Comparative accuracy analysis of the CAPM model and its multifactor extensions in explaining the stock returns

Anton Goncharov

Master's thesis

Abstract:

Given that the U.S. economy and its financial market have faced many challenges and obstacles during the time period between 01.01.2018 and 31.12.2021, there was an opportunity for me to test asset pricing models in a recent “stressful” economic environment. Therefore, the following master thesis aims to study, assess and compare the performance of the capital asset pricing model, Fama-French three-factor model, and Carhart four-factor model by evaluating them through the use of OLS regression models and formal testing. The data was collected from Yahoo finance and an online data library created by Kenneth R. French. The methodology considers 30 individual U.S. stocks and targets two specific performance metrics, adjusted R-squared along with residual standard errors. The findings reveal that during the sampling period, the models’ average explanatory power in terms of adjusted R-squared increases when the capital asset pricing model is extended by the extra factors being HML, SMB and Momentum. However, when performing formal testing of residual standard error which is the second performance metric, I was not able to identify a significant difference or improvement in the performance of previously mentioned models.

Abstract:

Given that the U.S. economy and its financial market have faced many challenges and obstacles during the time period between 01.01.2018 and 31.12.2021, there was an opportunity for me to test asset pricing models in a recent “stressful” economic environment. Therefore, the following master thesis aims to study, assess and compare the performance of the capital asset pricing model, Fama-French three-factor model, and Carhart four-factor model by evaluating them through the use of OLS regression models and formal testing. The data was collected from Yahoo finance and an online data library created by Kenneth R. French. The methodology considers 30 individual U.S. stocks and targets two specific performance metrics, adjusted R-squared along with residual standard errors. The findings reveal that during the sampling period, the models’ average explanatory power in terms of adjusted R-squared increases when the capital asset pricing model is extended by the extra factors being HML, SMB and Momentum. However, when performing formal testing of residual standard error which is the second performance metric, I was not able to identify a significant difference or improvement in the performance of previously mentioned models.

Keywords

U.S. financial market, CAPM, Fama-French three-factor model, Carhart four-factor model, performance of OLS regression models, stocks, returns, asset pricing models, explanatory power

Keywords

returns, Carhart four-factor model, explanatory power, U.S. financial market, stocks, CAPM, Fama-French three-factor model, performance of OLS regression models, asset pricing models

Language used: English

Date on which the thesis was submitted / produced: 3. 5. 2022

Thesis defence

• Date of defence: 1. 6. 2022
• Supervisor: Karel Brůna
• Reader: Viktar Dudzich

Vysoká škola ekonomická v Praze

Master programme:
Finance and Accounting