For the lovers of total transparency among
my readers there is good news on the horizon as I am told that the FastMatch
sponsored Tape for FX is “one or two banks short of becoming a reality”.
That a Tape is being promoted by FastMatch
is no surprise, CEO Dmitri Galinov has long been a proponent of FX heading
towards an exchange model complete with a Tape, but there is some surprise at
the level of support for the venture – something I believe is driven by the
increasing desire of some players to become agency players rather than
Is the Tape a good idea? I understand its
value in an environment in which a product is traded across multiple venues in
one centre but does the concept work in FX with its global,
multi-jurisdictional and equally fragmented framework?
I like the idea of a Tape for informing
TCA, having better data means we can, theoretically at least, make more
informed decisions (I will point out the standard disclaimer about past
performance not reflecting future returns at this point – liquidity is not a
rigid concept). A big question is, I suppose, when the data should be made
My understanding from talking to people
familiar with the project is that the FX Tape will publish all trades from
certain venues up to a specific amount and that trades beyond that amount will
be published one day later. This makes good sense, for as I like to point out
(repeatedly!) no one will want to win a large EUR/NOK trade for instance if it
is going to published for all to see before the risk can be effectively
In this instance the market maker will not
only quote the customer much wider to avoid winning the trade, which is a bad
outcome for that customer, but they will also create a much larger market
impact by clearing the risk quickly, before the trade is published.
So the fact that larger trades are going to
be delayed by up to a day is a good thing. But does this not dilute the value
of the Tape itself? It is already undermined to a degree by the fact that not
all venues and participants have bought into the concept – market data is still
a very lucrative business for some – so doesn’t the fact that larger trades
will not be published until much later further dilute its value?
The answer depends upon how people are
going to use the Tape. If they are looking to provide TCA to prove best
execution then yes, there is real value in this. Of course some customers may
have to wait longer than they currently do for a TCA report but that may not
matter to some. If the Tape is going to be used as an audit mechanism it has
Will it improve execution performance for
customers? That is less clear in my eyes for as I have also (repeatedly)
observed in the past, total transparency of order and best execution do not go
together in FX. This is especially the case when there are so many nimble
players keying on any pattern in markets ready to jump in front of an order for
a few pips.
This may not seem much but it can become a
self-fulfilling prophecy in many ways, for as one firm jumps in front of an
order they have spotted, so other less sophisticated “sniffers” will be able to
spot the move and jump on it, which in turn….I am sure you get the picture.
A challenge for someone using an algo to
break an order down in an FX market that has a Tape is that the order will,
inevitably, be easier to spot. Current thinking is if you don’t want market
impact you don’t go to a lit venue, if the Tape also includes dark venues, such
as single dealer platforms or other aggregation methods, the number of lit
venues increases exponentially.
There is already a growing body of
customers unhappy at having to assume market risk by using an algo but they
feel, according to some I have spoken to, that they are being pushed in that
direction because too many of their providers no longer want to provide risk
transfer pricing. That is, of course, the right of the provider to decide (they
will still claim to be “client-centric of course) but it doesn’t exactly help
that end user whose expertise lies elsewhere – anywhere – than FX.
There is also the issue that not only will
the customer’s order be easier to spot if they are running the algo, but just
it can be just as easily identified if they are using some bank-sponsored
Signalling risk is becoming a more widely
discussed topic in FX circles thankfully because I feel after two years of
banging the drum, that it is still under-estimated. What people are forgetting
in the current environment is that as banks move more towards the agency model
at the expense of their principal business, so that bank’s algo strategies can
be more easily spotted. After all, if a bank isn’t really in the principal
game, how hard can it be to spot one of its algos operating for a customer?
Principal business being in the market is a great “mask” for larger client
orders because it confuses the sniffers and thus buys some crucial extra
seconds or minutes for the original customer order.
It is into this growing agency environment
that we are launching a Tape, and while I laud the enterprise I have concerns
that it merely serves to further speed up the transfer of risk. Even if the
Tape is delayed by 15 minutes the customer will want the bulk, if not all, of
their order executed before the first trades hit the Tape and the pattern can
It is easy to criticise a new idea and I
don’t want that to be the impression from this column because I like the idea
of a Tape, as I stated in this
column earlier this year. I am just not sure the proposed model is the end
I have long argued we need a better
high/low mechanism in FX and a Tape provides that. I also have been of the
opinion that we need a better measure of market levels than that provided by
Matching or EBS Market for option triggers and limit orders given the decline
in influence of those two venues – the Tape can provide that.
My concerns about transparency are very
real because I believe if proven correct they will negatively impact the end
user of the market – generally speaking the only people larger orders come from
Where I think the Tape can really help is
in highlighting market impact from orders (even if it does, as I previously
stated, actually increase that impact), providing a better audit mechanism,
and, potentially, offering the best solution for determining when orders and
triggers should be hit.
The latter is a tricky issue because a
delay in reporting is of no use to someone holding limit orders or option
barriers – they need to know the second they need to be executed. This can be
overcome by providing a feed that people can subscribe to that alerts them when
certain criteria is hit – for instance if a level trades three times in at
least one million units an option barrier breach is valid.
Yes, there will be the eternal complaint
that this means re-writing legal documentation, but surely this is an issue
that a body such as ISDA could be debating and, potentially, providing a
solution for? The cost of re-writing legals is too often provided as an excuse
for inertia in this industry, hopefully the creation of a Tape will provide a
great opportunity to do something about actually making the market a fairer
place for end users.
So it’s good the Tape is getting close, but
I still want to see it in action to be comfortable that it is not leaking
information that shouldn’t be out there.
Finally, one outcome of this innovation
could be quite worrying for those who have stepped away from the principal risk
model, for if a customer does see their order creating too much of an impact
they may go back to the risk transfer mechanism, and while there may be less
competition for that business there will be enough to maintain a good spread on
Total transparency is often seen as a good
thing but too often the people that want it least are the customers – shouldn’t
that tell us something?