Objective –This study examines how the goal of an investor is to establish an optimal investment risk structure which maximizes profits by incurring fewer losses at a certain level of market risk. This research aims to determine the accuracy of the CAPM and APT models in predicting stock returns, as measured using Mean Absolute Deviation (MAD).Design/Methodology –The methods employed is a quantitative approach. The population for this study includes companies included in the LQ45 index during the 2020-2022 period consisting of monthly observations spanning from January to December. A purposive sampling technique was used to select 30 sample companies. The reason for using the LQ-45 index is because this index is an index in which there are 45 issuers. Apart from that, the shares included in the LQ-45 calculation are considered to reflect the movement of actively traded shares which will influence market conditions, consisting of shares with high liquidity and market capability, as well as growth prospects and fairly stable financial conditions.Results –The MAD calculation shows that the APT model is more accurate than the CAPM model. The choice of model use can be adjusted to the preferences of each investor. CAPM is a forecasting model that only uses market return factors, making it suitable for investors who want to predict stock returns easily and simply. On the other hand, APT can be used by investors who want to know in detail what macro factors influence changes in stock prices.Research Limitations/Implications – Increasing the length of time that researchers spend doing their research is recommended in order to improve the accuracy of their forecasts about future stock returns.Novelty/Originality –This instrument is highly beneficial for investors who are looking for a clear and effective technique to anticipate the returns on their stock investments. Investors who want a detailed grasp of the precise macroeconomic issues that affect swings in stock prices may find that applying the Arbitrage Pricing Theory (APT) is useful.
Read full abstract