Questions and answers about
the economy.

Dark trading: what is it and how does it affect financial markets?

Dark pools enable an opaque form of trading in financial assets that has raised concerns among investors, brokers, exchanges and regulators. Detractors argue that the lack of transparency damages asset pricing in financial markets, while advocates claim that it expands access to those markets.

‘Dark trading’ is an anonymous form of financial exchange that is becoming increasingly mainstream. In the United States, the percentage of the value of trading executed ‘in the dark’ doubled between 2008 and 2012. In terms of volume, dark trading venues executed nearly 40% of transactions in US shares in April 2019.

Dark trades are facilitated by ‘dark pools’ – a growing class of platforms that do not offer pre-trade transparency. In other words, market participants, other than the submitter and the pool operator, are unaware of the existence of orders submitted prior to their execution. Traders do not have to make public either the price or number of shares of a dark order. But once executed (that is, the order becomes a trade), they must be made public in a timely fashion.

How have dark pools developed?

The proliferation of dark pools has been driven in part by a greater reliance on technology for trading in financial markets. It is also a response to changes in regulations, as regulators increasingly focus on investor protection and making financial markets fairer and more transparent. Efforts in this regard include enactment of the 2005 Regulation NMS (RegNMS) in the United States, and the 2007 Markets in Financial Instruments Directive (MiFID) in the European Union (EU).

Opaque market structures are not new. But prior to the 2000s, dark venues and structures had a combined low market share and were mainly used for block trading – the trading of large quantities of shares that meet a minimum per trade threshold (for example, a single trade of 10,000 shares). With the effects of technological advances and the implementation of regulatory interventions, dark trading has become mainstream.

Dark trading is growing in Europe. In European markets, the volume of trading executed in dark pools accounted for 9.1% and 9.6% of all on-exchange activity in April and July 2019, respectively.  These are significant levels of trading activity.

In 2018, the EU implemented a provision that imposes what is called a double volume cap (DVC) of 8% on stock-level volumes executed in dark venues over any 12-month period. This was introduced as part of the MiFID II package of rules. Despite these efforts, dark pools remain popular in EU markets.

Efforts to rein in dark trading activity are not limited to the EU. Australian and Canadian regulators have also introduced measures to reduce the volume of transactions executed in dark venues. These efforts suggest that regulators and policy-makers around the world have a dim view of dark pools.

But the evidence is not quite that clear cut. In reality, and based on emerging research evidence, the effects of dark trading on the quality of markets – the features that indicate how well they are functioning – are contextual.

Who is using dark pools?

In theory, a dark pool added to a market where the only previous market structure for trading is a ‘lit’ exchange (where order submission is transparent by comparison – think of the standard limit order book) will attract uninformed traders. This also results in a concentration of informed traders on the lit exchange. Uninformed traders are those who have no prior information of the value of the instrument that they are trading, unlike informed traders who do have this information.

Uninformed traders will gravitate towards the dark pool because their risk of being affected by having insufficient information compared with an informed trader is lower in a dark venue. On the other hand, informed traders – who are wary of the costs of delay in the execution of their orders in dark pools – will largely stay in the lit market.

Thus, traders self-select their trading venues based on how much information they hold, and this has implications for the risk of adverse selection. This is the risk of an uninformed trader trading with another trader who has more information. In this scenario, the uninformed trader will be likely to pay more or accept less money than is optimal for the asset that they are trading.

This has implications for associated characteristics of market quality, such as liquidity – the willingness of uninformed traders to trade in the first place. Uninformed traders will understandably be reluctant to trade when the probability of trading with a more informed trader is high. This reluctance drains the market of the liquidity that the informed trader needs to exploit the information they have – without an uninformed trader to take advantage of, information is useless to the informed trader and the market.

When informed traders trade with their information, they help the market to discover the ‘fair’ price for the asset they trade. Thus, in the context of dark trading, the two classes of traders self-selecting where they trade based on their needs has implications for overall price discovery in the whole market, comprising the lit exchange and the dark pool. This self-selection improves price discovery under normal conditions.

Research shows that volatility is a critical driver of the overall dynamics of self-selection into dark and lit venues for trading (Zhu, 2014). It also suggests that there is a variable relationship between volatility and the share of trading activity in dark pools. Specifically, at a sufficiently low level of price volatility – that is, in normal conditions – the proportion of trading in dark pools for a given asset will increase with volatility. But when volatility becomes excessive, trading in dark pools decreases as volatility increases.

This variability is driven by the pattern of informed and uninformed traders selecting where they trade, but only when market conditions are normal. In other words, it holds when volatility is moderate and the spread between the ask and bid prices on the exchange is narrow. Under these conditions, uninformed traders gravitate towards the dark pool because they face lower risk of adverse selection there.

At the same time, informed traders concentrate on the lit exchange because the gap between the price asked by the seller and the price at which the buyer is willing to pay – the exchange spread – is not excessive. In this case, the cost of execution risk in the dark pool is greater than the benefit of potential price improvements it may offer. For example, regulations in Australia and Canada require that the price at which regular-sized orders are executed in dark pools be better than on a lit exchange.

This dynamic changes once volatility in the exchange exceeds the maximum level needed for informed traders to avoid the dark pool. In this scenario, informed traders start to migrate to the dark pool in search of uninformed counterparties with whom to trade, and in an effort to avoid the widening exchange spread.

So, when uninformed traders of whom they could take advantage are scarce in the lit market, informed traders may start entering dark pools in order to reduce their transaction costs, due to wider spreads, and to increase their profits (see Hendershott and Mendelson, 2000; Nimalendran and Ray, 2014).

