«Robert P. Bartlett, III Justin McCrary University of California, Berkeley University of California, Berkeley School of Law School of Law February 12, ...»
Dark Trading at the Midpoint:
Pricing Rules, Order Flow and Price Discovery
Robert P. Bartlett, III Justin McCrary
University of California, Berkeley University of California, Berkeley
School of Law School of Law
February 12, 2015
JEL codes: G10, G15, G18, G23, G28, K22
Keywords: Tick size; high-frequency trading; dark pool, internalization; fragmentation; market quality 1
Dark Trading at the Midpoint:
Pricing Rules, Order Flow and Price Discovery
We exploit a regulatory discontinuity in the minimum pricing increment, or tick size, to explore the mechanism by which tick size regulations benefit “dark” trading venues at the expense of traditional stock exchanges and the associated effects on displayed liquidity and price discovery. Under Rule 612 of Regulation NMS, orders priced at or above $1.00 per share are required to be priced in penny increments, while all other orders can be quoted in subpennies. Using all trading data from 2011-2013, our regression discontinuity analysis of this $1.00 cut-off confirms that the penny tick size facilitates “queue-jumping” of exchanges by dark pools; however, we show that the most significant portion of such queue-jumping occurs by means of trades at the midpoint of the NBBO—a form of trading that is generally considered to have significant benefits for institutional investors. Further, while queue-jumping in dark venues is associated with a significant decrease in the quantity of displayed liquidity on public exchanges, we show that the largest effect is on liquidity providers utilizing high frequency trading (HFT) algorithms, thus revealing an important inverse relation between dark trading and HFT.
Finally, notwithstanding the drop in displayed liquidity associated with dark trading, a tick size rule that encourages queue-jumping of exchanges’ limit orders books is associated with an overall grater rate of actual trading. In light of recent concerns about HFT, these findings suggest that the deterioration of liquidity on public exchanges facilitated by wider tick sizes may be less detrimental for market quality than previously believed. They also emphasize the difficulty of stemming the flow of trades away from stock exchanges without prohibiting midpoint trading within non-exchange venues.
1. Introduction Recent years have witnessed a dramatic transformation of U.S. equity trading from a market dominated by just a handful of conventional stock exchanges to one involving dozens of dispersed trading venues. Urged on by Regulation National Market System (Reg NMS) in 2005, a host of non-exchange “dark” venues now exist that compete successfully with exchanges for trades in U.S. listed equities. As of the end of 2014, these venues accounted for more than 30% of all consolidated trading volume, leaving approximately one dozen stock exchanges to compete fiercely for the remaining volume. The fate of the once storied New York Stock Exchange provides but one example of the disruptive forces at play.
Accounting for nearly 83% of the consolidated volume of NYSE-listed stocks as recently as 2000, the “big board” accounted for just over 20% of such volume as of the end of 2014—its very survival now in the hands of the Intercontinental Exchange, a rival futures exchange that acquired the NYSE in 2012.
Within this environment, recent theoretical and empirical work has established a somewhat surprising link between the fragmentation of U.S. equity markets and the minimum price variation (MPV) for quoting equity securities (see, e.g., Kwan, Masulis, and McInish, 2014; Bartlett and McCrary, 2013; Buti, Rindi, and Werner, 2011). The reason arises from Rule 612 of Reg NMS which requires all orders submitted to exchanges and priced at or above $1.00 per share be priced in penny increments but which allows trading to occur in subpenny increments. In principle, a liquidity provider that faces a long queue of limit orders priced at the national best bid or offer (NBBO) on a public exchange can therefore turn to so-called “dark” trading venues to submit more aggressive, non-displayed orders to be executed in subpenny increments against incoming market orders. In this fashion, current MPV rules help contribute to the increasingly fragmented U.S. trading market, but at the potential cost of impairing price discovery.
In particular, by facilitating queue jumping of exchanges’ limit order books, existing MPV rules may discourage traders from providing publicly displayed liquidity used to infer market prices (Buti, Rindi, and Werner, 2011.) A pending proposal by the Securities and Exchange Commission (SEC) to increase the MPV to $0.05 only heightens these concerns given the potential such a change might have for increasing the incidence of queue-jumping in dark pools.
