«Abstract Two important characteristics of current equity markets are the large number of trading venues with publicly displayed order books and the ...»
The impact of dark trading and visible
fragmentation on market quality
Hans Degryse Frank de Jong Vincent van Kervel
CEPR discussion paper 8630
TILEC discussion paper No 2011-26
CentER Discussion Paper Series No. 2011-069
Two important characteristics of current equity markets are the large number of
trading venues with publicly displayed order books and the substantial fraction of
trading that takes place in the dark, outside such visible order books. This paper evaluates the impact of dark trading and fragmentation in visible order books on liquidity.
We consider global liquidity by consolidating the limit order books of all visible trading venues, and local liquidity by considering the traditional market only. We ﬁnd that fragmentation in visible order books improves global liquidity, whereas dark trading has a detrimental effect. In addition, local liquidity is lowered by fragmentation in visible order books, which suggests that the beneﬁts of fragmentation are not enjoyed by market participants who resort only to the traditional market.
JEL Codes: G10; G14; G15;
Keywords: Market microstructure, Fragmentation, Dark trading, Liquidity CentER, EBC, TILEC, Tilburg University.
CentER, Tilburg University Corresponding author, TILEC, CentER, Tilburg University, Email: firstname.lastname@example.org.
The authors thank Albert Menkveld, Juan Ignacio Peña, Mark Van Achter, Gunther Wuyts, seminar participants at the 2010 Netherlands Authority for the Financial Markets workshop on MiFID, the 2010 Erasmus liquidity conference (Rotterdam), the European Retail Investment Conference (ERIC) 2011, the CNMV international conference on securities markets (2011), and seminar participants at the K.U. Leuven and Tilburg University for helpful comments and suggestions.
Electronic copy available at: http://ssrn.com/abstract=1815025 1 Introduction Equity markets in the US, Europe and Canada have seen a proliferation of new trading venues. The traditional stock exchanges are challenged by a variety of trading systems, such as electronic communication networks (ECNs), broker-dealer crossing networks, dark pools and over-the-counter markets (OTC). Consequently, trading has become dispersed over many trading venues –visible and dark– creating a fragmented market place. These changes in market structure follow recent changes in ﬁnancial regulation, in particular the Regulation National Market System (Reg NMS) in the US and the Market in Financial Instruments Directive (MiFID) in Europe.
An important question is how market quality is affected by the many and different types of competing venues. In this paper, we study the impact of market fragmentation on liquidity, which is an important aspect of market quality. We investigate the impact of different types of fragmentation by classifying trading venues according to pre-trade transparency into visible and dark venues, i.e., with and without publicly displayed limit order books.
According to this deﬁnition, US stocks have a dark market share of approximately 30% and European blue chips of 40%.1 Recently, the SEC is conducting a broad review of current equity markets, and is particularly interested in the effect of dark trading on execution quality.2 Both the impact of fragmentation in visible order books and dark trading on equity markets have since long interested researchers, regulators, investors and trading institutions.3 In a recent study, O’Hara and Ye (2011) ﬁnd that fragmentation lowers transaction costs and increases execution speed for NYSE and Nasdaq stocks. They do not distinguish, however, between the differential impact on liquidity of fragmentation stemming from visible and dark trading venues. The main contribution of our paper is that we disentangle the liquidity effects of both fragmentation in visible order books ("visible fragmentation" for short) and dark trading. In addition, we address regulatory issues of fair markets and retail investor protection. To this end, we distinguish between liquidity aggregated over all trading venues (global liquidity) and liquidity of the traditional market only (local liquidity). Global liquidity is available to investors using Smart Order Routing Technology and 1 Speech of SEC chairman Mary Schapiro, “Strengthening Our Equity Market Structure”, US SEC New York, Sept 7, 2010, and Gomber and Pierron (2010) for Europe.
2 See the speech of Schapiro, and the SEC concept release on equity market structure, February 2010,
File No. S7-02-10. 3 See section 2 for a review of the academic literature.
