An empirical analysis of the limit order book and the order flow in the Singapore Exchange equities market

Abstract

This thesis examines the trading behaviour of investors in the equities market of the Singapore Exchange (SGX). One of the main aims is to better understand how the heterogeneity of market participants with their strategies may be linked to the properties of order book dynamics. The conclusions drawn from the three essays could potentially influence the decisions of regulators, exchange operators, and investors in securities markets.

The first essay investigates client order execution strategies in terms of how they place orders across the limit order book. The results suggest that a multi-price trading strategy, where investors place a “network” of buy and sell limit orders simultaneously, is a viable strategy and might even be profitable enough for investors to persist with the strategy. It is also shown that institutional investors tend to place more aggressive orders than retail investors.

The second essay examines the effects prior order transactions, limit order book attributes, and market variables have on order submission and cancellation choices. It is found that while order transactions have a positive same order type order-by-order serial correlation, when order flow is aggregated by time intervals, the correlation of the changes in aggregated order flow of the same order type between consecutive time intervals is negative even over short time periods of five seconds. The positive order-by-order serial correlation for orders of the same type is present regardless of whether the incoming transaction is by an institutional investor or a retail investor.

The final essay studies how changes in the limit order book and market conditions affect the aggressiveness levels of orders after their submissions and thus influence order cancellations. Traders who place orders near the top of the order book actively monitor those orders and decisions to cancel orders are significantly affected by changes in order book conditions such as order depth, stock volatility, market volatility, price returns, and net trade imbalance. In managing and balancing non-execution risk and option-to-trade risk, the probability that traders cancel their orders also differs depending on whether their orders have moved further away, remain unchanged, or are nearer to the best same-side prices.