High frequency trading in Australian derivatives markets

This dissertation investigates price discovery dynamics in Australian equity and options markets. This research provides empirical evidence of an important issue in the market microstructure literature in high frequency trading (HFT) and high frequency traders (HFTs). The increase in HFT has led to a shift in dynamics in equity and options markets due to the higher proportion of computer generated algorithms trading with, and also against, human traders. Each chapter in this dissertation addresses issues in the current literature that have been identified in an Australian context, which can assist academics and regulators understand more about HFTs and their effect on prices in smaller, less liquid markets. 

The first issue investigated is how HFTs affect liquidity and price discovery at the intraday level on the Australian Stock Exchange (ASX). Previous research shows that HFT increases the level of liquidity and the speed of price adjustment in both equity and options markets. However, very little is known about whether this is consistent in smaller and less liquid markets. This is mainly due to the higher proportion of computer based traders compared with human traders in smaller markets. For example, in smaller markets where the number of stock and option trades may be few and far between, it is imperative that we know whether HFTs have the same liquidity providing effects that have been shown to exist in larger markets. Due to the size of the Australian market and liquidity constraints, quote prices are used. Additionally, quote volumes, average duration between quotes and a market momentum indicator are added to test whether broader trading patterns exist. The results are consistent with the literature, in that prices adjust faster and liquidity levels increase. 

The second issue examined is intraday price discovery dynamics on the ASX from an event-study perspective. Across the literature, a majority of findings show that options usage increases around news announcements and that this contributes to the price discovery of equities post-announcement. Some studies conjecture that it makes no difference to price discovery whether or not equities have listed options. For equities with listed options informed traders, including HFTs, increase the speed at which prices adjust. However, many of these studies incorporate trade price data from the New York Stock Exchange (NYSE) and other large exchanges. Due to liquidity constraints and the small size of the ASX, this research uses quote prices for equities and options and focuses only on the pre-announcement trading period. Additionally, the scope of announcements examined is reduced to earnings announcements. A benchmark that replicates an average trading day is used to measure whether the price impacts in the lead-up to earnings announcements are significantly different. Some results are consistent with prior findings that informed traders turn to options markets prior to earnings announcements to hedge their portfolios and also speculate on price movements, which, in turn, affects the price discovery of the underlying equities. In particular, trading in the pre-announcement period is significantly different from an average trading day. The evidence suggests that there may be information leakage in the lead-up to earnings announcements, from which informed traders benefit. 

Compared with other issues studied in the literature, underlying equities become less important to the price discovery of call and put options in the lead-up to earnings announcements compared to the benchmark. When controlling for the cost of trading and intent to trade through quote bid-ask price spreads and total quote volumes respectively, results are mostly consistent. However, this changes marginally only for earnings announcements classified as ‘good’ and in the final minutes of trade prior to the good earnings announcements. Prior to events classified as ‘bad’, the equity is even less important to the price discovery process compared with the benchmark. In a further contrast to other findings in the literature, quote volumes have very little impact on the price discovery of call or put options. Furthermore, increases in both call and put quote bid-ask price spreads have negative impacts on their respective options prices compared to the benchmark. On an average trading day these effects are opposite - that is, an increase in quote bid-ask spreads has positive effect on their respective options prices. All results are robust to changes in the earnings announcement classifications.