Today’s release of
Virtu’s financials, as well as the deal it has signed with JP Morgan, probably
make this the appropriate time to share my thoughts on the non-bank market
making sector in general, in particular the value of being in the FX business
for these firms.
A big factor behind
the deal from the JP Morgan perspective (and the bank has been reticent in
talking about it in public) is, according to my sources, the cost saving by
outsourcing the technology. Internal technology teams come with a hefty price
tag, as any banker can tell you. A firm like Virtu provides credibility in the
market place at a much cheaper cost.
Of course this type
of deal is also helped by what could be termed the equitisation of the US
Treasury market where credit is intermediated and the all-to-all model is
growing in popularity (driven by regulations mostly). Virtu is effectively
providing the bank with direct market access at a fraction of the cost.
The deal also, in my
eyes, reflects the challenging nature of markets as far as Virtu is concerned
and the firm’s desire to diversify its income stream.
FX has generally been
a good market for the firm but so far this year revenues are on the decline on
a year-on-year basis – they are still relatively healthy, of course, it’s all a
matter of degrees. This reflects the changing nature of the FX market, where
volumes and volatility are generally at historically low levels and only
interspersed by sudden, short-lived bouts of increased activity.
For firms that rely
upon high volume and volatility the changing nature of FX is not necessarily
Not all firms do, of
course, but I think it’s fair to say the majority are in the hands of activity
levels and those levels are in decline, in spite of a very uncertain
geopolitical and macroeconomic outlook – two ingredients that have,
historically, brought increased FX volatility.
This means the firms
have to look, as many are doing, for “partnerships” with banks. Virtu has done such
a deal in FX, of course, with Bank of New York Mellon and rumours abound of
other banks involved with the various non-bank firms, the ubiquitous XTX being
prime amongst them.
This makes sense and
my understanding is the partnership model has been one of the cornerstones of
the XTX plan, but even this diversification comes with challenges.
For a start, the type
of deal often spoken about outsources flow from a bank to the non-bank market
maker (NBMM), but it doesn’t, according to my sources, necessarily involve all flow. The bank can still, to a
degree, cherry pick and only pass on the more toxic flow. Some of the NBMMs are
that good of course they can profitably handle flow that other liquidity
providers cannot, but generally speaking there is still an invisible brick wall
between the market maker and the genuine end user flow – and that remains the
key to a really successful FX business.
So toxic flow will be
passed through but do the NBMMs want that? Of course they don’t. I have been
told by several sources in the hedge fund world that literally weeks after
agreeing to add a non-bank firm to their portfolio of streaming liquidity provders
the firm has sought to cherry pick the volume – along the lines of “we can make
money out of your euro and yen flow but we don’t want to see your sterling”.
I don’t think this
flies because the hedge fund will ask the question ‘what else do you do for
me?’ and the answer will be not a lot –banks still loss lead the odd bit of FX
business because, frankly, they can.
So there is a real
risk that as these arrangements become more commonplace the bank will still
cherry pick the flow – “this stuff makes me money, I’ll keep that and see what
I can do with it, the balance I will hand off to these guys”.
The challenge for the
NBMM at the moment is that it is still at the end of the food chain when it
comes to this original business – it’s hard to cherry pick scraps.
So the key for these
firms becomes getting closer to that end user flow – and being able to handle
both good and bad.
There is also the
question of whether the business structure can stand up to increased toxic flow
and, potentially in FX, greater regulatory and compliance requirements.
The NBMM model is
predicated upon being a low cost provider – as the Virtu-JPM deal highlights –
but that model also means fragile profit margins in the event a client gets too
good at their market timing. A requirement of getting closer to the end user
flow in FX is also the ability to quote in size and warehouse some risk – some
NBMMs can do this, some can’t. The latter have market impact and are generally
shunned by the end user, the former, however, have to consider the risk
management impact of such a policy.