Electronic copy available at: http://ssrn.com/abstract=1815025 local liquidity is accessible to investors who tap the traditional market only. We furthermore improve upon previous research by employing a new dataset that covers the relevant universe of trading platforms, provides stronger identiﬁcation of fragmentation and allows for improved liquidity metrics.
Our main ﬁnding is that the effect of visible fragmentation on global liquidity is generally positive, while the effect of dark trading is negative. An increase in dark trading of one standard deviation lowers global liquidity by 9%. The effect of visible fragmentation has an inverted U-shape, i.e. the marginal effect is declining when fragmentation increases. Employing our most conservative estimates, the optimal degree of visible fragmentation improves global liquidity with approximately 32% compared with a completely concentrated market. In addition, we ﬁnd that the gains of visible fragmentation mainly hold for liquidity close to the midpoint, i.e. at relatively good price levels, but to a much lesser extent for liquidity deeper in the order book, which improves by only 12%. This result suggests that newly entering trading venues with visible order books primarily improve liquidity close to the midpoint. Furthermore, compared to small stocks, trading in large stocks is more fragmented and its liquidity beneﬁts twice as much from fragmentation. This suggests that competition between trading venues is ﬁercer for large stocks than for small stocks.
While global liquidity beneﬁts from fragmentation, we ﬁnd that the market quality at the traditional stock exchange is worse off as local liquidity close to the midpoint reduces by approximately 10%. As such, investors without access to Smart Order Routing Technology are worse off in a fragmented market, especially for relatively small orders.
We address the impact of fragmentation on market liquidity by creating, for every ﬁrm, daily proxies of visible fragmentation, dark activity and liquidity, employing information from all relevant trading venues. Speciﬁcally, we study a period before fragmentation set in, January 2006, until the end of 2009, when markets were already quite fragmented.
Similar to Foucault and Menkveld (2008), we select all Dutch mid- and large-cap stocks, which are relatively large with an average market capitalization approximately twice that of the NYSE and Nasdaq stocks analyzed in O’Hara and Ye (2011). We measure the degree of visible fragmentation by the Herﬁndahl-Hirschman Index (H H I, the sum of the squared market shares) based on executed trades on all visible trading venues. Dark trading is captured by the market share of traded volume on dark venues and OTC. Then, for each stock we construct a consolidated limit order book (i.e., the limit order books of all Electronic copy available at: http://ssrn.com/abstract=1815025 visible trading venues combined) to get a complete picture of the global liquidity available in the market. Based on the consolidated order book we analyze global liquidity at the best price levels, but also deeper in the order book. This is important, as the depth of the order book reﬂects the quantity immediately available for trading and accordingly the price of immediacy. Next to global liquidity, we also analyze local liquidity, available at the traditional exchange only.
Our panel dataset helps to identify the exogenous relation between liquidity and fragmentation by means of ﬁrm-quarter ﬁxed effects. The inclusion of ﬁrm-quarter dummies implies that the impact of fragmentation on liquidity stems from variation within a ﬁrmquarter, making the analysis robust to various industry speciﬁc shocks and time-varying ﬁrm speciﬁc shocks. Furthermore, in order to address concerns about endogeneity of visible fragmentation and dark trading, we use instrumental variables. Similar to O’Hara and Ye (2011), we use as instruments for visible fragmentation the average order size of the visible competitors, and also the number of limit orders to market orders on the visible competitors. Dark trading is instrumented by the average dark order size.
Our ﬁndings on liquidity can be related to several recent studies. The positive effect of fragmentation on visible trading venues is consistent with competition between liquidity suppliers, since the compensation for liquidity suppliers, the realized spread, reduces with fragmentation. A similar argument is made in Foucault and Menkveld (2008), who study competition between the LSE and Euronext for Dutch stocks in 2004, and ﬁnd that fragmentation over these two traditional stock markets improves liquidity. The negative impact of dark trading is consistent with a "cream-skimming" effect between dark and visible markets, since the informativeness of trades, the price impact, strongly increases with dark activity. The "cream-skimming" effect is predicted by Zhu (2011), who argues that informed investors face low execution probabilities in crossing networks and dark pools because they typically trade at the same side of the order book. Consequently, dark markets attract predominantly uninformed traders, leaving the informed trades to visible markets.
