As someone who has been a regular critic of the benchmark fixings, especially the 4PM London fix, readers may be expecting a resounding “hear, hear” following the Bloomberg report on “manipulation” of exchange rates. Well, sorry to disappoint you, but…
My problem has long been with the process, not just of the 4PM fix but of all fixes, specifically the lead time customers had to allow for placing orders. It is a nonsense, as I have stated before, that in this super-quick electronic messaging era, people are expected to submit orders 5 minutes before fixing, let alone the 15 minutes that some mechanisms demand.
But the claims being made against the industry in the Bloomberg story, which I have to stress the organisation should be praised for covering, suggest collusion, something that is easy to allege, but often harder to prove. Cutting to the chase, a bank trader could be seeking to push the market in a direction ahead of a large order at the 4PM fix, but there is no way that trader knows every order in the market.
True, if three or four banks are talking to each other ahead of the fix they can get an idea of the orders to be executed, and equally the asset management industry that blithely continues to use this flawed mechanism tends to behave in a herd like fashion so what one is doing several others could copy, but there is still the unknown factor. As an example, the stories are legion (I have one or two myself) of people executing what are considered to be huge orders, only to have the market not only soak the order up but continue in the other direction.
The point of this? You may have a customer looking to sell a yard of dollar-yen, but there could equally be someone you don’t know about, who is looking to buy a yard and a half. You can collude all you like, but if the larger order is outside of your circle such collusion is worthless – this is what makes the FX market unique.
Reading the Bloomberg story, I am more struck by the allegations that bank traders knew a client was going to sell or buy something, they adjusted their position accordingly and looked to close the adjusted position when the customer order was executed. This is not collusion, it is more simple – and serious – than that. It is front running, something that always has been and always will be illegal. Thus it is a breakdown (if proved) in oversight at some banks and has nothing to do with the fix itself.
I would illustrate my point with two similar ways a customer can execute a large order through a bank. On one hand they can use a bank-provided algo strategy, which executes against different streams of liquidity, but can, explicitly, be shielded from the bank’s own trading desk. On the other, they can submit an order for the fix, which goes to the trading desk.
Why? Surely these orders should be subject to the same Chinese Walls that exist in the algo world where it is the customer’s choice whether the order is exposed within the bank or not?
The professional way to execute large orders (and I am sorry to bang this drum yet again) is to use an algo strategy, one that comes with transparent (and immediate) TCA. The mechanism is the same as a fixing, it just operates in a fairer fashion and utilises modern technology the right way, it reduces the opportunity for inefficiencies or bad practice to seep in.
I also have one other problem with the collusion angle. We operate in a different environment from even two years ago. Several bank traders have told me over the past year or two that they are banned from talking to traders at other banks in light of the Libor scandal. That suggests that some of the practice alleged in the story took place (if proved) in a different environment.
I see the Bloomberg story as yet another indication that the fix is flawed and needs to be mended, but equally I have to stress that I don’t think it is about collusion as such, in the fashion seen in the Libor fiasco.
This story is also, sadly, yet another reflection upon how poorly the asset management industry has operated over the past decade or more, during which it has ignored the FX component of its business and is only now worrying about it when returns become more patchy. If there is some good news to be had from this sorry situation it is that these professional money managers will hopefully focus on every aspect of their business, including FX and operate in a more professional fashion.
For the FX industry the problem is more prosaic – it needs to fix a flawed process. I note that ACI has issued a release suggesting that dealers sign up as having read and understood the Model Code, which covers the alleged irregularities highlighted in the Bloomberg story. This confirms to me that what is alleged to have taken place is nothing new – ACI’s Committee for Professionalism identified it as a potential issue when it first penned The Model Code. That said, having people explicitly acknowledge best practice codes can not be a bad thing. Looking ahead, it would be a good thing for the industry if the ACI’s Committee for Professionalism – the body responsible for the Model Code – and the world’s FX committees looked at a best practice for submitting and accepting orders for the fix. As I have been saying for some years now, the process is broken and needs fixing.
So, collusion? No. A flawed process that allows people to transgress established rules and procedures? Absolutely. This is a loophole that has long needed to be closed, perhaps the Bloomberg story will be the straw that breaks the camel’s back. I, for one, certainly hope so.