Read time: 4 min

Data Paints an Interesting Picture for Platforms

While all seems rosy in the FX world
following the FX committee
surveys, there are a few interesting
questions over activity on those platforms
that publish data. Firstly, why weren’t
results a little more positive and secondly,
where did all the yen volume go?

For those looking for an illustration of
what fragmentation has done to the FX
market, April, and in particular the yen
numbers, would seem to be a good place
to start. Across the major centres,
average daily volume in USD/JPY rose
by more than $285 billion in April 2013
from the same month the year before.
EUR/JPY volume in those centres rose
by about $60 billion, meaning yen
volume probably rose by around $350
billion per day.

Historically we would have looked at
those numbers and then looked for the
expected surge in activity on EBS and,
more latterly, CME. What did we actually
get? Well, they were busier over the same
survey period, EBS rose 17% year-on-year
and CME 13%, but that was across all
currencies, not just the yen and those
numbers are nowhere near the almost
trebling of yen volume from the surveys.

Clearly other platforms grabbed some
of the flow, but market sources say they
cannot pick out a venue or venues that
may have witnessed such a surge. Of the
other platforms that publish numbers,
Thomson Reuters was unchanged year-
on-year and the yen remains something of
an irrelevance on the Matching platform
so it was unlikely to see much of a
bounce. Elsewhere, FXall had a good
year, but again it was spread across the
board, and of course the platform includes
forwards and swaps in its numbers, which
Profit & Loss estimates is about half of
all activity. Additionally, yen swaps
activity almost doubled in the UK and it
was in the one day to one week tenor, an
area FXall is very strong in. Finally,
Hotspot FX also rose strongly. In fact,
aside from Thomson Reuters (which has
had a couple of strong months since), all
platforms were up about 15-20%.

In volume terms, the total year-on-year
increase from the four platforms that saw
a rise in activity was just under $55
billion. Assuming this is all spot yen,
which it probably wasn’t, that still leaves
the small matter of an unaccounted for
$300 billion.

Banking sources say they saw a strong
increase in activity on single dealer
platforms, but an even bigger increase was
API-based and via aggregation venues.

This suggests that much of the “missing”
volume went through the myriad of
aggregation venues available, suggesting
the democratisation of the FX market,
which started about three years ago, is
really picking up some serious steam.

This supposition is supported by the
market share data released by the JSC,
FXC and Tokyo Committee. In the UK,
the First Quartile of banks (six in total)
handled 72.55% of all USD/JPY flow
(across products, not just spot), which
comes out at roughly $365 billion
dollars a day. This compares with the
63.52% of April 2012 USD/JPY flow
seen by the same banks, amounting to
about $148 billion. It is a similar picture in EUR/JPY when the top six
saw a more than 10% increase in market
share amounting to something like $32
billion equivalent.

It is a similar picture in the US,
although the FXC breaks it down by the
top five banks. Here the top group’s share
of activity rose by around 5% in both
currency pairs, a volume increase of
around $80 billion. In Japan, the top eight
institutions saw their market share of yen
products rise by 5.3%.

Aside from the existing platforms,
however, the April data may also offer
food for thought for some of the newer
entrants to the multi-dealer world, for it
seems increasingly likely that a single
model is unlikely to survive, unless that
model is predicated upon bespoke
liquidity. Having a traditional-style ECN
with better technology may not be the
answer some people thought it was when
the new models started working their way
into the system, especially as the
aggregation venues that are seen to be
doing best, are those that offer either bank
aggregation, or multi-dealer platform
aggregation, but not both in one solution.
While the MDP aggregation venues did
okay in the month according to sources at
some providers, a greater surge appears to
have come from the single dealer
aggregation venues as the guaranteed
nature of bank liquidity once again proved
a winner.

For the more mature offerings, the April
data highlights a few uncomfortable
trends. Clearly, EBS Direct is going to be
a very important business for the firm if it
wants to continue the rebuilding process
that is now almost 18 months old.
Likewise, Thomson Reuters can look at its
FXall purchase and see the opportunity to
extend the bespoke liquidity work that
firm has done to help avert any decline in 
matching volumes that may occur through

The big surprise was CME not seeing a
huge surge in activity as clearly the world
is buying into the cleared model. CME
probably didn’t grab a chunk of business
that month for a couple of reasons; firstly,
there are still issues for a lot of ‘real
economy’ firms in trading futures, and
secondly, the banks are still reluctant to
get heavily involved there. It was also a
“non-roll” month for CME, which would
normally mean a decline in activity, but
interestingly the exchange saw a decline
in activity compared to other non-roll
months in 2013, although yen activity was
strongly higher across the survey periods.

The April data from the FX committees,
then, paints an interesting picture.
Fragmentation is clearly alive and well,
and the likelihood is that much of the 350
extra yards each day was spread thinly
around several providers, both old and
new. Aggregation venues clearly did better
than most, but the ultimate winners would
appear to have been, once again when
things get busy, the banks.

Non-bank market makers typically do
not hold large risk and April was clearly a
month in which, for yen traders at least,
risk warehousing had to be a part of the
game plan – and some of the risk they had
to hold was undoubtedly large – even if for
a few minutes. More pertinently for the
non-bank market makers, those that are
only aggregated through the multi-dealer
platforms would have seen a decline in
activity. More than ever, for this model to
succeed, the non-bank market makers need
to be connected to the ultimate end users.

There is also something in these reports
for the end users of the FX market. Profit
& Loss is hearing from banks that they
analyse when their clients are trading with
them and via what mechanism. The school
of thought being if the client is with you
over a visible venue when things are
sedate, it is OK to keep pricing them when
things get busy. Alternatively, and this is
food for thought for that group of clients
that tend to have a loose interpretation of
the word ‘relationship’, more banks are
looking closely at those firms that trade
anonymously for much of the year and then
head straight for an aggregated or single
dealer feed the minute things get volatile.

There is a lot to feel content about in
these surveys, but there is also plenty for
all sides of the industry to ponder. A new
market structure is evolving and we now
seem to be witnessing the first empirical
evidence of that. 

Paul Gogliormella

Share This

Share on facebook
Share on google
Share on twitter
Share on linkedin
Share on reddit

Related Posts in