The foreign exchange
industry is a creature of habit and as several potential disruptors can
testify, changing participants’ behaviour can be a difficult – on occasions
impossible – task. This means that what I am about to discuss may not happen in
a long time, but that doesn’t mean it’s not worth raising one of the many
paradoxes in the industry now.
Most people agree that
liquidity is at a premium and a growing body of participants agree with one of
my regularly aired concerns that slippage and information leakage are becoming
bigger and bigger issues. Executing what were once considered normal amounts
into the FX market now comes with a health warning and how stealthy you or your
executing agent is in the public market can be directly correlated to the cost
How liquidity is
accessed has changed to quite a degree in recent years, the time of a direct
relationship with one or two banks is long gone, replaced by aggregation or
multi-dealer platforms. Some of the more advanced customers have also (finally)
worked out the benefit of a good algo execution tool, which had led to a
reasonable increase in their use.
The other aspect of
the business that has changed is that, in the majors, especially at top of
book, the spread merely serves as an indication there is a bid and an offer.
Even in the traditionally quiet early Asia session spreads in EUR/USD are
around the 0.2-0.3 of a pip mark, in USD/JPY they’re 0.5 and even those two
traditional devastators of a dealer’s P&L, Cable and USD/CHF are both
sub-one pip. In Europe and the US morning of course, they are even tighter.
These spreads are good
for pretty small amounts, however, and top of book is often viewed as an early
warning signal by some market makers on the look out for significant business.
This means that good, significant liquidity is at a premium more than ever, but
spreads are still razor thin.
So why is it that
liquidity providers on some platforms pay the brokerage where the liquidity
If a client uses
aggregation they pay a fee, if they are on an ECN they pay the brokerage (and
often the prime brokerage as well), and if they use a third-party supplied algo
guess what? They pay a fee.
The only time they
don’t pay a fee it seems is when they put liquidity providers in competition
for their business (which suddenly becomes visible and as such harder to turn
for a profit) meaning spreads are compressed even further.
Many are the occasions
I have had to listen to a salesperson at one platform or another bang on about
how clients on their venue generally get choice or 0.1 pip pricing in
reasonable amounts. Now, thanks to a conversation with a friend last week
during which this question popped in our heads, I can ask, why would the LPs
continue to pay that fee and is this model fair?
Some of this business
is non-correlated of course so there is real value in winning, and then
internalising, it. But as principal risk becomes more and more frowned upon,
and more firms head to the churn and burn of the agency world, where is the
value for a liquidity provider in paying what is still a hefty fee on some of
these platforms for business they are going to have to turn quickly?
I have argued for some
time that liquidity is mispriced in FX, just as credit was, and part of that
mispricing is that the user doesn’t pay to access it. When spreads were above
one pip in the majors and two-three pips in second tier pairs I understood this
model entirely, the LPs were being paid in that spread.
The rise of agency,
the (misguided in my opinion) view that FX is going to end up a fee-based market
and a greater acceptance on the part of participants that liquidity is no longer
a “given” should probably lead to a re-assessment of how and who pays for what.
We can carry on with
the usual “client centric” clichés and platitudes but the harsh reality is that
to most LPs the “clients” they are meeting on these platforms are nothing of
the sort – they are merely looking for best bid or offer from a group of LPs
and hitting it. Because they don’t happen to be professional traders doesn’t
make them a “client” in the modern era, for many have access to algos, have
execution teams or, more pertinently, are taking advantage of the
infrastructure that allows them to put even the smallest ticket into
from LPs I have asked about this in the recent past suggests that the platforms
where they are in direct competition, on which they but not the liquidity user
pays the fee, are generally the least profitable. Surely then at some stage,
LPs start to question the amount and quality of liquidity they pump down these
As I noted at the
start of this piece, changing habits is very difficult in this business, but I
wonder if a few of the platform providers with the asymmetric fee model
wouldn’t be well served by starting to have a conversation with some of their
clients about paying some sort of fee? For just as we have seen when extreme
events hit markets, the downstream liquidity provider is the first to be cut
off, and I was interested during the UK referendum results to note that more
business seemed to be going through the ECNs than the RFS models.
This could be because all
the users of the latter were scared off by the uncertainty, but it could also
be because the pricing wasn’t exactly crash hot as the LPs realised that
streaming to such a model was a dangerous proposition in such conditions.
Either way as the big introspection
among bank LPs especially continues, I wonder how long it will be before they
start to question the value of streaming their premium liquidity down some of
these channels. If that does happen, the venues in question will face a
brokerage hole that can only be made up by the user, which is why I suspect now
is a better time to be having what will be a difficult conversation, rather
than when the user knows you’re on weaker ground.
Liquidity is a very
valuable commodity, especially in the current event-driven environment. Isn’t
it about time the consumers started paying for it?