Liquidity supply and adverse selection in a pure limit order book market

Based on a structural model we analyze adverse selection costs and liquidity supply in a pure open limit order book market. Given the discontenting empirical model performance reported in the previous literature, we relax restrictive assumptions of the underlying theoretical model concerning order book equilibrium and the distribution of market order volumes. We demonstrate that the resulting revised econometric methodology delivers considerably improved empirical results. Employing the alternative approach in a cross sectional analysis we provide evidence that adverse selection costs are more severe for smaller capitalized stocks, and find empirical support for one of the main hypothesis put forth by the theory of limit order book markets, which states that liquidity supply and adverse selection costs are inversely related. We also show that adverse selection component estimates based on the formal model and those obtained using popular model-free methods are closely correlated. This result indicates the robustness of the structural model, but also provides a theoretical underpinning for the application of the ad hoc method to limit order book data.