A Comparative Analysis of the Determinants and Pricing of Liquidity in Floor and Electronic Trading Systems

Abstract

In recent years many stock exchanges have moved away from floor-based to automated-based trading systems. However, the choice between these alternative trading systems is a major concern for stock exchange regulators and designers, and the impact of their merits on market characteristics (e.g. liquidity) is controversial. This thesis is motivated by the desire to shed light on this controversy, and therefore aims to offer a comparative analysis of various aspects of liquidity under floor and automated trading systems. More specifically, within the context of different trading systems (i.e. floor versus electronic), this thesis examines three empirical issues: firstly, the determinants of market-wide liquidity and its time-series behaviour; secondly, whether market-wide and firm-specific liquidity are priced in assets returns; and finally, whether the cross-sectional variations in firm-specific liquidity could be explained by the cross-sectional variations in information asymmetry and divergence of opinion.

The findings of this thesis can be summarized as follows. Firstly, market-wide liquidity is significantly influenced by market returns, market volatility, interest rate variables and the announcement of macroeconomic indicators. Market-wide liquidity also shows distinct day-of-the-week regularities and a distinct pattern around holidays. The impact of some factors on market-wide liquidity, and the time-series behaviour of market-wide liquidity on the floor trading system in some markets is higher than that on the electronic trading systems. Secondly, market-wide liquidity has a significant impact on assets returns, and after controlling for its effect, firm-specific liquidity has a significant effect on risk-adjusted returns. The liquidity premium required on market-wide and firm-specific liquidity, for some proxies of liquidity in some markets, is higher on an automated trading system than on a floor trading system. Finally, firm-specific liquidity is negatively related to the level of information asymmetry. However, the evidence for the impact of divergence of opinion on firm-specific liquidity is inconclusive; a higher level of divergence of opinion results in higher liquidity, which supports the optimistic view; and firm-specific liquidity decreases with divergence of opinion, which is consistent with the view that disagreement among investors is a source of risk. Additionally, after automation, the impact of information asymmetry (divergence of opinion) on firm-specific liquidity is greater (lesser) than that before automation.

Overall, this thesis demonstrates that the design and the structure of markets is closely linked to the latter’s performance and that the change to automated trading systems has significant implications for liquidity. As such, this study should be a valuable reference point for stock exchanges that have introduced automation, or are considering doing so.