Trying to decipher platform performance
during last month’s UK-inspired mayhem is almost as difficult as keeping track
with the latest shenanigans in UK politics following the vote to leave the EU –
but I’ve had a go and there seem to be some interesting angles.
I am not talking about the technology –
that worked across the board with no issues reported and liquidity fairly
steady, if not spectacular. Rather I am talking about volume data and,
obviously, month-on-month the news was uniformly good. As I hinted at when
reporting on one of the volume releases, however, I am increasingly hearing of
a divide when it comes to the model that garnered market share.
My sources are telling me that they saw
more volume on the ECN/exchange-type model than they did on the disclosed,
request-for-stream (RFS) model.
I have to stress this is anecdotal because
no firm breaks down its activity by model, however I am told that, across the
major providers, EBS Direct volumes were slightly up, as were those on FXall’s
RFS model; while EBS Market, and FXall’s OrderBook (and Thomson Reuters’
Matching) saw much stronger growth. Market making sources tell me that volumes
were also lower through aggregation channels, to which can be added the opinion
from buy side sources that liquidity was better on the all-to-all models than
via the aggregation channels.
In one way I find this information a little
surprising – after all LPs are always very keen on knowing who they are quoting
– but in another I think it reflects the changing nature of the industry.
There was a real and genuine apprehension
in FX circles over how the market would operate during the UK referendum, this
was not limited to market makers, it was also apparent amongst investors,
punters and hedgers – no one, it seems, was immune. In such circumstances, not
only are sensible hedgers likely to hedge early or avoid the FX market during
the mayhem, but LPs are going to be less happy to provide the type of pricing –
in size – which often goes over the RFS/aggregation model.
In other words, the genuine end user
function of the FX market kind of idled in the background while the speculators
took over – and if that happens most can only find liquidity on the
ECN/exchange model. This partly because the LPs have pulled back from directly
streaming to a fair chunk of the speculative community – they have also, more
pertinently, shied away from the “liquidity downstreamer” model.
We have all seen the explosion in the
number of firms claiming to be liquidity providers (and then refuting that tag
by citing the “tier-one” liquidity they receive – a liquidity provider initiates liquidity rather than recycles
it – rant over!) as well as those firms that have stepped out of the heavily
regulated retail FX space and moved into the “institutional” market using –
again – “tier-one” liquidity sources.
With everyone so nervous about the market
in the run up to, and immediate aftermath of, the UK vote, these “tier-one”
liquidity sources did what they do whenever things get busy – they turned off
the re-cyclers. This left the latter with only the ECN/exchange models as
sources of liquidity – hence they push their volumes that way.
If nothing else, the data from June
suggests the disconnect between the speculative community and the major bank
liquidity providers has increased. It could also suggest that when volatility
hits the market, volume goes to the all-to-all model because it is also the
only place where so many participants (who were previously cherished clients in
the world of the big market share ideal) can go to trade.
There is also the factor that when things
get volatile, LPs are often keen to switch off their less important (in terms
of volume and quality of client/counterparty) streams, but they can’t switch
off their own sources of liquidity.
Now this all sounds like good news for the
platforms – and for those that used to be single model businesses it is. After
all, if you relied upon RFS for volume you probably had an average June, and if
you rely upon the ECN/exchange model, you probably had an average few months
leading up to June.
The situation is not, however, altogether
Looking at the first six months of this
year, of the six firms reporting more than $10 billion per day in volume (CME,
EBS, FastMatch, FX SpotStream, Hotspot and Thomson Reuters), only CME and
FastMatch achieved a new high water mark for 2016. This could be viewed
positively of course, after all it shows how well the others had been doing in
the previous months, except….
Putting aside FX SpotStream which recorded
its second highest month and is something like 35% up on H1 2015, the other
three venues all handled lower volumes in H1 2016 than the same period in 2015
– as did CME. Hotspot and Thomson Reuters are lower in fact, that the same
period in every year going back to 2010, and EBS and CME’s first half only
escapes that tag because of a pretty horrible H1 2014.
So we have a situation wherein the four
firms that are not growing from a small base all saw activity drop in the first
half, in spite of what was one of the more major events of recent years.
Of course there was another major event in
recent years, so I thought I would take a look at the ADV across firms from
January 2015, when the SNB so thoughtfully created volatility, and June 2016
when the British public did likewise.
