Limit-order completion time in the London stock market

Abstract

This study develops an econometric model of limit-order completion time using survival analysis. Time-to-completion for both buy and sell limit orders is estimated using tick-by-tick UK order data. The study investigates the explanatory power of variables that measure order characteristics and market conditions, such as the limitorder price, limit-order size, best bid-offer spread, and market volatility. The generic results show that limit-order completion time depends on some variables more than on others. This study also provides an investigation of how the dynamics of the market are incorporated into models of limit-order completion. The empirical results show that time-varying variables capture the state of an order book in a better way than static ones. Moreover, this study provides an examination of the prediction accuracy of the proposed models. In addition, this study provides an investigation of the intra-day pattern of order submission and time-of-day effects on limit-order completion time in the UK market. It provides evidence showing that limit orders placed in the afternoon period are expected to have the shortest completion times while orders placed in the mid-day period are expected to have the longest completion times, and the sensitivities of limit-order completion time to the explanatory variables vary over the trading day.