Dark trading and market quality

Moderate levels of activity in UK dark pools appear to support market quality in equities.

Changes in regulation, coupled with advances in tech, have led to a proliferation in new classes of trading venues. One of the most important are ‘dark pools’ – where orders with no pre-trade transparency are executed. In other words, where there is no visibility of the order book.

Dark pools attract big institutional investors for a number of reasons. Not only do they offer firms the opportunity to execute large trades while avoiding significant price impacts.  In most cases they provide better prices than lit venues. 

For this reason, dark trading has become increasingly important in Europe and the US. In European markets, the total value of dark trades stood at more than €898 billion for the 12-month period ending March 2014. Representing more than 9.6% of all equity trades in Europe.  In the US, dark trading was estimated to be around 31% of consolidated volume in 2012. 

These developments have raised questions about the quality of the price discovery process in markets where dark trading is prevalent. Global regulators and others have asked if dark pools reduce transparency. For instance, some have argued dark pools affect investors’ willingness to display quotes in lit venues, leading to a decline in overall market quality. 

Importantly though, no-one has really tested these claims. Existing studies have only looked at cases where the impact of dark trading on trading quality is measured for a single lit exchange. No research has looked across the whole UK market. 

To fill this gap, we aggregated data from the UK’s four main trading venues – the LSE, BATS Europe, Chi-X Europe and Turquoise, using high frequency data for 288 of the largest stocks in the London equity market. The stocks used are constituents of the FTSE 350, which consistently form part of the index over the five-year sample period (June 2010 – June 2015).    

This large-scale data grab is important because finance theory suggests that dark pools could be beneficial to market quality. 

On paper at least, you would expect ‘uninformed liquidity traders’ – an institutional investor, say, trading because of fund redemptions – to gravitate towards dark pools. Just as you would expect ‘informed traders’ – an investor with superior information based on expert analysis – to stick to lit markets. 

The former want to conceal their liquidity driven trades. The latter want the information in their trades out into the open and impounded in prices, ideally after they have traded on it. 

All other things being equal, this should result in a reduction in pricing errors and noise that you normally associate with uninformed traders firing their orders into lit markets, where they may be mistaken for ‘informed’ trades. Hence if uninformed traders select dark pools, and informed traders select lit venues, overall market quality improves.

But we all know too much of a good thing can be ultimately bad. Consistent with this notion, it would be reasonable to expect that when even informed traders start to trade in dark pools, or dark trading passes a certain threshold, market quality will be harmed (indeed there is research to support this idea from both the US and Australia). So it makes sense, in theory, that dark trading is beneficial only up to a point.

 

Dark trading is not detrimental to market liquidity up to around 15% of total trading - if measured in value.

Consistent with theoretical expectations, our results suggest that the level of dark trading we see in London at the moment does not appear to be harmful to aggregate market quality. In fact, for the full sample of stocks we assessed, we found that dark trading is not detrimental to market liquidity up to around 15% of total trading - if measured in value. Our estimates are uncertain but one thing is clear: current levels of dark trading are not detrimental to market quality.

Once MiFID II becomes operational, at the beginning of 2018, this value will not be allowed to exceed 8% for each stock. 
However, it should be noted that the threshold is sensitive to stock-level liquidity. In other words, it is significantly lower for very liquid stocks (approximately 9% or 10%) and far higher for less liquid ones (at around 35%). 

The intuition behind this result is clear. For liquid stocks, dark trading is unlikely to reduce adverse selection to a significant degree – particularly given that the information available on such stocks is likely to be commonly held by a large number of market participants. 

For stocks that are not very liquid and only trade sporadically, it is likely that investors trade in the dark because they want to reduce their price impact. Not because they have superior information that they do not want to show their hand on.

Note: The views expressed in this article are those of the author alone, and should not be taken as an official FCA position.

 

01/08/17