Finding Best Execution in the Dark*

by Esmeralda Lyn, Ph.D.

Over the past three decades, the equities markets all over the world have seen periods of fragmentation and consolidation as new players enter the industry, or two or more entities combine to unify liquidity pools.   Despite major industry mergers, traders currently have more venues to send order flow than ever before.  These include the primary exchanges, electronic communications networks (ECN), alternative trading systems (ATS), and crossing networks which are opaque venues more commonly known as “dark pools”.  There are currently over 30 dark pools in the United States today. None of the trades of dark pools are reported to the consolidated tape keeping the marketplace as a whole unaware of the volume being executed.

The market share of dark pools has doubled from 5% to 10% from 2008 to 2010 and has intensified since then. Rosenblatt Securities reported that dark pools are getting a larger piece of a shrinking pie since trading volume in U.S. equity has declined to the lowest level since 1999. Today, as much as 40% of trading in the U.S. equities takes place away from public stock exchanges.   Ironically, more routing options have not necessarily made attaining “best execution” any easier.  Both regulators and investors have struggled not only to define best execution but also to create strategies and rules to establish a fair and transparent environment in which to conduct business.

All else being equal, the most favorable price for the customer at the time a stock is traded appears to meet the definition of best execution. When developing its regulations for the marketplace, the SEC had decided to use price to distinguish the good from bad, but there are other factors or costs which affect the quality of entering or exiting a position. The more sophisticated the investor, the more the execution price becomes only one aspect of execution quality. The explicit costs include brokers commission while the implicit costs include the spread, price impact and opportunity costs, which are much larger than explicit trading costs. This is where the concept of implementation shortfall is so important – i.e. to look at the change in price from the time a trading decision is made until the trade is actually completed.

 Dark Pool Structures

Not all dark pools operate in the same ways.  One primary differentiation is between dark pools set up by sell-side firms and those set up by independent providers of trading systems. Sell-side firms have been matching internal order flow for years, but the advent of new technology and increased electronic trading have greatly increased volume.  The rationale for sell-side firms is simple: sending order flow to the different exchanges and ECNs costs money and cuts into their commission margins.  If they are able to pre-match their different customers’ orders against each other and against the firm’s proprietary trading desks before sending the orders on to the open market, they are able to eliminate the need for a market maker to provide liquidity and, thus, they avoid paying for execution. In theory, this method also works to the customer’s advantage—there is speed of execution and minimal market impact.  Many sell-side firms, including almost all the bulge bracket broker dealers, now offer this as a standard.

A second differentiator among dark pools is how frequently stock is crossed.   Some dark pools only look to cross at set scheduled intervals over the course of the day.  ITG’s Posit is an example of a scheduled crosser.  The NASDAQ also offers a crossing mechanism where buyers and sellers enter orders, which the system looks to cross if the other side is available at the scheduled time.  Other newer dark pools, such as Liquidnet and Bank of America Merrill Lynch MLXN, look to cross continuously throughout the day.  Some dark pools only allow buy-side participants while others allow both buy and sell-side traders.

Types of Orders       

Dark pools must also decide what types of orders they will accept.  The largest decision is between committed and uncommitted liquidity.  Committed liquidity is a method used by dark pools to police their community of investors and prevent market manipulation.  Committed liquidity will be executed without any notice to the investor once it is sent to the dark pool.  With uncommitted orders, the investor is notified that the other side exists and must confirm that she wants to trade.

In pass-through order types, sell-side firms are sending their liquidity through their internal pools before sending them on to the traditional market venues.  This order type also encompasses dark pools that constantly “ping” the pools with what amounts to Immediate or Cancel orders.  The least aggressive order type some dark pools employ involves some amount of human negotiation.  Essentially, indications of interest are expressed anonymously over the network and, if a counterparty exists, some further human interaction is necessary in order to trade.

Costs and Benefits    

While the benefits to the existence of dark pools are relatively self-evident, the costs are subtler and some are definitely unintended.  The main benefits to trading via a dark pool are the ability to execute anonymously blocks of stock with very little direct market impact.  Information leakage can significantly cut into trading profits when the market is able to ascertain the direction of order flow.

Dark pools by their nature involve counter-parties submitting orders in the hopes that the other side exists and is willing to trade.  The main costs to the investors all relate to the construction of that marketplace (namely, the smallest price movements called tick prices).  The time costs can be significant if the other side is not looking to trade.  Along with time costs are the obvious opportunity costs associated with not trading.  When analyzing implementation, traders are forced to balance the benefits that dark pools offer against the potential costs of not executing stock.  Relating directly to this issue is the number of successful fills.  Compared with a traditional upstairs marketplace, dark pools have a higher percentage of unfilled orders.  This can be attributed to the lack of negotiation room plus not seeing all available liquidity. Either way, this is a definite cost to trading.