The informed traders’ migration to the dark pool would result in uninformed traders leaving the erstwhile safety of the dark pool for the lit exchange. This would, in turn, lead to an overall loss of trading activity in dark pools and a net gain by lit exchanges.

What are the effects of dark trading?

New research examines the Covid-19 pandemic as a case study. The authors explore the impact of the Covid-19 shock on financial markets and regulatory restrictions on dark trading to investigate whether the predicted dynamics of when informed and uninformed traders engage with dark pools hold (Ibikunle and Rzayev, 2022).

This work finds that high levels of volatility on lit exchanges are linked to an economically significant loss of market share by dark pools to lit exchanges, as predicted by theory. The implications for market quality of the net loss of market share by dark pools to lit exchanges during periods of high volatility are mixed.

Generally, the shock of Covid-19 on financial markets negatively affects liquidity – in other words, the ability to trade large quantities of assets promptly and with little or no impact on the price. But this negative effect is statistically and economically significantly lower for shares that are traded on both lit exchanges and dark pools in comparison with similar shares traded only on lit exchanges because of regulatory restrictions.

Conversely, the dark trading makes the loss of efficiency in the price discovery process worse than it would have been had a share been traded only on a lit exchange.

Recent evidence on the effects of dark trading during normal trading conditions – low volatility and narrow exchange spread – appears more positive. Although an increase in dark trading is associated with an expected loss of liquidity by lit exchanges (see, for example, Comerton-Forde and Putni?š, 2015; Foley and Putni?š, 2016), some research suggests that the liquidity and efficiency of the price discovery process improves for the overall market, comprising both lit exchanges and dark pools.

Further, analysis of a sample of 288 of the largest UK shares being bought and sold across trading venues in London investigates the effects of dark trading on characteristics of market quality (Ibikunle et al, 2021). The results show that the market benefits when dark trading occurs at low to moderate levels.

As dark trading increases, the risk of adverse selection recedes for uninformed traders. This is then associated with an improvement in overall market liquidity. It appears that when uninformed traders can trade in the dark, where they are more likely to avoid being targeted by informed traders, they are less reluctant to submit orders to the market.

So, dark pools encourage the provision of liquidity that otherwise would not have been offered in a world where they do not exist. The increased trading activity driven by the availability of dark pools dilutes the proportion of informed trading in the overall market, leading to a fall in the risk of adverse selection faced by uninformed traders.

The migration of uninformed trading volume to dark pools is also linked to a reduction in the noise observed in the price discovery process and an improvement in informational efficiency in the market as a whole.

Specifically, since uninformed trading activity is responsible for much of the excess variability in price as displayed on lit exchanges, when they migrate to dark pools – which neither display order prices nor acknowledge the existence of orders – the price displayed on lit exchange becomes less noisy.

This makes it easier to observe the fair price for a tradable asset. As most dark pools (for example, in Europe) execute orders in line with the price displayed by lit exchanges, the efficiency of the price discovery process improves for the market in aggregate.

But this work also shows that the relationship between market quality characteristics and dark trading varies (as predicted by Zhu, 2014 and reported for an Australian sample by Comerton-Forde and Putni?š, 2015).

This implies that at higher levels, dark trading could harm characteristics of market quality, such as liquidity and price discovery. Using the level of adverse selection risk faced by uninformed traders as an inverse indicator of market quality given its close alignment to both liquidity and price discovery, the impact of dark trading on market quality is estimated to be about zero when the percentage value of dark trading in the overall market is roughly equal to 14% (Ibikunle et al, 2021).

Essentially, the initial effect of dark trading on market quality characteristics is positive and continues to be so as the value of trading conducted in dark pools increases as a percentage of the overall market value comprising all trading venues, including dark pools and lit exchanges. But when dark trading value is at about 14% of total market value, an inflection occurs and the effect of dark trading turns negative – and this continues as the value climbs higher.

But this estimate varies depending on the level of trading activity across shares. When the 288 shares in the study sample are split into quintiles, the thresholds for the quintile 5 and 4 (largest trading) shares area around 9%, while the estimated thresholds for quintiles 3, 2 and 1 are 14%, 25% and 23%, respectively.

The disparity across the quintiles appears to be indicative of the extent to which shares with different trading activity rely on transparent and opaque trading venues. Quintiles 1 and 2 shares, those with lower trading activity on the London Stock Exchange, historically execute a high proportion of their trading volumes via broker-dealer arrangements that are not dissimilar to dark pools with respect to pre-trade transparency requirements. As a result, the tolerance of these shares for opaque (dark) trading is higher on average than those of shares with higher trading activity levels, which almost exclusively trade on lit exchanges.


Recent regulatory efforts emphasise investor protection, transparency and fairness, all of which are served by the enhancement of liquidity and efficiency of the price discovery process.

Based on the evidence from recent studies (for example, Ibikunle and Rzayev, 2022), the goal of these efforts is furthered by dark pools operating alongside lit exchanges. It is important that policy-makers are careful not to eliminate the benefits of dark trading for market quality by arbitrarily imposing restrictions on it.

Recent post-Brexit regulatory actions by the Financial Conduct Authority (FCA), which indicate that the regulator would no longer automatically apply the DVC requirements as part of its regulation of equities, are consistent with this view.

Where can I find out more?

Who are experts on this question?

  • Carole Comerton-Forde
  • Gbenga Ibikunle
  • Talis Putnins
  • Khaladdin Rzayev
  • Haoxiang Zhu
Author: Gbenga Ibikunle
Picture by serts on iStock
Recent Questions
View all articles
Do you have a question surrounding any of these topics? Or are you an economist and have an answer?
Ask a Question
Submit Evidence