There is no evidence that dark trading at the level of the caps in new European Union regulation harms market quality according to research published by the UK’s Financial Conduct Authority.

The FCA said this month in an occasional paper, Aggregate Market Quality Implications of Dark Trading, that dark trading at current levels does not appear to be harmful to market quality in the aggregate UK equity market.

MiFID II, which comes into force in the European Union at the beginning of next year, places double volume caps on dark pool trading. The caps on dark pool trading are 4% on any single venue and 8% market-wide in any 12-month rolling period with exemptions for Large In Scale (LIS) trades above a certain size specified by the European Securities and Markets Authority.

The FCA paper concluded: “Our results imply that there is a threshold at which dark trading may start to negatively affect market quality. We estimate that threshold, for our full sample of stocks, to be when dark trading value is approximately between 11% and 17% of total trading by pound value, depending on the specific market quality attribute examined.”

The researchers said their study is the most comprehensive examination to date of dark trading in Europe, since they analyzed the universe of FTSE 350 stocks from June 2010 to June 2015.  The study examined the following UK trading venues –  the London Stock Exchange, Turquoise, Bats and Chi-X Europe (which have since merged) – to test competing theoretical predictions on the implications of dark trading on the quality of the aggregate market, comprising both the lit and dark sections.

They concluded there was no evidence of a negative effect of dark trading on market liquidity until dark trading value as a proportion of total trading value is more than 15%.

“The results also indicate that dark trading reduces adverse selection risk and noise in the price discovery process until it attains 16% and 11% of total trading value respectively,” said the paper. “Results also indicate that, when we adjust our measure of adverse selection risk to account for increased trading volume in a high frequency trading environment, that threshold increases to 17%.”

The study continued there was no evidence that dark trading at the 8% MiFID II cap is detrimental to market quality.

The paper added: “Based on analysis conducted using data from mid-point dark order books as well as lit order books, we find that the overall quality of the market improves at moderate levels of dark trading.”

The researchers commented that MiFID II emphasises investor protection and trading quality in financial markets.

“We have shown in this study that the latter aim is furthered by the existence of dark pools operating alongside lit exchanges,” said the paper. “It is important that policy makers take care not to eliminate the market quality benefits of dark trading by arbitrarily imposing uniform dark trading restrictions for all stock sizes.”

FIA EPTA, the body which represents European firms who trade with their own capital, questioned the timing of the study:

@TheFCA #mifid2 #darkpool @ESMAComms Not clear on why this DP study wasnt done before. Random choice it was 8% cap. https://t.co/6Luhcxdbc4

— FIA EPTA (@FIAEPTA) August 1, 2017

Last month the inaugural research paper from Plato Partnership, the not-for-profit industry group representing asset managers and broker dealers,  examined  the potential impact of double caps on dark trading and the closure of broker crossing networks under MiFID II.

Professor Carole Comerton-Forde, Professor of Finance at the University of Melbourne and author of the paper, warned that the buyside will have to make choices on how to achieve best execution given the potential fragmentation of block trades across multiple venues. In addition, the algorithms and smart order routers will have to adjust to  dark trading being shut down in certain stocks when the caps are reached. The FCA research found that 8.35% of the sample had dark trading in excess of the 8% cap.

The full FCA research paper can be read here.