Magazine Issues » December-January 2015

EXCHANGES: Dark, lit or just dusky

Dark cloudsHigh profile lawsuits, a best-selling expose and unwelcome regulation have brought dark pools under the spotlight. Nicholas Pratt discovers some buy-side firms are calling for a return to basics.

It has been a turbulent year for Europe’s equity markets, not only in terms of volatility, but also for market structure and dark pools. These off-exchange venues were originally designed to allow buy-side traders a place to transact large block orders anonymously and avoid costly market impact. 

But with so many dark venues operating at the same level of anonymity and with often different participants, confusion is taking hold, if not at the professional trader level, then certainly among senior management, politicians and regulators.  

The new Markets in Financial Instruments Directive (MiFID) II rules are committed to implementing controversial volume caps on the use of dark pools, despite the fact that Markus Ferber, the European Parliament rapporteur on MiFID II, has expressed misgivings and David Lawton, director of markets at the UK’s Financial Conduct Authority (FCA) has called them a “clunky” policy measure.  

The caps will limit monthly trading on dark pools to 4% per venue and 8% across the region based on total volume of trading for a financial instrument. The problem is that no one knows how the volume caps will be enforced. At present no consolidated tape exists in Europe, a facility that would show all the trade dates on various execution venues in Europe and which is essential to measuring volumes. 

“Regulators are between a rock and a hard place,” says J.P. Urrutia, European general counsel for ITG, which operates Posit, a block crossing network aimed at buy-side firms looking to execute large orders.

“Without consolidated tape there will be no transparency on how the dark cap data has been collected. Different regulators may have different approaches and this could lead to mudslinging at Esma [European Securities and Markets Authority] about the quality of data. However, a consolidated tape will not be available for some time and if the EU has to end up building a utility because the tender process does not provide a viable commercial candidate, I don’t know where it would obtain the money for it.”

A simpler and more effective solution, says Mark Pumfrey, head of Liquidnet in Europe, Middle East and Africa, would be to impose dependency on creating value over and above lit – or non-dark – markets. “This means size and/or price improvement, a rule which has worked successfully in Australia and Canada. Unfortunately, this approach is not on the agenda for EU regulators who are already facing a tight deadline to meet their MiFID II requirements.

The central issue is around trust. Because of the pre-trade anonymity within dark pools, transparency between the dark pool operators and their members is very important. “There are inherent conflicts of interest in some dark pools that can create investor-confidence issues so clarity around who the counterparties are and the rules of engagement are very important,” says Pumfrey. 

Members’ interests have been highlighted by a pending court case in the US brought by New York attorney general Eric Schneiderman against Barclays, which alleges that the UK bank misled investors about the level of high frequency trading activity in its anonymous dark pool, LX. For its part, Barclays has robustly defended LX and vowed to “introduce evidence demonstrating how far off base these allegations are”.

For those operating lit markets, the world of dark liquidity is simply not transparent enough. “Right now with equity dark pools, it is hard to tell what is dark, lit or dusky,” says David Mercer, chief executive of LMAX Exchange, a multilateral trading facility for foreign exchange trading aiming to bring more transparency through exchange style execution to an over-the-counter market that has, effectively, been “dark” for many years. “About the only transparency you have around a dark pool is when it launches. I operate a lit pool and its main job is to create liquidity. I don’t know how you do that in a dark pool. How do you know that the other side of the trade has the requisite liquidity?”

Mercer is a believer in rulebooks rather than volume caps, as proposed in MiFID II. “As long as the rulebook is clear, you should be able to have any means of execution.” He also believes in open market access for execution venues, something which is not the case in many dark pools. “I think there needs to be clarity over membership.”

Mercer also believes that market impact and the predatory threat from high-frequency traders (HFTs) have been massively overplayed. “In a liquid market, market impact should not be a big issue. There are also an increasing number of algorithms available to slice and dice orders. 

“And a lot of HFTs are the best market-makers in the street. Similarly a lot of banks are the biggest HFTs. Everyone relies on HFTs for daily liquidity. People want benign orders with no HFT liquidity. That is wanting your cake and eating it.” 

Others within the dark pool market disagree. “You have to go back to the reason why dark pools were created,” says Liquidnet’s Pumfrey. “Exchanges were courting high frequency traders in order to boost volume and this meant that buy-side firms needed another market where they could trade at mid-point without wrecking an institutions’ share price. 

“And as institutions grow in size, market impact will only become more important. A book like Flash Boys highlighted some of the difficulties that buy-side firms have faced when trading and it has caused investors and pension funds to take notice. So the way managers trade will become a big part of RFPs [requests for proposals], alongside fund performance.”

Earlier this year 30 buy-side dealers formed a working group to address the issue of dark pools. “We are looking for a return to basics – simple rules of engagement but to do that at the venue level,” says Paul Squires, head of trading at Axa Investment Managers, one of the group’s founding members. 

The group, 43 members strong, recently met again with Turquoise, the dark pool operator owned by the London Stock Exchange, to discuss the October launch of its Black Discovery service, an offering that has been launched on the back of buy-side clients complaining about tiny fills and the fragmentation of venues. 

Turquoise’s plan is to launch a block crossing network that interacts with the random uncross timing of its dark pool. By consulting with the buy-side group, the access to this service becomes broker-agnostic– something that is important to certain dealers, says Squires. “We trust the venue insights of brokers. They have better resources than us and they are able to look at the market in aggregate but, of course, part of their routing logic means that they will cross orders internally or post passive orders on venues that will get them a rebate. We are under increased pressure to prove best execution so these concerns need to be addressed.”

Some of the group have asked a number of brokers to provide them with an algorithm that will not go through their smart order routers (SORs), that is relatively simple, and goes to Turquoise first. “It will define the minimum execution size (MES) eligible for the block discovery service and if there is a matching sell-side order, our broker will be given a block indication.”

Despite several of the brokers being approached to support this initiative, at the time of writing only four were able to meet the request – Barclays Capital, JP Morgan, Instinet and Societe Generale. “That was quite telling considering there are 43 buy-side firms involved,” says Squires. 

He believes that the only way to return to a simpler trading model is for the buy-side to impose more control. He cites rules in Hong Kong as well as France that buyers have to demonstrate a working knowledge of the algorithms they use. And buy-side firms are taking initiative, sending out more questionnaires to their brokers and getting involved in initiatives, such as Turquoise. 

In the first month of trading, the Block Discovery service has produced an average trade size of €132,000, as opposed to €16,000 in the normal pool. What has been disappointing is the small size of the fills which is in part due to the small number of brokers involved and the fact that the MES is only applied at block indication level rather than for actual executions.

Nevertheless, the results from the Block Discovery service and its bespoke algorithms will be used to give buy-side users an ongoing monthly guide to help them assess the quality of execution they receive elsewhere and what constitutes an acceptable level of trades that brokers cross in their own internal pools, says Squires. “This is what the regulators want in terms of taking more responsibility for our execution.”

©2014 funds europe