I was going to discuss a paper released by the Bank of
England late last week entitled A discrete choice model for large
heterogeneous panels with interactive fixed effects with an application to the
determinants of corporate bond issuance but not only was I asleep by the end of the
title, I was also taught never to discuss things I don’t understand – and I
don’t even understand what a “discrete choice” is!
want to continue last week’s theme of the relationships in FX, this time
looking at the choices around the make up of aggregators.
remains the preferred channel for liquidity consumers and sometimes this is for
the good and sometimes the bad when it comes to relationships. Generally
speaking most liquidity providers understand the days of “only you” in anything
other than a huge amount are over, so they are happy to be aggregated with
their competition and be judged on the quality of their pricing. That is the
The bad is
that aggregation has been used by too many consumers to spray the market, often
enough for their business to become worthless (actually revenue negative) to
One advance in the industry has been that if you are a
player who sprays the market your relationship is probably already toast, along
with your execution quality (unless your “best” execution policy is tailored to
your execution style, in which case it’s all a bit of a waste of time isn’t
Although it started with the banks, all LPs now seem to have
been through the process of analysing their flow to determine the value of
their relationships and they are having some frank and open discussions with
some of those relationships over whether they actually want to see their
business in the same fashion they previously have.
So the bad (in terms of liquidity consumer behaviour) seems
to have been taken care of when it comes to aggregation, but that isn’t the end
of the story.
also the question consumers ask of their LPs, “will you be there for me
in torrid times?” The recent evidence suggests not, and if there is any good
news in that it is that everyone has plenty of hard evidence to demonstrate who
the solid, liquidity partners are in this industry.
Of course, the consumers need to be realistic and remember
that the price may not be pretty but it is a price – and equally LPs need to
understand that most relationships are delighted to receive an ugly price rather
than suffer what too many apparently do – some or all of their LPs shutting
down. And it doesn’t matter, as some seem keen to claim, that you’re the first
LP back in – the damage is probably already done, both to the clients’ P&L
or execution quality (often both) and your relationship with said client.
I still have some sympathy with the LPs who prefer not to be
aggregated with the public platforms – that can encourage machine gunning and
multiple hits. I am also not sure that the ECNs in an aggregation stream should
be happy if an LP is getting hit on multiple channels – they may be the channel
the LP decides to cut.
This means that the consumer has to decide what sort of
trader they want to be – and aggregate accordingly. The fact that in events
liquidity seems to swing to the ECNs is a positive for the multi-dealer
platforms, but is it enough to make up for the lack of larger ticket
availability at other times? More pertinently for the signalling risk and
subsequent slippage larger tickets can cause?
The first, and most important, question when looking at with
whom one should populate an aggregator has nothing to do with any of these
issues in my opinion.
The key is how do the LPs use your liquidity? If they turf
it straight out, dump them – unless you’re in the instant gratification game of
course. After all, if you hit an LP and you know they’re going to hit the
market you can very easily place a bid somewhere and pick up a pip or two on a
regular basis – perhaps rather than latency arb it can be termed “chicken arb”?
More seriously, consumers need to ask direct questions of
their prospective LPs, starting with a request for detail around their risk management
strategy? What does the LP do if you give it five out of a total of 20?
Generally speaking this shouldn’t be a big issue, especially if the flow is
“natural”, i.e. not intended to generate short term alpha, but is it to some
liquidity providers? The consumer needs to know.
I stated last year that non-bank and bank market makers are
very alike, and further evidence of that can be seen in how the LP world seems
to be split into two general styles that straddle both camps. There are those
that are comfortable with risk (and I don’t mean for 10 seconds) and they don’t
run to the market every time they get hit – and there are those that aren’t and
The key to the risk style has to be internalisation rates and
anecdotally I sense some of the non-banks are continuing to build their rates
while some of the banks are plateauing and even going in the other direction.
The consumer has to know – is my LP a true principal in the market or just an
Although this will take some time to play out, I suspect
that at some time in the future we will see our top group of LPs defined not by
type of institution, but by their ability to prove that they not only stick
with their customers through thick and thin (and flash crashes), but also by
how they can demonstrate higher internalisation levels or, at the very least, longer
hold times. That is a real definition of a professional, efficient and credible
institution – and it is to be hoped that companies that fit this bill rise to