How to Catch a Black Cat in a Dark Pool

June 2011 Issue
How to Catch
a Black Cat
in a Dark Pool
or five things you wanted to know
about Dark Pools but hesitated to ask
How to Catch a Black Cat in a Dark Pool
ATMonitor Commentary
This is not an academic paper on theoretical discussions but rather a series of practical questions and
answers that members of MyATMonitor have asked and industry experts answered. Our primary goal is to
bring knowledge that will be useful to traders on the buy side. In fact, this philosophy is well reflected in the
very heart of MyATMonitor, a reliable, independent and trusted peer-group network of and for buy-side only
institutional traders.
This publication has been compiled from ongoing Q&A activity on the MyATMonitor Expert Panels. At the
time of publication, the Expert Panels on MyATMonitor are Dark Pools, Commission Sharing Arrangements,
EMS/OMS Relationships, Fragmentation of Liquidity, MiFID II and Transaction Cost Analysis and Best Execution.
The ATMonitor team would like to thank all members and experts that have generously contributed to the
success of MyATMonitor.
ATMonitor Team.
How to Catch a Black Cat in a Dark Pool
ATMonitor Commentary
Experts Panellists (in the order of appearance):
Chris Jackson is head of EMEA Execution Sales at Citi. He is responsible for sales
of the firm’s electronic trading and execution platform to European Institutional and
Hedge Fund clients. He joined the firm in September 2009 from Merrill Lynch where
he was latterly responsible for sales of the electronic, portfolio and transition trading
product to European clients. Prior to taking that role, Chris spent three years with
Merrill Lynch in New York where he was responsible for sales of the international
portfolio trading business. Chris joined Merrill Lynch London in 1997 having started
his career at SBC Warburg.
In addition to his role at Citi, Chris is also co-Chair of the FIX committee in Europe.
Natan Tiefenbrun joined LSEG in 2009 to develop a pan-European dark pool, and
became commercial director of Turquoise following LSEG’s acquisition of a majority
stake earlier this year. At Turquoise he is responsible for product development, sales
and marketing. Prior to that, he spent 13 years at Instinet, a pioneer of electronic
trading, latterly as President of the Asian & European businesses. Over the years
he was responsible for the conception & delivery of Instinet’s global EMS platform
for single-stock, portfolio & algorithmic trading, the after-hours crossing network,
transaction cost analytics, and for servicing the needs of quantitative investors from
indexers through to high frequency traders.
Sal Rodriguez joined Citi’s London Electronic Trading Team in November 2010. He
is head of EMEA electronic sales trading with responsibility for managing the team of
sales traders that manage electronic flow from the firm’s Institutional, Hedge Fund
and Wholesale clients. Prior to joining Citi, Sal spent time with UBS and Goldman
Sachs in London as an electronic sales trader. He started his career at Morgan
Stanley in London where he spent 10 years as a pan European cash sales trader.
Charlie Guy is part of the Chi-X Europe business development team. He joined
Chi-X Europe in May 2008 from the Royal Bank of Scotland. Chi-X Europe is the
largest European equity exchange by number of trades and the 2nd largest by value
traded, according to the Federation of European Securities Exchanges.
The content of this report is provided for informational purposes only and has been obtained from sources believed to be reliable,
but it is not necessarily complete and its accuracy cannot be guaranteed. It is not intended as an offer or solicitation for the purchase
or sale of any financial instrument or as an official confirmation of any transaction. Moreover, this material should not be construed
to contain any recommendation regarding, or opinion concerning, any security. Any views expressed in this report are those of the
individual experts and do not necessarily represent the official view of ATMonitor or participating organisations.
How to Catch a Black Cat in a Dark Pool
ATMonitor Commentary
Editor’s Choice: Selected Questions and Answers from ‘Dark Pools’ expert panel.
Which dark pools are most effective in generating larger
sizes of executions?
Chris Jackson, Citi
All dark pools are capable of delivering large size executions. We’ve found however, that block orders will
tend to reside more in dark venues that have a reputation for lower toxicity—consequently, knowledge
of the large order is less likely to leak into the market. In our experience, dark pools that can place
restrictions over the type of counterparts allowed to interact with the pool are likely to be less toxic
and so attract larger blocks. Unlike MTF dark pools, Broker Dealer pools are able to restrict access in
this way—consequently, a Broker dark pool will often attract larger blocks. At Citi, we strictly control the
nature and type of counterparties that we allow to access the pool giving us a very low level of toxicity.The
average resting order size we see in our dark pool Citi Match is in excess of $230k (April, 2011).
Natan Tiefenbrun, Turquoise LSE Group
Several factors determine trade sizes achieved by different dark pools:
1. Periodicity/Frequency of matching—dark pools that aggregate liquidity at a point in time tend to result
in larger trades than continuous matching models.