In this study, we investigate whether the prevailing MPV rule harms the incentive of liquidity providers to display trading interest by facilitating queue-jumping in dark pools. We offer three central empirical findings. First, to the extent the MPV rule facilitates queue-jumping in dark pools, it does so primarily by facilitating trades at the midpoint of the NBBO—a form of subpenny trading that is generally considered to have significant benefits for institutional investors. Second, queue-jumping in dark venues is associated with a significant decrease in the quantity of displayed liquidity; however, our empirical analysis indicates that the largest effect is likely to be on liquidity providers associated with high frequency trading (HFT). Finally, we demonstrate that notwithstanding the drop in displayed liquidity associated with queue-jumping, an MPV rule that encourages queue-jumping of exchanges’ limit orders books is associated with an overall greater rate of actual trading. In combination, these findings suggest that the deterioration of liquidity on public exchanges facilitated by wider tick sizes may be less detrimental for overall market quality than previously believed.
In our empirical tests, we consider a trade as potentially queue-jumping exchanges’ displayed order books because of the MPV rule where the trade both executes away from an exchange in a subpenny 3 increment and the execution price rests within the prevailing NBBO.1 While any trade meeting these conditions might reflect queue-jumping because of the MPV, we empirically measure the incidence of two distinct forms of subpenny queue-jumping that differ in the offsetting benefit they provide to investors through price improvement. In permitting subpenny trades in Reg NMS, the SEC justified the practice to enable broker-dealers to provide price improvement over the NBBO for incoming marketable orders.2 As emphasized in Buti, Rindi, and Werner (2011) and Dick (2010), however, the fact that Reg NMS does not require a minimum amount of price improvement provides no assurance that a marketable order will receive a meaningful improvement over the NBBO in a dark venue to compensate for any adverse effect such trading might have for liquidity provision on public exchanges. Consistent with this concern, Delassus & Tyc (2011) document a sharp increase over time in the incidence of subpenny trades that offer only de minimis price improvement (e.g., $0.0001 per share), suggesting subpenny trades can be motivated simply to “step ahead” of exchanges’ limit order books.
In contrast to such trading (which, in keeping with industry practice, we refer to as “stepping ahead”), a separate class of subpenny trades often execute at the midpoint of the NBBO. This latter form of subpenny trading, which we refer to as “subpenny midpoint trading,” typically arises from nondisplayed orders from investors with instructions to execute at the NBBO midpoint against marketable orders submitted to a trading venue. As modeled in Buti and Rindi (2012), such orders can be used by aggressive traders to undercut depth at the top of exchange limit order books given that these orders provide superior pricing to an incoming marketable order. However, the fact that these trades maximize price improvement for both sides of the transaction by splitting the spread between buyer and seller has historically diminished concerns that they might harm price discovery without any offsetting benefits.
Indeed, the SEC adopted this position in devising its controversial “trade-at” rule included in the agency’s pending pilot study to increase the MPV to $0.05.3 In light of the agency’s concern about the effect of increasing the MPV on queue jumping and price discovery, the rule prohibits a venue not displaying the NBBO from filling an incoming order unless it can provide price improvement of at least $0.05 per share.
The rule specifically exempts, however, any trade executed at the midpoint of the NBBO, limiting its reach to subpenny stepping-ahead.
Notwithstanding widespread concern about the relation between MPV rules and stepping-ahead, the means by which a wider MPV facilitates queue-jumping remains an open question in light of the recent findings of Kwan, Masulis, and McInish (2014) (KMM). Using a proprietary dataset that classifies the trading venue for each reported trade in a dark facility, KMM examine the effect of the prevailing MPV rule on dark trading by exploiting an exception to Rule 612 for stocks priced under $1.00 where subpenny quotations are permissible. They find that the effect of increasing the MPV on a dark venue’s market share was localized entirely among trading venues classified as dark electronic communication networks 1 Queue-jumping can also occur without subpenny pricing where a trade occurs on a non-exchange venue at the NBBO. Because such trades occur at the same (penny) price increment as the NBBO, the minimum pricing increment should have no effect on their incidence. As discussed below, such trades can be problematic for brokers in light of their best execution obligations.
