Chris Concannon, partner at Virtu Financial, who sits on the CFTC?s Subcommittee on Automated and
High Frequency Trading, talks with P&L?s Julie Ros about his move from equities to FX, and the role that
Virtu plays among the new breed of FX market makers.
Julie Ros: You were an early proponent of Dodd-Frank, lobbying
extensively for it while overseeing US markets at Nasdaq. What
prompted your move to Virtu in 2009?
Chris Concannon: I spent most of my career in cash equities,
having started at the Securities & Exchange Commission as an
attorney, before joining a law firm. I then moved to the exchange
world joining Island ECN until it merged with Instinet and later,
Nasdaq, where I was quite active in the Dodd-Frank world post-
crisis. It became clear to me that the world would change pretty
dramatically over the next five years in terms of where and how
financial products would be traded and I wanted to participate in
that kind of evolution. I don’t think it’s the sort of thing that
happens rapidly, but I see it slowly changing.
JR: What changes have you seen so far?
CC: Obviously Dodd-Frank has been slow to be implemented,
but what we have seen is people preparing for it even before the
rules are written by rethinking their business. If you look at the
Volcker Rule, it hasn’t even been defined yet and people have
already evolved and changed their businesses just on the threat of
it. I see large banks evolving pretty dramatically, trying to solve
some of the cost challenges they absorbed over the years as a
result of the change in business model. I see the end user across
all products becoming much more sophisticated. When you get
an end user that adopts a technology like an e-trading screen,
they want to see quotes and get access to markets electronically,
so that’s a huge predictor of what the model will look like.
JR: Tell me a bit about Virtu.
CC: We are an electronic market making firm with offices in
Singapore, New York, London, Dublin and Los Angeles, but soon
we will be moving our London to Dublin and LA to Austin. The
asset classes we trade are cash equities in North America, South
America, as well as in Europe and Asia. We actively trade futures
– financial futures, commodities (softs – sugar, corn, wheat), and
all Comex products (metal and energy). We have a fairly sizable
footprint in all energy products. We trade FX as market makers,
so making markets, capturing spread across locations around the
globe; we can trade from pretty much anywhere. We connect to
most old and new ECNs. All bullion products, gold, silver,
platinum, palladium, and futures in those products as well. We
are a big market maker in all US, Europe and Asia ETFs. Within
FX, we trade 75-80 pairs, in spot and forwards and some NDFs.
JR: What do you do in LatAm?
CC: Cash equities, futures and FX in Mexico and Brazil.
JR: At our Forex Network Chicago conference last September,
you said: “We are not HFT, because it has not been defined yet”.
Do you know whether you are an HFT yet?
CC: If everyone’s an HFT, then we’re an HFT. I like to take a
step back and say you can’t access these markets manually – no-
one accesses the markets manually and in FX, no-one quotes
manually. They may manually update their forward rates, but everyone auto quotes so by definition every bank that streams a
quote to us is an HFT. If someone has a different definition other
than computer-generated quoting, then I don’t know what we are.
But if you have an auto quoter anywhere in your organisation,
you are by every definition I’ve seen around the globe, by every
regulator, an HFT.
JR: There has been some talk of rolling spot potentially being
considered a swap – do you have an indication of CFTC’s
thinking on this?
CC: It was originally because NDFs were not excluded and
people started asking if having different settlements for spot is
really that different. I think that’s just people technically reading
the exemption and not taking a step back and looking at the
policy of the exemption. I think it’s an issue created by attorneys
in the private sector that are now asking for clarification. CFTC
hasn’t answered the question, which is why everyone is
concerned. That was not the intent of CFTC, and it was
definitely not the intent of Congress.
JR: So are you a proponent of more regulation of the sector?
CC: In FX, no. Dodd-Frank is more about taking privately
negotiated, very liquid financial products and putting them in a
cleared facility with some trading transparency. I don’t think FX
needs that. The driving force behind Dodd-Frank was not the FX
market. The FX market is one of the most efficient markets, and
it has gotten to that efficiency through commercial needs, not
JR: In your dealings with CFTC, is it your impression that they
share this view?
CC: They’re more focussed on the retail side of FX, where we
have seen concerns raised around margin and the average
lifespan of the retail client; I think that concerns them.
JR: Are you concerned there is a risk of unintended
consequences from Dodd-Frank?
