Consumer sentiment and time-varying betas: Testing the validity of the consumption CAPM on the Johannesburg Stock Exchange

ABSTRACT

We test both the conditional and unconditional versions of the consumption capital asset pricing model (CCAPM) on the Johannesburg Stock Exchange, for the period 1988–2018, and compare its performance with that of the CAPM and the Fama-French three- and five-factor models. We use the consumer confidence index as an instrument to parameterise shifts in betas over time in conditional models. In order to study the robustness of the results at a higher frequency than that of consumption data, we use the mimicking portfolio of the stochastic discount factor tied to the model. Our results show that in all cases the conditional CCAPM performs satisfactorily, outperforming both the CAPM and the Fama-French three-factor model. These results suggest that South African consumption growth and consumer sentiment help explain a large fraction of the expected returns in the Johannesburg Stock Exchange.

KEYWORDS:

• CCAPM
• CAPM
• Fama-French model
• consumer confidence index
• stochastic discount factor
• mimicking portfolio

JEL CLASSIFICATIONS:

• G12
• G15

Notes

1 Classic Utility Theory generally assumes that the marginal utility of consumption is always positive but decreasing as consumption increases, which is consistent with the positive SDF that arises in any market that satisfies the law of one price and leaves no arbitrage opportunities.

2 This literature is too large to summarise here. Chapter 5 in Campbell (2018) provides a complete review of the main findings in stock return predictability.

3 All data is publicly available at Rojo-Suárez & Alonso-Conde (2020).