a fundamental valuation factor and a behavioral sentiment factor.These factors explain 46.9% of total variance and were used as predictors in a multiple regression model for IDX excess return.Regression results show both factors have significant negative impacts on returns (β = -0.382,p < .001),with the model accounting for 29.2% of return variability (F(2,77) = 15.86,p < .001).Findings corroborate value-reversion and sentimentreturn dynamics, demonstrating that overvaluation and extreme sentiment each herald lower subsequent returns.The composite PCA approach mitigates multicollinearity and enhances explanatory power compared to single-proxy models, extending anomaly forecasting techniques to emerging markets.For Indonesian investors, the dual-factor strategy offers a robust tool for timing market entry and exit by combining fundamental screening with sentiment monitoring.