Title: Limit-order-book models and high-frequency-trading
Authors: Lars Wendt - University of Duisburg-Essen (Germany) [presenting]
Abstract: Order book data from the London Stock Exchange is evaluated regarding order execution times and the effect of high-frequency trading on those. From raw order data the order books of a diversity of securities from different market segments are reconstructed. Based on the order books the time to first fill and completion of the orders are measured and compared to a previous work. It is found that order execution times have declined drastically. The idea of strategic runs has been used to identify high-frequency trading. This measure is applied to compare order execution times of non-high-frequency trading to high-frequency trading. The mean order execution time of high-frequency trading orders is lower than their normal trading counterpart. A possible reason for the decline of order execution times may be a fast changing spread. From the order book the best prices for every point in time are reconstructed and the order code of the order that sets this price is recorded. With this data at hand it is evaluated how many best-price-setting orders lead to a first fill or complete execution and if this differs for high-frequency trading and normal trading. It cannot be found that best-price-setting high-frequency trading orders are executed more often than their usual trading counterparts.