How long is it, for
example, before a regulator starts treating the NBMMs, for compliance purposes,
as a bank? If they’re holding risk and quoting in larger size how are they
different? Will they have capital requirements under Basel? These are unknowns
but do represent a risk to the low cost provider, because in FX it is not all
about being best at top of book or being able shift relatively small amounts
quickly – it is about dampening market impact from an end user’s hedging policy
amongst other things.
I found it
interesting in the Virtu analysts’ call that CEO Douglas Cifu reiterated the
firm’s desire not to build a large sales force and change the nature of the
firm. He also used the ECNs and exchanges as a benchmark for its FX revenues. Inevitably
deals such as that with JPM (and others) will mean some relationship function
and that means the benchmarks have to change. The NBMM business cannot remain
just about trading skill – agency and risk warehousing has to come into the
equation at some stage.
I think this issue
once again highlights how FX is different. The non-client, trading only model works
in equities – it will probably work in the more commoditised (and cleared)
fixed income markets, but will it work in the OTC, uncleared world of FX,
where, as the FX committee data the other week shows, a lot of customers like
to deal on a direct, disclosed basis and use risk transfer?
There are two other
issues that will emerge at some stage – one can be handled the other I don’t
think can as easily.
Firstly, taking the
“easy” challenge first, the NBMM model in FX is still, to a degree, reliant
upon the banks. Not only is credit provided through the prime brokerage unit,
but generally speaking to get anywhere near the end user flow the NBMM has to
go through a bank (at this point see cherry picking section above!) The NBMM
has to be careful not to tread too hard on the toes of its PB provider – start
getting into areas that impact the PB’s wider business and watch the fees head
north. Overall though this aspect of the relationship seems balanced, although
capital pressures on the banks will inevitably mean PB fees edge higher rather
The other interesting
issue for me is that the whole non-bank market in FX is just too crowded for
what business is available. I cannot see how there will be room for a large
number of providers in FX terms because there is only a certain amount of
“good” business on offer and there are dozens of firms likely to be chasing it.
are also continuing to rise and inevitably we will see the entry of new firms
that are simply faster because they are smaller. In this situation there is a risk
that the historical hunter becomes the hunted and again, with small margins,
losing regular pips to speed merchants is a real challenge for the business.
Taking aside this
question of the “shark tank” model that will emerge in certain quarters, there is also the matter of the bigger firms
and their impact on one’s business. I found it interesting that in such a
highly correlated market like global equities, Virtu did well in the US and
Asia, but saw revenues decline in Europe. Partly this could be traumatic market
conditions around Brexit but it is also likely to be the reflection of
increased competition from other non-bank firms – again XTX springs to mind –
in that area. The same could be suggested when looking at the declining FX revenues.
I don’t think this becomes a
straight case of winners and losers – Virtu for one has the footprint to be
able to hit back hard – but I do wonder if too many firms means they end up
eating each other’s lunch?
As I stated earlier, I am
generally optimistic over the fortunes of some non-bank firms in this space
(Virtu being one of them) but not all. I
don’t see FX going to an equities-style cleared model and as such the key
remains, as firms like Citadel, Virtu and XTX to name just three are
trying, getting closer to the end user flow.
For those seeking to remain a
straight proprietary trading firm majoring in market making I see their
continued success being predicated upon at
least two of three conditions:
The continued goodwill of their credit providers
The continued retreat on the part of the banks from the FX
business – leaving the non-banks as last man standing
Markets getting busier
FX is a simple game that starts
with the end user customer flow. Some banks seem to be “juniorising” their
sales and relationship teams, inevitably meaning there is a period of
opportunity for others – the non-bank firms amongst them.
This window is relatively limited
though so the time for these firms to act is now. I am not sure given the changing
nature of markets that the prop trading model works in FX – it is all relative
of course, Virtu made over $37 million net FX income in the first half – and I
am sure that to succeed in this space you will need scale and diversification.
Tougher times are inevitably
coming, and as is always the case, your success or otherwise is driven by how
well you handle and survive the bad times.