| Resumo: | We provide new evidence on the role of real money balances in terms of explaining equity risk premia by using a rich cross-section of average stock returns (associated with 11 major CAPM anomalies). By estimating Euler equations associated with a cash-in-advance (CIA) model, we find that such model produces substantially smaller pricing errors than the baseline consumption model, while still generating lower estimates of the risk aversion coefficient. The estimates of the parameter governing the share of cash goods are highly significant and plausible in economic terms. A transaction-costs model and a money-in-the-utility model perform considerably worse than the CIA model, both in terms of statistical fit and in terms of the plausibility of the structural parameter estimates. Moreover, a linear version of the CIA model also largely underperforms the corresponding non-linear model.
|
|---|