This study investigates the convergence process toward efficiency of daily top losers. We find that significance of order imbalance coefficients decreases with increasing time interval, indicating evidences on convergence to market efficiency. A time-varying GARCH model is employed to examine the relation between order imbalance and volatility. The significance of order imbalance coefficients shows a decay pattern, which also supports convergence to market efficiency. We develop an imbalance-based trading strategy and can not make profits from these daily top losers under bid/ask price. A nested causality approach, which examines dynamic return-order imbalance relation during price formation process, confirms the results.