Due to the increasing number of dark pools, liquidity at any given one varies considerably. If a trader wants to buy or sell large amounts of a stock, she/he may need to submit orders to multiple dark pools.  In addition to the time and resource costs related to spreading orders across dark pools, the direct costs include ticket costs if fills come back from a number of venues.  There is a high probability that such a strategy will result in the trader missing liquidity.

Although it is likely the US equities market will remain fragmented, dark pools will probably see the same wave of consolidation that ECNs have seen in recent years. The real question is, will investors be able to efficiently utilize all the tools at their fingertips?   In addition, will the existence of dark pools help or harm price discovery? The evidence from existing empirical studies is inconclusive although there is preliminary evidence that the existence of dark pools reduces the noise in exchange order flow and, thus, may improve price discovery.  Liquidity could also be a problem.  Although buyers and sellers are matched in a dark pool at some agreed upon price, for there to be liquidity, buyers and sellers have to participate in dark pools.

Regulation of Dark Pools

As technology, trading styles and volumes in the equities markets have changed, the SEC is continually challenged to fulfill the original mandate of the Securities Act of 1934 of protecting investors, maintaining a fair, orderly and efficient markets, and facilitating the formation of capital. The National Market System was instituted in 1975 and Regulation NMS in 2005 to provide much greater transparency to the marketplace.  The rationale was simple: if investors could see all current quotes and then could route to the venues listing those quotes, all would be protected.  Although the marketplace would still be fragmented where different trading venues compete for order flow, best execution would be secured by routing to the venue with the best pricing.  The crux of the regulation was that best execution and investor protection focused solely on price.

As they have grown, dark pools have been subjected to even more regulation.  This is especially pertinent for larger pools.  If they are executing more than the SEC- prescribed 5% threshold of the daily volume in a name, they must provide open access to all market participants.   These pools would be forced to display their quotes, which would negate most of the inherent benefits of being a dark pool.  How this regulation plays out will greatly affect dark pools in the years to come.

Two recent releases from the SEC simultaneously demonstrate the Commission’s concerns that the benefits of dark pools to market participants are preserved and are consistent with its regulatory objectives. While the changes proposed in these documents have not yet been implemented, the SEC’s sensitivity to issues associated with market fragmentation is clearly evident from the releases, wherein the commission states that among its regulatory goals are efficiency, price transparency, best execution of investor orders, and order interaction. Within the topic of undisplayed liquidity, the Commission invited comments on three issues in particular:  the effect of dark pools on order execution quality, the effect of dark pools on public price discovery, and fair access to sources of dark pools.

Where do we go from here?

Do dark pools undermine market functioning as a whole?  Fragmentation has made it increasingly difficult to execute large blocks of stock and dark pools have certainly exacerbated this situation. One thing is certain, dark pools benefit from the price discovery that the marketplace freely provides them.  Since there are no published quotes from a dark pool, it relies on other investors to trade freely in the traditional marketplace to price stocks.  The exchanges make no money but the dark pools benefit by not having to make a market. In highly liquid stocks this may not be an issue, but in thinly traded issues losing twenty to forty percent of the volume to a dark pool obviously affects the way a stock trades. Another important issue is the growing number of separately organized dark pools, which increases search costs for market participants to find hidden liquidity.  Regulators and policy makers, here and abroad, have launched various committees to address issues of whether trading in the dark is consistent with the spirit of a free, open and efficient marketplace. Canada now requires dark pools to offer a better price compared to the exchanges or limit their trades to large blocks.  Since then, the sharp price swings in stocks have been reduced significantly.  Is this the model that the U.S. should follow?

It is impossible to discount the massive effect technology has had upon the global equities markets.  Technology has changed the role of the modern trader as they not only seek liquidity with minimal market impact, but also are forced to comprehend and react to an ever-changing trading landscape. While the names of the participants may change, it appears the U.S. and the rest of the world will be dealing with market fragmentation for at least the near to mid- future.  Finally, we must understand that dark pools do have an important and useful role for investors but they are not a panacea for all traders’ execution problems. The SEC is clearly amenable to preserving the benefits of dark pools, if it can be done consistent with promoting other regulatory goals including market transparency and improved price discovery.  It is also clear that the equity markets will continue to evolve and innovation will be introduced.  Trading volumes in different markets will determine which innovation should survive, and dark pools seem to be winning.

*The long version of this paper was originally published in the Journal of International Business and Law, volume 12, issue #1  “Finding Best Execution in the Dark: Market Fragmentation and the Rise of Dark Pools,” by Edward Eng, Ronald Frank and Esmeralda Lyn.