Purpose This study aims to examine whether the consumption capital asset pricing model (CCAPM) can explain the Australian equity market premium under incomplete markets with heterogeneous households. It provides the first household-level test for Australia and revisits prior work that relies mainly on aggregate consumption data, the complete market assumption and implied relative risk aversion (RRA) estimates that are implausibly high. Design/methodology/approach Using microdata from the Household, Income and Labour Dynamics in Australia longitudinal household survey (2006–2017), we construct four measures of household consumption and estimate the stochastic discount factor with unconditional and conditional CCAPM Euler equations under complete and incomplete market settings. We then extend the analysis from the market portfolio to size and value portfolios. Findings Household risk sharing in Australia does not fully smooth idiosyncratic consumption shocks. In the incomplete market setting, the CCAPM with disaggregated consumption matches the observed market risk premium with an RRA of about 2.16, whereas the complete market model requires an implausibly high RRA near 50. Total Consumption and Essentials explain the market premium, while Non-Durables and Services better capture premia associated with the equally weighted market, size portfolios and value portfolios. Conditioning on lagged market information modestly lowers effective risk aversion. Originality/value This study provides the first household-level CCAPM test for Australia under incomplete markets. It shows that cross-sectional dispersion in marginal utilities and careful measurement of household consumption restore sufficient pricing kernel volatility to reconcile the CCAPM with observed equity premia and clarify how consumption risk shapes asset prices.
Sun et al. (Thu,) studied this question.