MI3 Academic Research: Exchange Competition, Fragmentation, and Market Quality
Plato Partnership has released a new piece of independent academic research into how the market quality effects of fragmentation depend on the type of exchange competition. The research was conducted by Michał Dzieliński and Björn Hagströmer from Stockholm Business School, Stockholm University, and Chengcheng Qu from the College of International Management, Ritsumeikan Asia Pacific University as part of Plato Partnership’s MI3 Academic initiative.
Executive Summary
Market fragmentation is arguably the biggest change to the equity market microstructure in the past two decades. Around 2007, regulators in the US and Europe opened up for competition between exchanges. The same stock could then be traded at multiple venues, not only where it was listed. Nowadays, only 40% of the trading activity in European stocks takes place at the listing exchanges. In addition to the competition from other exchanges (known as multilateral trading facilities, MTFs), trades are routed to less transparent venues such as dark pools, single-dealer platforms, and over-the-counter markets.
This paper investigates how different types of exchange competition influence market quality, in terms of market liquidity and market efficiency. Previous research has established that market fragmentation is good for liquidity, but fragmentation is in turn a consequence of exchanges competing with each other, potentially in several dimensions. This paper seeks to disentangle those dimensions, and their impact on market quality.
Exchange competition may be categorized as follows:
Liquidity. The most obvious area of competition is prices, which for trading services primarily boils down to exchange fees. If competition leads to lower fees, it should result in tighter bid-ask spreads and greater depth.
Speed. Exchanges can attract trade flows through technological innovations, such as timely data feeds, server colocation, and fast matching engines. An alternative approach is to attract slow traders by screening out the fastest participants, e.g., by introducing speed bumps.
Compliance. With more than 30 listing venues in Europe and a lack of a unified rule book, the pan-European offerings of the MTFs can be another competitive edge. Foreign investors may be able to reduce their cost of compliance by routing all their flow to the same MTF.
This paper focuses on liquidity competition, measured by how much the MTFs improve the bid-ask spread, or by the fraction of market depth provided by the MTFs. An interesting finding is that the correlation between liquidity competition and the degree of market fragmentation (measured as the MTF share of trading volume) is low. This suggests that exchanges attract trades not only due to improved liquidity, but also for other reasons, such as speed and compliance.
To identify the effects of exchange competition, this paper studies the sudden centralization of trading in Swiss stocks in July 2019. At that time, the Swiss financial market’s status of regulatory “equivalence” with the EU was revoked. The Swiss equity market went from having around 30% of the trading activity at the MTFs, on average, to zero.
The results show that after the centralization of trading, the liquidity in Swiss stocks declines. The bid-ask spreads widen and the market depth falls. This is consistent with prior studies showing that fragmentation benefits market liquidity. Surprisingly, however, the effects are not stronger for the stocks with higher fragmentation before the event. Instead, it is the stocks with the greatest liquidity competition, which display the largest drops in liquidity.
Specifically, the effective bid-ask spread in stocks with high liquidity competition increase by 23\% after the centralization. For other Swiss stocks, the increase is merely 5\%---on par with a set of German stocks that are used as a control group. Similar results hold for the quoted bid-ask spread. Furthermore, the depth posted in the order book drops by 33\% for Swiss stocks with high liquidity competition, and only by 17\% for other Swiss stocks. For the German control group, depth falls by 7\% over the same time period.
The findings hold for liquidity measured on aggregate across exchanges, as well as for the Swiss exchange in isolation. The latter is relevant to investors who do all their trading at the listing venue, which is common for, e.g., retail investors.
The results show no consistent effect of exchange competition on the efficiency of prices.
The study concludes that different types of exchange competition have different impact on market liquidity. Market fragmentation is associated with liquidity benefits, but that boost hinges on the exchanges competing on liquidity provision. Thus, if the goal of policymakers is to reduce overall transaction costs, they should prioritize competition on liquidity among the fragmented marketplaces.
