The Inverse Market Impact of Limit Orders : Quantifying short-term mitigating effects of liquidity adding to price shifts

Abstract

The research focuses on establishing a method to measure market impact for a single limit order, as this has been largely neglected in traditional market impact literature. The aim of the research is to find the variables most likely to influence the impact from the point of view of existing research, to establish their connection to the impact observed using statistical methods and to verify the observations using a machine learning model. The research was able to identify two variables that could be directly related to the impact of the limit order, as well as to identify what kind of results could be expected from the data. The research ended up choosing the limit order booking size in terms of shares, as well as the limit order distance from the mid- price. The distance was measured as the price of the limit order book posting and the mid- price and acted as a proxy for the limit order book shape prior to the order posting. With the features chosen, we expected to find a decreasing, concave relationship between the limit order size and little to no discernible relationship between the posting distance and impact. The research found that as there exists a strong, linear relationship between the subsequent price shift and the limit order posting distance and size. Surprisingly, the posting distance was even more highly connected to the impact following a limit order posting, suggesting that there is a direct link to the order book shape in the impact function for limit orders. The research suggests that the limit order impact should be thought of as an inverse of the impact for trades, as limit orders do not have a direct effect on the mid- price changing but rather mitigate the price changes. The limit order impact is thus inverse compared to the impact of trades due to the liquidity adding.