Determining the information share of liquidity and order flows in extreme price movements

Liquidity and order flows have been found to be major causes of extreme price movements (EPMs) in previous studies. However, few studies have clarified whether the impacts of these factors to EPMs are transient or permanent. In this paper, we represent the fluctuation of liquidity as a time series of price. The measurement of permanent price impact is converted to the price discovery problem solved by a quantile vector error correction model. Empirical results using the high frequency data in the Chinese stock market indicate that both liquidity and order flows contribute to the permanent component of the EPMs. However, liquidity is the dominating factor, which accounts for more than 60-80% of the information share in EPMs scenarios.