2 A trader seeking immediate execution will ordinarily obtain it by submitting a marketable order to be executed at the NBBO against resting limit orders.
3 Spurred by Congress to explore whether decimalization of stock prices harmed the liquidity of small capitalization firms, the SEC proposed a tick size pilot in 2014 to increase the MPV from a penny to $0.05 for a test group of securities. The agency’s longstanding concern with queue-jumping and price discovery, however, led it to create a subset of securities that would also be subject to a trade-at requirement.
To isolate the mechanism by which the prevailing MPV rule affects the incidence of dark trading and the supply of displayed liquidity, we turn to a regression discontinuity (RD) analysis of subpenny trading at the $1.00 cut-off established by Rule 612. We hypothesize that a primary channel through which the MPV rule affects the incidence of dark trading is through its effect on midpoint trading rather than through stepping-ahead. Specifically, we conjecture that for small orders investors generally prefer to buy (sell) securities through posting limit orders at the bid (ask) on exchanges’ displayed limit order books rather than taking liquidity through placing marketable orders and, consequently, paying the spread.5 As emphasized by KMM, however, where spreads are constrained by the penny MPV, investors seeking to place aggressive buy (sell) orders will be forced to join long queues at the national best bid (ask) given the price-time priority rules by which limit orders are filled on exchanges.6 It is in this environment that dark ECNs can compete with exchanges by offering midpoint pricing that enables investors to post nondisplayed midpoint orders that execute against marketable orders in the venue.
Although turning to dark ECNs raises execution risk, the significant price improvement offered by midpoint pricing combined with the nontrivial execution risk of posting orders to an exchange should result in smart-order routing protocols first checking dark ECNs for midpoint liquidity. In contrast, when quotes are no longer constrained by a penny MPV, the finer pricing increments will more accurately reflect heterogeneous pricing among investors on public exchanges, which both lowers spreads and shortens quote queues at any single price point, including the NBBO. Both effects should reduce the attractiveness of using nondisplayed midpoint orders relative to placing displayed orders on exchanges.
Using TAQ data from 2011-2013, we provide evidence that the current MPV rule has precisely this effect on the incidence of midpoint trading in dark venues. In particular, we show that the probability of midpoint trading in a dark pool at the $1.00 cut-off reveals an approximate 11% discontinuous increase as 4 We interpret KKM’s definition of “dark ECN” to include venues that match market orders to posted limit orders and that match market orders with other market orders. As such, we treat dark ECNs as equivalent to so-called “crossing networks” managed by broker-dealers. KMM’s concept of a “dark ECN” should also be distinguished from a conventional ECN (such as Lavaflow or Bloomberg Tradebook). Conventional ECNs display to the public the best available quotes of their subscribers either directly (e.g., through Finra’s Alternative Display Facility) or indirectly through incorporation into a formal stock exchanges’ displayed liquidity. Through 2014, for instance, Lavaflow published its quotations as part of the quotation feed of the National Stock Exchange. Dark ECNs, in contrast, do not display to the public the price or depth of any limit orders submitted by their subscribers.
5 We ignore for the moment the effect of exchange access fees and rebates, which can also induce investors to trade by means of posted limit orders on exchanges. We address these fees and rebates in Section 5. Additionally, we focus here on small orders to put to the side concerns about price impact. Where an investor trades a large order (e.g., a block trade), concern about revealing its trading interest before consummation of the full order may induce an investor to use a dark pool for reasons having nothing to do with the size of the MPV. Because large orders by institutional investors are commonly broken apart into smaller “child” orders for execution, small orders can originate from both institutional and retail investors.
6 This effect is also driven by the ban on locked markets contained in Rule 610(d) of Reg NMS. For instance, if the NBBO is at