What would traditionally have been called
the incumbents, EBS and Thomson Reuters, didn’t fare well, seeing a near 25%
and 21.5% drop respectively. Hotspot was down 16% and only CME had anything
remotely like a positive outcome, activity in June was a whole 0.8% higher than
January 2015. Comparing the same months, FX SpotStream was, by my estimates,
about 40% higher and FastMatch was 90% higher – both are, it needs to be
remembered, still in the early growth phase of their business.
It seems clear to me that while the UK
referendum brought short term relief to the platform providers, that was all it
was – short term relief that helped alleviate the pain of what was threatening
to be a pretty poor six months.
Looking at the four largest providers there
is what looks like a long term downward trend. CME had a good June but in
general terms its numbers have stalled over the past 18 months, whilst the
other three – EBS, Hotspot and Thomson Reuters – are having to come to terms
with much stepper declines. All will, I suspect, have a different take on
At EBS the negative is obvious – numbers
are not rebounding as strongly as would be hoped – and I notice that for the
past two months the Icap press release has stated that EBS data includes spot,
outright and swaps, which suggests spot volumes are (albeit slightly) lower
than actually published.
The good news for the firm, however, is
that it has firmly established itself in one of the few growth areas
geographically – China. It is hard to find someone who doesn’t believe the
Chinese FX market will continue to grow and EBS seems well positioned to
benefit from that. Whether it will make up for the obvious decline in the
majors is unclear at this stage, but there will be more than a few people
inside the firm grateful for the effort that brought such an important new
currency into its sphere of influence.
China is also, I suspect, part of the story
at Thomson Reuters and one day there will no doubt be a post mortem within the
firm as to how it allowed EBS to so dominate what is an important and growing
FX market. The firm seemed to have a good position on China a few years ago and
that has, in FX trading terms at least, slipped.
Elsewhere, it is hard to escape the
conclusion that, as I predicted years ago, it has proven impossible for the
takeover of FXall not to lead to at least a degree of cannibalisation of the
My understanding is that Matching had a
good month on the back of its position in sterling – dealers tells me some good
tickets went through on June 24 and the days immediately thereafter – but the
overall numbers tell their own story. The combined spot volumes at Thomson
Reuters are lower than they were for Matching alone just a couple of years ago.
Again, it is not necessarily all bad news,
though, for my sense is that as the industry pushes further and further down
the path of better conduct and more transparent modus operandi Thomson Reuters is ahead of its immediate peers in
this area and that could well provide a boost in months and years to come – it
is a long road, though, so patience is needed.
For me, the negative tone at Hotspot is
less accentuated. It is more a question of why the firm isn’t growing rather
than why its volumes are falling. The actual numbers around Hotspot are not too
volatile – in H1 2015 it was handling ADV of $28.9 billion, in the same period
in 2014 it was $29.4 billion and in 2013 it was $30.6 billion. In the first
half of this year ADV was $27.8 billion.
So while the trend is lower at Hotspot –
and that must be something of a concern to its owners – you sense that the
platform is one or two active participants away from matching its numbers of
recent years. Of course, finding, acquiring and holding those participants is
the big challenge and as I have noted before in this column, the sales model at
Hotspot has not historically seemed (to this uninformed onlooker at least) to
be particularly beneficial to growing a business.
The good news is that model has changed,
it’s just taking time to bed in maybe. Hotspot’s other positive is it has
owners with decent pockets and a commitment to the business, but that can be
said for a lot of other platforms out there, and as such it is less of a
differentiator than it may have been perhaps three years ago.
As far as CME is concerned, I have long
been of the (public) opinion that it had hit its sweet spot in terms of the
type of participant using the exchange. Extending its reach into more entrenched
users of the OTC market was always going to be difficult and it seems to be
proving just that. There is nothing overtly negative about this, unless of
course, you are paid by CME to produce ever-increasing numbers!
Markets have continued to move in July so
there is the chance of a further positive wash through into this month’s data,
but I doubt it will be that noticeable. The fragmentation of the multi-dealer
venue industry continues to bite and unlike a few years ago, we no longer exist
in a world in which new providers can claim to bring new volume to the
For the past 18 months – and this could
also be a knock on effect of “SNB Day” – it has become apparent that, rather
than introducing new volume, the game now is about taking a piece of your rivals’
pie. The real problem is that, as we are likely to find out in September, the
pie is shrinking – not dramatically, but enough to cause concerns – and all
incumbents are going to have to think hard, possibly change their view of the
world, and more probably put their hand in their pockets.