CC: Yes. I think the unintended consequence is, ‘What about the
rest?’ If you push products into cleared facilities, NDFs and
options, and it becomes a little bit difficult, because we live and
breathe the world of OTC and cleared, how do you marry the two
and manage collateral and margin? It’s very difficult, and
sometimes it’s very inefficient to have OTC and cleared – whether
it’s a future and OTC product, whether it’s a future and cash
equity product – there’s margin that’s being housed in those two
asset classes that needs to be funded. What’s happening in FX is
kind of what’s happening elsewhere due to Dodd-Frank, and that
is the need for additional margin. You have your FX options,
NDFs and spot all in the same prime broker account, but now
there’s an exchange that’s calculating margin based on the naked
FX option and NDFs but doesn’t see the spot. This creates a
demand for capital where, prior to this Dodd-Frank
implementation, everything was cross margined in one account
and held by the prime. What I think eventually could happen is that it can drag other products into clearing. ‘If we have to clear
this, why not clear the rest?’ is what we may hear from end users.
JR: Do you see FX spot going this way?
CC: I can’t say we’re making that prediction right now, but you
can see there will be a demand to move more product into
cleared over time. As the cleared gets worked out and people get
comfortable with how it works and how things margin, this
demand for capital is one reason that could drive other products
in. Unless there’s a solution for the demand for increased capital
– that much capital really starts to push products into cleared –
so I don’t sit here and think that spot will be cleared because of
options and NDFs, but over time, you’ll start to see more people
say, ‘Let’s create a cleared spot and see if that helps.’ Then you
would be able to hold cleared spot for spot, so that could happen.
JR: Where does CLS sit?
CC: Our trades go through it, so we benefit from CLS. I don’t see
it ever going away, I see it adding more services over time. It’s
providing a netting function that is critical to settlement of spot. If
you look at the history of clearing in other products, it always
starts with netting and the second step is attaching a guarantee
with margin associated with that guarantee; so netting always
comes first and then clearing. I don’t see CLS providing a
guarantee in the future, but the market may seek one at somepoint.
JR: HFT is suffering because of low volatility, increasing PB
costs, as well as general trading costs in FX. What do you see as
the future of HFT?
CC: We’re in a growth mode in FX. Even with dips in volumes,
our penetration in the FX market continues, so we’re very excited
JR: Where are you focussed?
CC: We’re growing across the number of pairs we trade as a
market maker and we’re growing with respect to the venues that
we trade in, as well as the counterparties that we trade with. A lot
of new growth is coming out of Asia – CNH, INR – CME just
launched their new CNH contract back in February, there are
new venues popping up every quarter, these are all things we’re
JR: Do you see FX trading shifting to Asia as a result of
regulations in the US and Europe?
CC: I don’t know if it will shift, but we definitely see Asia as a
growth spot for FX and see that continuing; the demand out of
Asia from counterparties just keeps growing.
JR: Is there the same development level in terms of high velocity
CC: Not really. The same people here are there, so it’s all the same
market. The end users are different and how they come into the
market is very different, and that’s what’s developing – not as fast
as it has developed here or in Europe – but that’s where you see
development. How the end user accesses the market is becoming
more sophisticated in Asia, it’s not a single bank portal that they’re
accessing exclusively, so that’s getting more developed.
JR: Where are they trading?
CC: Aggregators – both new aggregators in Asia and the current
aggregators are having more success there.
JR: Do you use prime brokers?
CC: We have had several PBs since 2008. It makes sense to have
multiple PBs – partly for the relationship, partly for the protection.
JR: Smaller HFT firms are struggling, but the bigger ones are
faring better, why?
CC: We live and breathe scale. We trade so many different
products across so many different geographies that we benefit
from sector-based volatility. Any given asset class that we trade
will have different, unique volatility events that if you’re in just
one or two, you may or may not be experiencing. Given the
breadth of our geography, we create our own FX volume. We are a
big end user of spot currency needs, because we’re trading in
Europe and the US and therefore have so many cross border trades
going on within the firm that we have a very large FX demand.
JR: Do you trade currency options?
CC: No, but we anticipate that we will as it goes through its
transition to cleared. We’re waiting until it gets there and then
will look at it, as opposed to trying to work on it now when it’s
not really there.
JR: Do you think fragmentation is good for the FX market, or is
there a point when there are simply too many venues?
CC: The FX market is enormous and I don’t know what the
sustainable level of fragmentation is, but what I have experienced in
my career is that the industry finds the right level. You tend to get a
burst of new entrants and some survive and some don’t. I think
we’re just at the beginning of that process where we continue to see
more entrants and either a consolidation by M&A or attrition. It’s
just hard to say where that sustained level will be…20, 15, 10?
There’s also a geographic element to which market survives. Also,
can a new entrant live on two pairs or do you need 10? Will there
be a platform that does exceptionally well in Asia, or do they all
have to be in New York? I think the market has enough liquidity
and users to support a sizable number of venues.
