||This paper studies the effect of accounting conservatism on the relation between equity value and accounting data in the US banking industry. The bank industry offers a unique opportunity to measure accounting conservatism with relative precision by studying one important accrual---loan loss accrual. In addition, bank regulator plays an important role in influencing variations of accounting conservatism across banks as well as over time. We develop a simple model from the existing valuation models of Ohlson (1995) and Zhang (2000) to show how accounting conservative bias should impact the relation between the equity value and accounting data. We find that measures of accounting bias are incrementally useful in explaining the valuation of bank equity. Specifically, the change in cumulative accounting conservative bias is positively related to the market value, which is consistent with the prediction of our valuation model; the cumulative accounting bias is negatively related to market value, which is not consistent with the prediction from valuation model. Further analysis shows that the negative effect of cumulative accounting conservatism on valuation only exists for a particular group of banks with regulatory enforced conservatism. This result is consistent with enforced accounting conservatism capturing the conflict of interests between bank regulator and shareholders (Watts, 2003). In return regressions, we do not find significant evidence for the effect of accounting bias in explaining the contemporaneous stock return.