Ebook The Role of Market Makers in Electronic Markets: Liquidity Providers on Euronext Paris
Starting in Toronto 30 years ago, electronic trading systems (ETSs) spread relentlessly across the globe (to Tokyo in 1982, Paris in 1986, Australia in 1990, Germany and Israel in 1991, and Switzerland in 1995) and now predominate as a trading mechanism worldwide. To a surprising extent, the onslaught has swept aside many of the hallmarks of the world’s leading stock exchanges. The dealer agency system at NASDAQ, for example, reported in late 2006 a smaller share of trading volume in NASDAQ-listed stocks (35%) than the sum of the Instinet (26%) and Brut (13%) ETSs that NASDAQ acquired during 2004-2005. No less startling, the panel and post brokers that make up the NYSE’s distinctive “crowd” are quickly disappearing from the floor. Block transactions of more than 10,000 shares have fallen from 52% of the NYSE’s trading volume in 2000 to 21% in 2007 (NYSE website, 2007), and automated trading of smaller orders is now responsible for more than half of all stock transactions (Bloomberg.com March 19, 2007). The legacy systems of market makers (e.g., NYSE specialists, NASDAQ dealers, Euronext Liquidity Providers) who facilitate buyer and seller search are steadily being displaced by trading algorithms.
Exchanges around the world are yet to settle on a definitive role for market makers. Among the top 7 markets represented in the Dow Jones-Stoxx Global 1800 Index, the London Stock Exchange, NYSE, NASDAQ, and Xetra operate with market makers across all stocks, Euronext Paris operates with market makers for selected stocks, while the Tokyo Stock Exchange and the Australian Stock Exchange do without them altogether. Market makers certainly continue to play a significant role in dampening price volatility by calling out liquidity suppliers to fill the distant steps of the limit order book. Most liquidity suppliers go flat when a mean-reverting market begins to trend. This is precisely when market makers can commit inventory and attract unposted liquidity to offset order imbalance. The central purpose of this paper is to explore what remains of the role of market makers in this new electronic trading environment.
Aitken, Cook, Harris and McInish (2007) find that the presence of a market maker increases effective and realized spreads as well as price impacts using a sample of 2,330 matched pairs from the seven leading markets worldwide. Although they matched stocks to NASDAQ to secure the broadest possible sample involving all ten deciles worldwide, ACHM’s (2007) Euronext Paris subsample arose principally from traded value deciles 1-5. This absence of the more thinly-traded firms raises the question of the validity of the application of their conclusions to the prohibition of market makers on Euronext Paris. The present study seeks to address this issue by matching Euronext Paris stocks with and without Liquidity Providers across all traded value deciles.
Euronext Paris provides a natural experiment in which to compare execution cost across thickly and thinly-traded stocks. Euronext Paris only permits Liquidity Providers (ALs) for selected stocks. Specifically, ALs are not permitted for members of the Euronext 100 index. The next 150 largest securities (those included in the Next 150 index), plus those stocks in the AEX, AMX, BEL 20, CAC 40, PSI 20 or SBF 120 have the option to engage an AL. All listed firms not falling into the aforementioned categories and wishing to have their stock traded continuously are obliged by Euronext Paris to have an AL (maximum 2 per security). So, the illiquid continuously-traded securities on Euronext are AL-traded, by definition. Our research design takes these stocks and finds the best possible matched pair on Euronext that is not AL-traded. We then regress execution cost measures on design differences and control variables plus their interactions.
Current events in the financial services industry offer two strong motivations for studying the net effect of market makers on execution costs. First, stock and futures exchanges are consolidating rapidly. Shareholders of the New York Stock Exchange and Euronext Paris agreed to merge in December 2006, and trading as a joint exchange began in April 2007. A natural question therefore arises as to which trading mechanisms should be retained from one of the exchanges in preference over the other. Venkataraman (2001) found execution costs were higher among large market capitalization stocks listed on NYSE (where there is a trading floor with market makers in operation) than among matched-pair stocks listed on the Paris Bourse (where trading takes place in a purely electronic environment without market makers).
Second, both American and European regulators are moving to force broker firms to provide their clients with the lowest execution cost available. Specifically, the European Commission’s Market in Financial Instruments Directive (MiFID) will impose this “best execution” requirement on broker firms after November 2007. Implicit in such a regulatory requirement is the need for broker firms to benchmark the likely execution cost of alternative trade venues. Regulators, exchanges, brokers and their customers should all therefore be intimately knowledgeable about the extent to which a market design feature like the presence of market maker influences execution costs.
Our research examines a comprehensive set of four measures of trade execution costs across markets: quoted spreads, price impact, effective spreads, and realized spreads. Much of the previous matched-pair research suffers because it collapsed multiple market design differences into a single composite contrast on one or more of these costs. This study however can escape such criticism because we are explicitly studying a single design difference (the presence/absence of a market maker) across two subsets of a single market. There are therefore no other design features or institutional contexts confounded in our contrast.
The remainder of the paper is organized as follows. Section 2 reviews recent changes in the financial regulatory environment in Europe (i.e., the MiFID) and presents the market microstructure of Euronext Paris. Section 3 discusses the previous research on the role of market makers as a determinant of execution costs and hypothesizes the effect we expect market makers to have on execution costs in Euronext Paris. Section 4 explains our measures of trading costs and the methodology we used to infer trade direction. Section 5 describes the data, sample selection criteria, our matching methodology and provides pair-wise comparisons of execution costs across matched pairs, under different matching regimes. A conclusion is provided.
Download
PDF Ebook The Role of Market Makers in Electronic Markets: Liquidity Providers on Euronext Paris
Posted in :