2. Proximity to the “parent order”—dark pools in which the buy-side participate directly tend to result in
larger trades than those which receive their order flow from the algorithmic/smart-routing infrastructure
of brokers.
3. Whether orders must be firm—dark pools which allow non-firm orders (e.g. they will not match
without a further confirmation from the owner) can receive larger orders (since the same order can be
simultaneously represented elsewhere).
What is ‘toxic’ liquidity in dark pools?
Natan Tiefenbrun, Turquoise LSE Group
Brokers often evaluate dark pool executions based upon whether the trade price looks attractive 3-30
seconds after the fact. If trades in dark pools subsequently look unattractive, then perhaps they would have
been better not participating in that pool. By comparing the results from different dark pools, brokers can
identify those in which they are more susceptible to gaming or adverse selection. Turquoise works with
brokers to help them measure these effects, to understand their causes, and to mitigate them.
Chris Jackson, Citi
A trade in a dark pool against a toxic source of liquidity would tend to result in price action against
you—the toxic counterparty you have traded against would tend not to have natural inventory but to
opportunistically post orders in the pool (often both buying and selling) in order to find out or “fish” for
genuine order flow. Once a genuine order is identified, the counterparty may use a number of aggressive
How to Catch a Black Cat in a Dark Pool
ATMonitor Commentary
strategies that seek to take advantage of knowledge of the order. Dark pool providers may turn a blind
eye to toxic or predatory liquidity since it creates the appearance of additional liquidity in the pool whilst
any negative price impact may be more difficult to identify. Dark MTFs cannot exclude a counterparty just
because it may be perceived to be toxic, Broker dark pools can (if they choose to measure it). Citi assesses
the toxicity of counterparts and seeks to exclude toxic ones.
How do I know my orders are anonymous when in a
dark pool?
Natan Tiefenbrun, Turquoise LSE Group
By ‘anonymous’, I presume the question is both “how do I know my order cannot be seen by anyone?” and
“how can I be sure that nobody knows my firm is active in a given stock?” In respect of MTF dark pools, the
FSA insists on strict segregation of data. At Turquoise, for example, the commercial oriented staff cannot
see the details of any open orders in the Midpoint Order Book. Whilst BCNs might not be subject to the
same rules, brokers surely cannot afford the reputational damage that would result from allegations of
information leakage. And enough staff move between brokers to ensure that any unethical practices would
soon be outed. So I would suggest that, if you’re dealing with a reputable firm, whether MTF or BCN, you
should assume your orders are indeed dark and your trading anonymous.
Salvador Rodriguez, Citi
Protecting the client and their order is something all good Dark Pools should offer as standard. There
are various mechanisms that brokers employ to ensure that this is the case. At Citi, we naturally take this
very seriously as we will do everything we can to protect our client franchise. Consequently, we have
geographical and technological segregation from other business areas to prevent any form of anonymity
being breached internally.We employ stringent anti gaming logic that protects the order once it is received.
We also restrict access to CitiMatch which means we have the ability to control and reduce potential
toxicity. We have total control of who is actually allowed into our pool. Nor will we advertise any order
we may have in CitiMatch. We therefore also protect from breaching anonymity externally. It’s easy to
measure. Any post trade analysis will show you whether or not there has been any leakage by looking at
the reversion numbers for that order. It is also easy to measure real time by looking at your execution
price and comparing it to the prevailing market price.
Why do some algorithms seem to get more fills in
dark pools than others?
Chris Jackson, Citi
A lot of factors at work here:
1) Liquidity profile in dark pool—some dark pools will have high pass-through volumes (small orders on
way to market), others will have more resting block liquidity.
2) Trading objective of algorithm—different algos look for liquidity in different ways—a VWAP algo will be
How to Catch a Black Cat in a Dark Pool
ATMonitor Commentary
less likely to take a block than an aggressive algo as it will introduce deviation and therefore performance
risk relative to the specified benchmark. Implementation Shortfall and aggressive algos will.
3) The Citi algorithms interact with dark liquidity with varying levels of aggression depending on the
trading objective—algorithms on passive settings will tend to rest passively in the pool maximising
spread capture—often capturing 80-100% of the spread. Aggressive algorithms like Dagger 4/5 will
potentially cross the spread in the pool to capture more liquidity.
Natan Tiefenbrun, Turquoise LSE Group
Agreed—it’s all about the algo’s strategy, the number of pools it accesses, and what parameters the algo
applies to orders. A more passive or block-oriented algo strategy might apply a higher MAQ (minimum
acceptable quantity) to orders, which would avoid small fills at expense of reducing the match rate. Similarly,
it’s possible in the Turquoise Midpoint Book to opt out of interacting with orders demanding immediacy
(e.g. IOC), thus restricting interaction to other ‘patient’ participants—you might use this to be even more
‘passive’ in your strategy, but any decision to constrain matching (whether limit price, MAQ or IOC optout) will reduce the fill rate.
