We investigate whether stock prices reflect the asymmetric persistence of accruals and cash flows resulting from conditional conservatism. Using the Mishkin () test (MT), we provide further evidence on the earnings fixation explanation for the accrual anomaly. We also apply panel estimation techniques that significantly affect market efficiency inferences. Our results suggest that over our sample period (i) investors seem to partially anticipate asymmetric persistence in accruals and cash flows, (ii) the accrual anomaly originates in the mispricing of accruals in years of economic gains, even though the differential persistence between accruals and cash flows is greatest in years of economic losses, (iii) investors respond differently to accrual and cash flow surprises and, therefore, they do not naively fixate on earnings surprises, and (iv) after clustering standard errors in the MT by firm and year dimensions, there is no longer evidence of cash flow mispricing, while the statistical significance of accrual mispricing falls. All our findings contradict the earnings fixation explanation for the accrual anomaly. Our study has implications for understanding the accrual anomaly in relation to accrual dynamics, as well as for researchers interested in using the MT framework to test the rationality of investor expectations more generally.