The events of the past couple of years have
taught me to be wary when making statements about early investigations into one
practice or another, but obviously the news that a
US authority is investigating spoofing in FX markets needs addressing.
It is easy to dismiss yet another
investigation as yet another fund raising exercise by a regulator with political
ambitions but this is a little different, not least because the voice broking
businesses this is targeting are not exactly flush with cash – as I recall the
fine for RP Martins over Libor had to be reduced to save the firm going out of
What is different about this is that it
looks at a very interesting area of FX markets and at a practice that looks
strange to outsiders, perfectly natural to market participants, and is, quite
frankly, vitally important to many participants.
To start with, why would a broker “spoof”?
As my colleagues Galen Stops and Paul Gogliormella reported yesterday there are various
definitions of spoofing even within US regulatory circles, so actually pinning
the charge down on one practice is difficult. That said, the process a voice
broker goes through to build a price involves heavy negotiation and, if a word
or two is missed, could be seen as spoofing of a kind.
What do I mean? Well it’s quite simple. As
long as I can remember, a good forwards or options broker would come in early and work out
what they thought the “run” was. Bank traders, as they arrived would often ask
what the broker thought, run the numbers and, the majority of the time, accept
the run provided.
In return the bank traders would provide pricing in that run – in emerging markets FX where the New York Attorney General’s investigation
is focused this is normally in the one, two, three and six months.
It could be that outsiders see the broker’s
initial work as spoofing because it is providing prices that can’t be dealt
upon, but that could possibly influence the market.
That would be wrong, however, for everyone
party to the discussions knows the broker is providing an indicative run.
Therefore I would argue that is very much not
Equally, a broker may have a run in the
major tenors, but traders often want off the run prices, in dates that simply
aren’t priced. Often this is because a customer – a real end user – has
requested that date.
In this case the traditional process of a
voice broker is to work out what he or she thinks the price is (it’s quite a
linear process), suggest that as a level and then work hard to get a firm
price. Part of this could be telling customers that “there may be an interest”
in the specific date to see if they can entice a firm price from someone else.
Again, this is part of the process of
negotiation and is not, in my opinion, spoofing, rather it is a genuine
attempt to get a firm price in a tenor that does not have natural liquidity.
A third area that could be involved is the
broker tightening up a price that is already quoted. As an example, one trader
could have an offer at, say, 150 in the three months, while another has a bid
at 130. If one trader tells the voice broker that they “may pay a little
higher” or may “sell a little lower” the voice broker has something to work
with and can strive to bring the two closer together, or get another party
Generally speaking from my memory the
broker’s line would be something like, “150-130 in the three months may have
interest inside on either side”.
Again, this is not spoofing, there is no
intention to deceive, merely an intention to do what the voice broker’s job is
– bring two parties together through a process of negotiation.
Whatever the AG’s office in New York is
looking at, it must understand the critical role that negotiation plays in
these markets. The totally transparent, exchange type model simply doesn’t work
in more bespoke markets. Players are not going to stream four and a half month
USD/THB for example. As for options, the banks don’t like quoting publicly on big
tickets in the G10 so they certainly won’t do so on emerging markets.
If this process is banned or found to be
illegal the consequences for the end user are serious. Interest would
have to be put in the market and be there for all to see. Risk transfer would
be available but would be on much wider prices than are currently provided –
all of which means significantly higher hedging costs for corporations in
But what if there have been instances of
I have been pondering if this could happen
and it can, in one way deliberately and in another accidentally.
At the most blatant end of the scale, a
broker, using the hybrid model, can simply post bids and offers to their screen
that they have no intention of honouring. Equally, they could, in theory at
least, report trades (orally only) that have not occurred. They could get
away with posting rogue bids or offers by stating, every time someone hits a price, “sorry, price just
dealt, was about to tell you”. Alternatively, they could make a pretence of doing
the trade, only to come back with a refusal on lack of credit line grounds
(which is a perfectly acceptable reason for a trade not getting done of course).
That’s all very well, but that broker would
very quickly get a reputation amongst customers and peers for not honouring
trades and would not be getting brokerage – their lifeblood. Effectively they
would be self-regulated out of the market.
Another, much greyer area, brings together
all of what I have described above. A broker may not have a price in a certain
date and product, but may well be aware of pricing in futures. They could then
quote a price because they are confident they can execute the trade through the
Occasionally brokers are hit, can’t get out
of the trade because the futures have moved, and “wear the trade” until they
can get out of it – sometimes at a loss. This is not, I would argue, spoofing,
rather it is the acting as a principal in the market for a short time. As to whether that is OK or not is not something I really have an opinion on, but I would point out it is
why “points banks” existed in some centres, albeit to help facilitate
trades when credit wasn’t available.
So this is a very interesting
investigation. As I argued, taking away the ability to negotiate would
seriously impair genuine corporate hedging activities, but in the current
environment everything needs to be looked at.
The bottom line for me is simple, however.
In the cases of high frequency spoofing or quote stuffing, it is clear there is
no intention to deal. A voice broker cannot operate that quickly so knows they
are more vulnerable to being hit – especially with futures markets more
Everything then, comes down to the
“intention to trade” argument and I think is something that is pretty simple to
settle. When I discuss last look I often point to a voice trade as an area that
rarely, if ever, sees rejects. If two names can’t do each other the broker
works their socks off to get another name.
So defining “intention to trade” should be
easy in this case – did the brokers do the trade? If they did, it’s not
spoofing; if they refused on genuine credit grounds, it’s not spoofing. If they
were rejecting trades without good reason?…here we go again.
Colin_lambert@profit-loss.com Twitter @lamboPnL