Penyusunan Portofolio Saham Lindung Nilai Berdasarkan Principal Component Analysis Pada Indeks LQ45 Indonesi

Abstract

Stock investment carries significant risks, making hedging strategies essential. Research on hedging in the Indonesian stock index is limited. This study aims to construct a hedging portfolio using a linear combination of principal components to create orthogonal components for stocks in the LQ45 index. A 5-year time series dataset from 2018 to 2022 is employed. Analysis reveals four principal components explaining 84% of the total data variance, with a Kaiser-Meyer-Olkin (KMO) value of 90%. This indicates successful separation of stocks with different cumulative return patterns, return profiles, and risks. ANOVA statistical tests confirm significant differences between stock groups at a 95% confidence level. The portfolio built from these four components provides excellent hedging, as indicated by a high Sharpe Ratio. Even stocks with negative returns are offset by those with positive returns within the portfolio. Backtesting and forward testing demonstrate that this portfolio outperforms the combined stock index. This research reinforces that a PCA-based hedging approach with generated principal components can mitigate risk and enhance investment returns in volatile market conditions.