JR: How many markets do you connect to?
CC: We thrive in connecting – the firm connects to over 200
unique markets and 40 data centres.
JR: What has been the real market impact of internalisation?
CC: Internalisation is an interesting term. Let’s say a bank has
two clients – one has very aggressive flow and the bank’s
internalisation rate with respect to that client is not high so
anything they trade with that client loses money, so they widen
the spread and the client uses an aggregator and he never again
trades with that bank. The bank’s internalisation rate has just
gone up, yet that flow is now redirected into the market, so a
bank can experience a higher internalisation rate with less
volume. I never understood the goal of having a high
internalisation rate. What we see happening is flow being
released into the market that used to be included in internalised
flow and it’s now being pushed into the market. Again, the end
user has become more sophisticated, the flow is more
sophisticated and since we also live inside ECNs, we can see that
flow coming directly into the market. It’s an all-to-all market –
anyone can come in and trade against your quote – we’ve
performed exceptionally well in an all-to-all market. To the
extent internalisation rates are going up, but the flow is going
elsewhere and being redirected into the external market, we
receive a benefit from that outcome.
JR: Would you classify yourself as the type of client whose flow
CC: No. When it comes to the banks and the order flow that we
send them, it’s carefully selected and filtered to ensure that they
don’t get hurt by it.
JR: Have banks lost ownership of the market making function?
CC: No and they never will. They ultimately own the client, the
end user, so banks will always own the distribution channel and
as a result, they can pick and choose the most ideal margin for
market making. So there’s some market making that has margins
that are not ideal for a bank and it’s that area that they may
release into the market – just like widening out a client, what
they’re really saying is ‘I have low margin in this market making
with respect to this flow so I’m going to redirect this to another
venue’ in essence. So they will always be a dealer and always be
a distributor of FX. And they are, among the primes, the source
of credit in the market.
JR: What’s your view on firms such as Morgan Stanley, JP
Morgan, Goldman Sachs, scaling back certain prime services?
CC: I think they looked at the business and they looked at the
margin and they looked at the risk – the intraday risk that’s
involved in the business – and the math didn’t work out for them
and their risk officers. It’s a low margin game on a per unit basis
with what can be pretty sizable risk. But what I do see over the
next two years is that the ECNs in the markets that allow for this
risk are going to fix that risk problem for the PBs, so this won’t
continue, because venues just won’t succeed if they’re not offering
the primes that are clearing the end user with a solution for risk.
So for example, Reuters offers a gross risk limit, not a net risk
limit. Because it’s gross, it’s very hard to manage with respect to
a prime that lives and breathes in a world of net and sets their
limits based on net. So the gross limit is a very imprecise method
of controlling risk and that makes it not attractive for a PB. So
Reuters will have to create a solution that solves the problem for
JR: Do you see platform fees going up?
CC: When you have fragmentation like we’re experiencing, you
rarely see fees increase, you typically see them decrease.
JR: What about in PB?
CC: I think we’re at the tail-end of the fairly aggressive
commission compression period, but I don’t see them turning
and going up. In all likelihood, fees will stabilise where we are,
but not continue the drive to zero. But just like there’s
internalisation in trading spot among the banks, there’s
internalisation in the PB business – that is, ‘How much can the
PB avoid CLS by having its PB clients trade among themselves?’
So the bigger you are, the higher PB internalisation rate you
experience, and the lower your effective CLS rate will be over
time. That can be very profitable for a large PB.
JR: Are you seeing a move to a world of distinct agency and
CC: That’s the trend line. A lot of these new entrants in FX are
agents in other products and they’re adding FX as just another
product on their distribution channel. That seems to be a natural
move for them where they have the end client – whether it’s a
retail client or an institutional client – although we’re seeing it
more on the institutional side as more and more want to add FX
into their product mix. When their model exists as an agent, it’s
hard for them to exist as a principal with that same client
relationship. ITG is a perfect example, they have a phenomenal
relationship with the institutional community, a lot of credibility
with respect to price analysis and execution quality, and so they
bring that into the FX world and it’s hard for them not to be an
agent in that world.
JR: Since the FX market has been difficult over the past year,
what do you see as the allure of FX at this point?
CC: All the other markets are hurting, so FX is still good, comparatively speaking. If you live in cash equities, FX is a
pleasure given the cash equities issues that we’ve seen globally.
In 2012, after years of declines, cash equity volumes declined by
18% year-on-year, a substantial decline. If you’re in US cash
equities, FX looks great.
JR: Even with the stock market hitting highs?