The negative effect of dark trading can also be related to pre-trade transparency, as visible markets are more efﬁcient because of faster and cheaper access to information, in line with e.g., Biais, Bisière, and Spatt (2010) and Boehmer, Saar, and Yu (2005).
In line with our results, Weaver (2011) shows that off exchange reported trades, which mostly represent dark trades in his sample, negatively affect market quality for US stocks.
In contrast to our results, Buti, Rindi, and Werner (2010a) ﬁnd that dark pool activity is positively related to liquidity in the cross section. In their time series regressions however, similar to ours, the effect of dark pool activity on liquidity is economically insigniﬁcant and statistically marginally signiﬁcant. We contribute to Buti, Rindi, and Werner (2010a) by controlling for visible fragmentation.
In sum, our ﬁndings imply a deeper understanding of the more general conclusion of O’Hara and Ye (2011) that fragmentation does not harm market quality. We show that the composition of the fragmentation – visible versus dark – determines the total impact of fragmentation on market quality. Moreover, our conclusions especially relate to the issues raised by the SEC on the beneﬁts and drawbacks of stock market fragmentation, and show that the beneﬁts are not equally enjoyed by all stock market participants. This latter ﬁnding is particularly relevant to regulators who strive for fair markets and protection of retail investors.
The remainder of this paper is structured as follows. Section 2 discusses literature on competition between exchanges. The dataset and liquidity measures are described in sections 3 and 4. Section 5 explains the methodology and main results, while section 6 reports a series of robustness checks. Finally, section 7 concludes.
2 Literature on fragmentation and market quality
There is a trade-off between order ﬂow fragmentation and competition. A single exchange beneﬁts from lower costs, compared with a fragmented market structure. These consist of the ﬁxed costs to set up a new trading venue; ﬁxed costs for clearing and settlement;
costs of monitoring several trading venues simultaneously; and advanced technological infrastructure to aggregate dispersed information in the market and connect to several trading venues. Also, a single market that is already liquid will attract even more liquidity due to positive network externalities (e.g. Pagano (1989a), Pagano (1989b) and Admati, Amihud, and Pﬂeiderer (1991)). Each additional trader reduces the stock’s execution risk for other potential traders, attracting more traders. This positive feedback should cause all trades to be executed at a single market, obtaining the highest degree of liquidity.
However, while network externalities are still relevant, nowadays they may be realized even when several trading venues coexist. This happens to the extent that the technological infrastructure seamlessly links the individual trading venues, creating effectively one market. From a broker’s point of view, the market is then virtually not fragmented, which alleviates the drawbacks of fragmentation (Stoll, 2006).4 In addition, fragmentation might also enhance market quality, as increased competition among liquidity suppliers forces them to improve their prices, narrowing the bid-ask spreads (e.g. Biais, Martimort, and Rochet (2000) and Battalio (1997)). Conﬁrming a competition effect, Conrad, Johnson, and Wahal (2003) ﬁnd that Alternative Trading Systems in general have lower execution costs compared with brokers on traditional exchanges. Furthermore, Biais, Bisière, and Spatt (2010) investigate the competition induced by ECN activity on Nasdaq stocks. They ﬁnd that ECNs with smaller tick sizes tend to undercut the Nasdaq quotes and reduce overall quoted spreads.
Differences between trading venues may arise to cater to the needs of heterogeneous clientele. For example, investors differ in their preferences for trading speed, order sizes, anonymity and likelihood of execution (Harris (1993) and Petrella (2009)). In the US, Boehmer (2005) stresses the trade-off between speed of execution and execution costs on Nasdaq and NYSE, where Nasdaq is more expensive but also faster. In order to attract more investors, new trading venues may apply aggressive pricing schedules, such as make and take fees (Foucault, Kadan, and Kandel, 2009). The fact that some investors prefer a particular trading venue can also lead to varying degrees of informed trading at each exchange.