TITLE:
Empirical Testing of the Capital Asset Pricing Model (CAPM): An Assessment of Its Predictive Ability in Contemporary Financial Markets
AUTHORS:
Rongsen Chen
KEYWORDS:
CAPM Model, Prediction, Accuracy
JOURNAL NAME:
Modern Economy,
Vol.17 No.10,
October
19,
2026
ABSTRACT: This dissertation analyses the empirical forecasting power of CAPM for seven developed and emerging markets from 2003 to 2024. While the model has been a major contributor to finance theory and practice, over the last few decades, evidence has emerged that suggests it is not a good representation of the risk-return relationship in today’s financial markets. The study creates beta sorted portfolios using monthly returns of large listed stocks, tests the CAPM against the three factor and five factor models proposed by Fama and French, conducts a time series test, a Fama and MacBeth cross sectional regression, a rolling out of sample forecasting exercise, and analyzes two periods of crisis. What the results show is a definite conflict. The strict CAPM is rejected statistically in most markets, as the pricing errors are picked up by the multifactor models as they should be. The ordering, by the way, is swapped out of sample and the single factor CAPM generates the least out of sample prediction errors in 4 of the 7 markets. In times of crises, the level of predictability falls, mostly due to varying betas in less integrated markets. The results indicate that statistical rejection and practical usefulness are separable, and that the CAPM is a valid forecasting instrument.