Are independent dark pools viable in competition with
broker dark pools?
Charlie Guy, Chi-X Europe
Can I assume that by independent dark pools, you are referring to the pools such as those run by the
MTFs? If so, it is interesting that you ask if those pools are viable in competition with broker dark pools
and not the other way around since it is the MTF pools which are the larger.
I think the two types of pools offer different results and are going to draw in clients for different reasons.
Fills are likely to be larger with a broker orientated pool and it may also offer an outlet to fulfil commitment
to a broker. However these pools are not always open to as large an audience as the independent pools,
so brokers offering dark algos which need to sweep as much resting liquidity as possible are more likely
to be drawn to the MTFs.
We are already seeing concentration in the number of available ‘lit’ trading venues and it’s not hard to
imagine there will be similiar consolidation in the dark pools. However I do think there is scope for pools
to further differentiate on the size of trade they generate, ie minimum or no minimum size. Whether this
plays into the hands of the independents or brokers remains to be seen, but overall I think the readily
open access models of the independents is a strong plus point in their favour (but, of course, I am biased!).
Salvador Rodriguez, Citi
Each dark pool has its own trading characteristics. Broker pools differ from one another in the way they
interact and trade with liquidity. Each one is different. Some connect to external pools for example, others
do not.
Independent dark pools can succeed, however. Their success will be measured on how much liquidity
they have, the quality of the execution as measured against prevailing market prices and any reversion
associated with particular fills. One can also measure size of execution. Are fills only of 1 share lots? Or
are they more meaningful in size?
How to Catch a Black Cat in a Dark Pool
ATMonitor Commentary
Another question users of dark pools should be asking is how that particular venue is protecting an order.
• What anti gaming measures are in place? Who has access to and the ability to interact with that Pool?
Do they advertise to the market the composition of the pool real time?
• Brokers generally have the benefit of interacting with many streams of in house flow: PT, Cash and so on.
It is less clear how and with whom independent pools are interacting, which makes knowing the answers
to the above very important.
• At Citi, we strictly control the nature and type of counterparties that we allow to access the pool giving
us a very low level of toxicity.
Natan Tiefenbrun, Turquoise LSE Group
It’s an interesting question, and it has generated some interesting answers! Independent dark pools
certainly figure more prominently in the European landscape than they do in the US. Broker-pools (and
their clients) enjoy some advantages that MTF pools do not—including the right to control admittance
(more talked about than used), to trade anywhere within the spread, and to offer some optionality about
which categories of customer interact. Broker pools are also closer to the buyside and hence to the
larger ‘parent order’. On the other hand, there’s some disagreement about what type of access can be
provided and to whom, with some regulators pushing to limit broker pools to direct matching between
buyside clients—a restriction that would substantially alter the liquidity profile of many pools. Whilst
MTFs have to be non-discriminatory in terms of access and matching, they can of course appeal to a wider
audience. The need to tackle ‘toxicity’ which Salvador highlights is a real issue—and one which Turquoise
has a unique approach to—allowing customers to opt-out of interacting with immediacy-demanding flow.
I expect that a lot will depend on whether public/neutral pools are every broker’s second destination, or
whether linkages between broker dark pools ultimately reduce public venues to being “destinations of last
resort”. The outcome will likely depend on the contents of MiFID II. In the meantime, we’re working on
enhancements to our dark pool to make it more attractive for institutional block traders.
How to Catch a Black Cat in a Dark Pool
ATMonitor Commentary
About ATMonitor
ATMonitor is a new service designed to offer buy-side traders
a unique, valuable and independent resource in the field of
global equity trading and technology. Its member-only part,
MyATMonitor is a service available only to buy-side traders
and provides a unique environment where members can share
knowledge and experience directly with each other, through
forums and blogs, confident that service suppliers, consultants,
headhunters etc. are excluded from the dialogue. This exchange
is encouraged through opinions, editorial, research and analysis,
provided by experienced personnel with close knowledge of
the global equity trading business.
Visit us today at
To discuss any part of this report,
Jason Murray
[email protected]
+44 (0) 208 466 5650
+44 (0) 7817 901 757
Rufat Khamidov
[email protected]
+44 (0) 208 466 5650
+44 (0) 7424 364 511
For marketing, membership and
other business enquiries, contact:
Karen Delahoy
[email protected]
+44 (0) 208 466 5650
+44 (0) 7793 884 244