column two weeks ago on benchmark fixings prompted quite a response – to
the degree that I feel it is necessary to go into it a bit further and ask the
question, “Can there be a justification for offering mid-rate benchmark
execution for free?”
The question is especially pertinent when
one considers that such an act flies directly in the face of the
recommendations on benchmarks laid down by the BIS working group less than two
One thing that does need clearing up before
I go any further – several sources are suggesting that a non-bank provider is
one of those firms offering the mid-rate on the 1pm Bloomberg fix. I have yet to
get the stage where this can be verified with enough confidence to publish a
name but I would point out one fact. The working group recommendations
(as well as the global Code of Conduct) are aimed at all market participants and not just banks, and as such, if a firm
is found in contravention of the recommendations, a defence of “we didn’t know”
will not stand up.
Of course there remains the question of
what the relevant authorities can do if they find such open defiance of their
wishes (and I am told – again by multiple sources – that a bank is also giving
this fix away for free)? There is more than a little interest in what the
response will be. One correspondent even suggests we could see a “test case”
scenario that will either demonstrate that defiance of the recommendations is a
costly exercise or, more worryingly perhaps, the industry will come to see the
recommendations as toothless.
Either way, the majority opinion in my
anecdotal pool of evidence tends to agree with me that the industry is at risk
if such behaviour continues.
But is it really bad behaviour?
I ask this because one area guaranteed to
generate paranoia amongst bankers is the thought of making too much money out
of an agency business. There seems little doubt that some providers are able to
fairly regularly beat the benchmark – this has led to the prices charged for
benchmark execution being cut from what they were just a year ago.
Generally speaking we started out as an
industry at the $25 mark, we now appear to be in the early/mid teens with one
or two just below $10. This is a reflection of the performance of the TWAP
algos and, perhaps, better netting, and sources familiar with the matter tell
me that the relevant authorities are comfortable with such a trend – as I
pointed out in the earlier column, $25 was not a line in the sand or even a
direct recommendation, it was merely the starting point for the industry.
But here’s the problem – what does a
provider do if it is beating the benchmark and
charging a fee? It cannot – and this is where I think the buy side remains
culpable (as I have since this whole sorry mess started incidentally) – pass
the improvement onto the client because the latter only wants the benchmark.
They have little or no interest in tracking error, even if it is positive.
To my mind this was, and remains to this
day, a disgrace. That a so-called professional money manager will deliberately
shun better performance stuns me, and not only that – assuming they are being
shown the data by the provider – they have proof that a simple TWAP beats the
benchmark, so why not use that? By all means execute at the benchmark time, but
perhaps they should be taking the improved performance of their service
By not doing so the benchmark user is
creating a difficult situation for everyone – and putting the service provider,
potentially, in harm’s way with the authorities.
That said, I still have to question the
wisdom of not charging anything for a benchmark execution, because as noted two
weeks ago, there are latent costs to running any business. I note with interest
that the FIA’s Professional Trading Group in an oral submission to a US
congressional committee last month, cited the “significant expense” involved in
developing source code.
I understand that TWAP algos are simple egg
timers but there are still development costs – and last time I checked good
developers didn’t come cheap. Throw these in with the simple cost of office
space to house equipment and the personnel to oversee it and there has to be
some sort of fee involved.
Of course, the outperformance of the
benchmark may run to the sort of numbers that covers even these costs, in which
case I refer the reader to my earlier complaint about users accepting benchmark
execution as best execution!
To me there are three possible ways to
solve this issue. Either users of the benchmark accept out-performance and pay
a reasonable fee; the authorities find a way to financially penalise a provider
who defies the recommendations; or they nuance the recommendations to accept
that there are times a provider will beat the benchmark and as such they are
entitled to keep the revenues – provided they executed according to the rules.
In other words, if they do keep getting hit on the bid or offer and don’t have
to cross the spread that is OK, but the integrity of the algo must remain above
reproach and within the rules.
My concern with the latter course is that
while some providers may be able to beat the benchmark – and I do not accept
they do so every time – not everyone will. Assuming that they are using a TWAP
as they should, it is dangerous for a provider to think they will always beat
the benchmark. What do they do when they find they are matching, or slightly
missing it and have no fees on which to fall back?
If that happens they are then likely to be
in the same boat as those who accurately hit the benchmark, but have been
forced into mispricing the business rather than risk losing it (along with the
other business that is dangled by customers in return for cheaper pricing).
They will all be losing money for providing benchmark execution. In other
words, we end up right where we started, which is the start of a dangerous path
that no-one (I hope) wants to walk down.