Does dark trading hurt market quality? It is a question that has vexed policymakers for some time, and has attracted renewed focus recently following certain exchange initiatives to establish non-displayed trading pools for retail orders. Understanding the relationship between dark liquidity and market quality has become central to the debate on market structure as authorities around the world consider revisions to their respective regulatory frameworks. Measures to support fair competition between displayed and non-displayed trading venues should be the focus of those efforts.

Over the past decade, the pace of evolution in equity markets has accelerated as a result of the combined forces of technology, regulation, and globalisation. Consequently, average trade sizes have declined alongside significant increases in overall quote traffic and transaction volumes (largely owing to the adoption of systems that have improved operational efficiency and network capacity). Further, there has been a marked reduction in trading costs (both bid-offer spreads and commissions), and increasing fragmentation of liquidity across numerous trading venues.

Within this fragmented, decentralised environment, dark pools serve an important role for institutional investors, enabling them to achieve efficient executions through fewer transactions and anonymity. While dark pools have existed for decades in one form or another, their importance has perhaps grown as the public exchanges have in recent years become dominated by algorithmic and high-frequency traders.

The use of sophisticated algorithms to slice orders into small sizes and spread them across multiple venues to effect transactions is an established feature of today’s marketplace. Yet the existence of algorithms that work to detect larger blocks of liquidity that often sit behind ‘child’ orders and almost instantaneously adjust public quotes can lead to slippage, or adverse price movements. In such an environment, dark pools offer relative safety by enabling institutional investors to minimise market impact.

By restricting access to undesired market participants (such as high-frequency traders), and by not revealing orders, dark pools enable investors to minimise their information leakage and realise more efficient executions. Moreover, dark trading facilities provide the possibility of price improvement and reduced transaction costs by crossing orders at the midpoint of the quoted best bid and offer prices, thereby saving on both the bid-offer spread and on exchange fees.

Given these circumstances, it is easy to see why dark pools, or non-displayed trading more generally, continue to grow in popularity. Non-displayed trading away from public exchanges, including broker/dealer internalisation and dark pools, is estimated to account for nearly one-third of consolidated volume in the US – growth of almost 50% since the start of 2009. In Europe, turnover in dark pools doubled between June 2010 and October 2012 and almost tripled as a proportion of order book activity.

However, from a broader market integrity perspective, this growth in dark trading raises certain issues, such as concerns over transparency. Investors may prefer to not reveal their order, but benefit from collective transparency even though they may not wish to contribute to it individually. However, if no investors revealed their orders, in other words if all trading took place in non-displayed pools, all investors would suffer, as market quality would deteriorate with the evaporation of displayed liquidity.

Other issues include considerations over fairness, particularly in relation to access among different classes of investors to these non-displayed trading venues and the basis on which they compete with traditional exchanges. Moreover, some investors have raised concerns that the incentive to display orders in public markets is being undermined by certain off-exchange trading practices, particularly in the US, where over-the-counter market makers fill retail orders ahead of displayed limit orders by offering price improvement in fractions of a cent – so called sub-penny trading. In turn, these concerns have implications for public price discovery and the quality and integrity of markets.

Therefore, it is desirable to obtain a competitive equilibrium between trading in displayed and non-displayed venues that balances the need for investors to transact efficiently in dark pools whilst not harming overall market transparency and integrity.

To better understand the relationship between dark trading and market quality, CFA Institute conducted a study of dark pools and internalisation based on a sample of US stocks stratified across listing market and market capitalisation. For each stock, data on bid-offer spreads, top-of-book depth, dark pool and internalised volumes, and other variables were obtained over the period from the first quarter of 2009 to the second quarter of 2011.

Bid-offer spreads and depth, respectively, were regressed on internalisation and dark pool volumes and other explanatory variables. After controlling for factors known to affect spreads and depth, the study found that increases in internalisation and dark pool trading activity are initially associated with declining bid-offer spreads and increasing depth – that is, improving market quality. However, beyond a certain threshold, the relationship reverses, such that market quality initially improves but then declines as dark trading increases. Specifically, the results suggest that when a majority (over 50%) of trading in a stock occurs in non-displayed venues, market quality deteriorates.

This outcome could be attributable to the effects of competition and adverse selection. Initially, the introduction of dark pools (moving away from exchange monopoly) might cause more aggressive quoting in the public markets to compete for order flow alongside the dark pools, at least for liquid stocks. The tighter the spread quoted by market makers, the smaller the savings to be made from the alternative of a mid-point execution in a dark pool. But when dark pools dominate, investors could become disincentivised from posting displayed orders because of the reduced likelihood of those orders being filled on acceptable terms – particularly if the majority of retail order flow is internalised, which increases the risk of adversely selecting a more informed counterparty on exchanges. As a result, investors could quote less aggressively in the public markets, resulting in widening spreads and deteriorating market quality.

Overall, these findings suggest that market quality is best served by maintaining strong competition between trading on lit and dark venues, and avoiding a predominance of dark trading. Policy measures should be designed to protect the incentive to display orders in public markets while offering the possibility for meaningful price improvement to investors executing away from exchanges, maintain fair competition by ensuring consistent regulations between similar types of trading venues, and support transparency by ensuring consistent and consolidated trade reporting.

More fundamentally, these measures would enhance market integrity and underpin investor confidence in the equity market structure.