The FX industry is a vibrant, innovative place, but sometimes I think it forgets why it exists. This amnesia means as an industry, FX does not respond sufficiently on those occasions when people with no understanding of the nuances of the business suggest “improvements”.

I am not sure if I am reading too much into events, but so many news threads running through the industry at the moment seem to me to be idealising a totally transparent, all-to-all trading environment. This works in domesticated markets like equities, but it doesn’t work in global, institutional markets like FX – especially when we remember why we are here.

The latest instance of someone believing the FX market structure is wrong was in a conversation I had this week with a person who believed that best execution was “easy” and that we should be easily able to predict execution paths. 

It isn’t and we can’t.

Of course I understand that we have better data in the FX market now, there are some really good pre-trade analytics available, but all of these tools have one thing in common – they cannot predict the future. If they could the owners of that technology would be, as Alan Rickman said in Die Hard, “sitting on the beach earning 20%”.

Just as an investment prospectus covers itself by stating that “past performance is no indication of future returns”, so a pre-trade TCA tool should state, “past market activity is no indication of future activity”. Actually the clearest way to put it would be, “I know you sold GBP 200 million in a three tick range last year, but today it could cost you 100 points…or two…I don’t know”.

Where I think we are getting better is in behavioural analysis – and that most certainly can help a prospective execution, because humans and computers largely behave in a consistent fashion (even if they sit in the White House – where consistency consists of being random). But when it comes to previous data all it can do is give a trader a semi-educated idea of what may happen.

That of course is an advance on what we had before, which was to hit and hope, but I use the word “semi” because there are so many other factors that go into understanding market conditions. Even though automation dominates markets, there is still a psychology to them that can stuff up (technical market term) any execution strategy.

Which brings me to my main point – the FX industry needs to remember – in some cases understand even – that it exists for the end user in what could be termed the real economy. FX is different because it services corporates trading across borders and helps investors switch their precious investments around the world.

For these traders best execution is about executing a hedge efficiently without moving the market, not nicking one tenth of a tick on a trade. This means that the industry should continue provide bespoke risk transfer and risk warehousing to the appropriate clients – and the regulators should support that. A corporate or asset manager client doesn’t want to execute a hedge on an exchange, the order is too visible and they will get slippage on what should be an administrative trade. They want to spend minutes (at the most) on the FX hedging and they certainly want to trade to a specific date and do not want to pay the infrastructure costs of multiple connectivity.

It has to be acknowledged that FX remains largely a speculative vehicle, but I believe we have to get away from obsessing on that segment – that is a throwback to the days when market share was everything and flow quality came second or third to the prospects for a good lunch or a gong at an industry awards event.

It is this belief that FX is merely just another speculative vehicle for retail and institutional punters that is driving the argument for the all-to-all, totally transparent model. The industry needs to explain the vitally important work it does in enabling hedging much better than it currently does, so that the people likely to make decisions actually have the facts to hand – not a propaganda pamphlet.

Two years ago I wrote of my sense that several banks were moving away from quantity of flow towards quality and the intervening period has seen that start to manifest. There is more focus on the real economy segment to the detriment of the speculative, but I am not sure the industry narrative has changed sufficiently. It needs to, because the FX industry is changing and it needs to broadcast that fact.

To return to the subject of best execution there is a stark difference between the efficient hedge model that I am talking about for real economy clients and that in equities where, frankly, best execution is an oxymoron.

Not only do the trades making up an equities order often form part of the data for the TCA, which pollutes the quality of that analysis, but all they really do is hit the NBBO and care less about the market impact. In FX terms it’s the old story of the trader who sold GBP 25 million during the flash crash – starting at 1.2580 and ending with an average of 1.2150 – but because they hit top of book all the way down and the TCA report said they did their job, when in reality they contributed to the panic.

What this is really showing is a lack of imagination (which, in fairness, few allow you to use these days) and trading like a robot. It’s why automated trading has taken over to such a degree – people have comfort in adhering to the norm.

I would question whether being average – for that is what this is – warrants charging a fee?

On a related topic, I have just flicked through the last e-FX report from Greenwich Associates, and interestingly it suggests that e-FX volumes may have hit a plateau. There is decent data to back this assertion up, so it does seem that the progress of automation in the trading function has slowed to a crawl. 

One reason for this could well be customers understanding that mechanically hitting top of book with no thought to the consequences (market impact) is not the best way for them to operate. This in turn means that a lot of this valuable flow remains opaque by being executed in a risk transfer fashion into a big (and hopefully genuine) internaliser and therefore a whole cottage industry that grew up in the early years of this century is found looking for a new source of information on which to act in an often predatory fashion.

So while for financial traders there are undoubted benefits in total transparency of order (and I need to stress I remain a passionate proponent of transparency of action or conduct), for the people that the FX industry was really created for – the corporates and pension fund investors – the exact opposite is true, and I honestly believe that if this is better understood by some academics and overseers they might start coming to different conclusions about how they think the FX market should operate.

The market is not perfect – nothing is – but for the sector it is meant to service, and for whom it does such a critically important job, it is working just fine. That is the message that needs to get out.

Colin_lambert@profit-loss.com

Twitter @lamboPnL

Twitter @Profit_and_Loss

Colin Lambert

Share This

Share on facebook
Facebook
Share on google
Google+
Share on twitter
Twitter
Share on linkedin
LinkedIn
Share on reddit
Reddit

Related Posts in