Accuracy of Single- and Multi-Index Models in Stock Investment Portfolios: Study on LQ45 Shares after the Covid-19 Pandemic in Indonesia

This study discusses the problem of the accuracy of the single- and multi-index models to form a stock investment portfolio. This study aims to reveal information related to the accuracy of the single- and multi-index models to form an optimal stock investment portfolio after the COVID-19 pandemic in Indonesia. The results of the analysis can be used to choose which model is more accurate to form an optimal stock portfolio that can provide returns for investors. This research is devoted to determining the optimal stock portfolio using the single- and multi-index models. The data were collected from stocks included in the LQ45 list for February-July 2022 with 5980 observations consisting of 115 market returns, 1150 industrial sector returns, and 4715 individual stock returns. The analytical method used was regression analysis to predict the expected returns from single- and multi-index models, standard deviation of residuals, and Wilcoxon's differential signed-rank test. The results of the study show that stock investment portfolios based on daily data using the single index model approach and the multi-index model have their respective levels of accuracy with an accuracy rate of 95% (alpha 5%). The main finding is that the multi-index model proved to be more accurate than the single-index model, where the standard deviations of the multi- and single-index models were 0.0215 and 0.0280, respectively. Keywords: individual stock returns, market returns, multi-index model, single-index model, Wilcoxon’s signed-rank test. DOI: https://doi.org/10.55463/hkjss.issn.1021-3619.61.64