CC: The market is up, all the indexes are up, but the volumes are
JR: Do you see a more order-driven market structure leading to
flash crashes in FX?
CC: Given the current structure of the market, FX and cash
equities just can’t continue to be compared. Cash equities have
6,000 symbols. The FX market has, in terms of actively traded
products, maybe 100. So you have a very different, concentrated
market with regard to FX products. Equities has a broad list of
names that people trade either as an end user or as a market
maker. In addition, the market in cash equities is much more of
an “exchange-based market” than what is still a “dealer-based
market” in FX.
JR: What’s your view on kill switches?
CC: It’s a good topic – everybody’s talking about kill switches,
they sound great. If I have one, credit officers will be happy – but
it all depends on how you use the kill switch. That Reuters gross
limit is a kill switch, and it doesn’t work correctly. So kill
switches are great to talk about, but how they’re actually
implemented is probably more important.
I do think that in FX, one area the industry needs to focus on is
the erroneous transaction area – the fat finger, the unintended
trade – it’s about how the exchanges co-ordinate on breaking or
adjusting trades. What are the policies and who decides? When
you have this intertwined system of venues, one break, one
mistake doesn’t happen on EBS alone, it happens across EBS,
Reuters, Currenex and FXall. How the decision gets co-ordinated
to break and at what price, or to adjust and what price, should be
standard across the industry. Many of the venues put this issue in
the hands of the PBs to figure out, because they don’t want to be
held responsible. But really the venues should adopt policies
about what should stand and what should be broken. I think I
spent five years in cash equities in the US trying to figure this
out and what we came to was just one standard approach to
everything, regardless of the event.
JR: What was it?
CC: You figure out what the reference market is – so you have a
standard for what the reference market is at the time of the event
and then you pick a percentage away as the point of break or
adjust. Here’s the problem: If every time I make a mistake in the
market, the venues or the PBs reverse my mistake and I get back
to zero impact, then I will never fix my mistakes. I will always
let the market be my backstop for my mistakes. But if the venue
says to me, ‘You made a mistake, you’re going to eat a good
portion of your mistake’, then I’m not going to make a mistake
the next day. So there has to be a penalty baked into the adjust
policy that incentivises people not to make mistakes. The
problem is, the industry has always been a dealer-to-dealer
industry, so the dealers have traditionally worked out these events
and moved on, but it has converted to an all-to-all market where
an end user is making a mistake and if the market is trying to
work it out and say don’t worry about it, then that end user is not
going to fix those mistakes. You have to hit the dog in the nose
with the newspaper so to speak.
JR: What is your view on aggregators – are they simply reconsolidating fragmented markets and re-selling that or do they
have a role to play?
CC: It depends. A lot of them have a front end, so they’re
selling front end technology and how they integrate other
pieces of information into that front end is critical to the end
user. Some are selling API connections so you don’t need a
front end, you connect directly to them. So in that sense,
they’re selling a reconnected market through one connection.
Some are selling data, a unique data set that they’re collecting
and aggregating. Some are selling routing, an algorithm, the
beginnings of what are algorithms, that will take you to venues
based on a criteria that you’ve set. So there’s a huge
opportunity for aggregators. When you add up all the
destinations where you can execute a spot trade, it’s not just the
venues, it’s every bank’s portal and API, so it’s a long list of
destinations in FX to aggregate together to make one solution
for an end user. So I don’t see the demand for aggregators
lessening, I see it increasing.
JR: What’s your preferred type?
CC: Our own! We connect to everything so we don’t even think
of it as an aggregator; it’s our core trading technology.
JR: What would be your ideal trading venue – or does it already
CC: That’s a great question. There’s an assumption in the
marketplace that folks like us that operate heavily focussed on
technology and speed and being able to consume data quickly,
prefer the fastest venue, but that’s not the case. Every venue is a
little bit unique to us so it really depends on the type of flow that
comes into that venue, how we’re treated by that venue – are we
treated as a market maker? Do our fees and our relationship
reflect that we’re actually a necessary part of that market or are
we treated as though we’re an outside party gaining access to a
market that shouldn’t be something we have access to? The markets we do well in, the ideal market, is one that allows
all parties to participate and doesn’t put up unnecessary barriers
to entry, but that doesn’t have to be the fastest market. There’s
definitely a misconception in the market that it is ‘us versus the
banks’. There are folks that have entered the FX market that look
like us – electronic prop trading firms that do think of it as ‘them
vs the banks’, but we recognise the value of the bank distribution
model and we want to fit within that world. So not only are we a
client to the banks, because we have spot flow and spot needs
that they solve for us, but we also want the banks to be clients of
ours because they have spot flow and spot needs that may be best