The Last Look…

There is nothing like last look to get people fired up in the FX industry – as my mailbag for the past two weeks testifies – but I have to say my trigger is pre-hedging; and the latest GFXC paper on the subject bothers me greatly.

As is the case with the last look paper, the pre-hedging document is not a recommendation to change the FX Global Code, rather it is an explanatory note – and the fact it comes in at 17 pages probably highlights the complexity of the subject. My concerns are that firstly it doesn’t deal with the biggest pre-hedging issue out there, rather it introduces a new conduct complexity; and secondly that it could promote poor behaviour on the part of clients.

I’ve been around FX people for decades and naturally, given my history, I tend to engage more on the trading side of the business. The first time, to my best recollection, that I heard the phrase “pre-hedging” was around 2011-12 when I asked a trader how they coped with the massive flows at the London 4pm Fix in a one-minute window. I may be out by one or two years, but before then these words were not in the FX lexicon.

Don’t get me wrong, the practice, described in other terms, was going on – traders were selling and buying in front of large stops and, as history shows, there was plenty going on in the minutes leading up to the one-minute 4pm window. It was not, however referred to as “pre-hedging”.

So pre-hedging is really just a clever use of language to describe something that was already going on and there is a sense, as with last look, that too many people think the problem is solved with a disclosure. All I can say is it most definitely is not – furthermore, reading the GFXC paper only makes me feel the industry has to do a lot more about pre-hedging if it is to protect itself and give clear guidance to participants about how larger trades are handled, both at and away from the Fix.

The first, and biggest, problem in the paper from my perspective, can be found in one line, “Principle 11 is not applicable to confirmed fixing orders.”

I think we would find the results of a straw poll asking people with which order type they most associated pre-hedging, very interesting, because in my experience just about everybody associates pre-hedging with the Fix (a bunch of lawsuits and fines will do that). I am trying to work out whether the GFXC has its head in the sand over this issue, or has decided it lies in the “too hard” in-tray. I think it’s the latter, but that doesn’t mean we should brush this issue away, because there are nuances to the Fix that need clarification.

Under what can only be described as new guidance from the GFXC, there is no pre-hedging of fixing orders, only hedging of firm orders. The fact that this “hedging” takes place ahead of the actual trade execution and therefore directly influences the price the customer receives, means little or nothing apparently.

I refuse to believe that the FX industry is naïve enough to think it can get away with the argument that the executing party has market risk and is therefore justified in hedging ahead of the order. On one, maybe two occasions, (pre-)hedging of the Fix may go wrong, but given the (pre-)hedging doesn’t actually start until the order is known, the majority of times the final fixing price is rarely influenced in favour of the client.

We now have a situation whereby a fixing order is not “pre-hedged” it is merely “hedged”, and that surely introduces legal risk.

I should stress here that I continue to believe that the executing parties are in a bind because the size of flow is clearly too much for a five-minute window – hence why they pre-hedge according to clear analysis by their algo teams – but that doesn’t mean we shouldn’t try to improve matters, it means just the opposite.

We now have a situation whereby a fixing order is not “pre-hedged” it is merely “hedged”, and that surely introduces legal risk.

The Australian regulator ASIC is not perfect, but by most measures it is a reasonable market authority – it is neither influenced by politics, nor funded by the fines it generates. ASIC has looked at Westpac’s trading in OTC fixed income markets and decided that hedging expected risk ahead of the trade is criminal. Yes, it’s fixed income, but the FX world needs to pay attention to this – a rational regulator believes that hedging, or pre-hedging, a trade constitutes front running (and incidentally ASIC uses the Code as a benchmark for acceptable behaviour).

Westpac may win the day in Australian courts, but do we really want to go through the legal wrangle so often? To my mind, one of the greatest benefits from the FX Global Code should be helping firms avoid having to go to court.

To me the solution is simple – as it has been since the first time this whole sorry subject hit the agenda. If an order is pre-hedged, or hedged ahead of time, the client should receive at least the majority of the benefit resulting.

It is at this point that some of you raise the same objection as always – what if the pre-hedging goes wrong? First of all, as noted, it very rarely goes wrong and over the course of a year I am very confident that the balance is a significant positive for the executing party. Secondly, if it goes wrong, so what? The decision to pre-hedge should be a judgement call made by the executing party having analysed liquidity conditions and the likely impact of the order. If they get that analysis wrong, whose fault is that? When a trader decides to buy and the market goes down, who takes responsibility?

If a business is confident enough in its market analysis then over the course of a year they are going to be fine. Equally, I should stress that I suggested “the majority” of the benefit should go to the client, an executing party should be rewarded for doing a better job – I just don’t see why it should be all one way.

The second issue I have with the GFXC paper concerns how it could promote poor conduct on the part of the liquidity consumer. Consider the following:

I have a very large sell order to execute and my best execution policy states I have to ask four LPs, which I do.

Every one of those LPs starts to pre-hedge, prompting the market to tick lower, to the point that when I am actually ready to execute, it is 10 pips lower than arrival price.

The LPs claim to be able to offer me a tighter price because they have pre-hedged, although generally speaking proving it is a tighter price is incredibly difficult, but that price is likely based off a mid-rate lower than where it would have been had I not asked ahead of time.

The LPs are collectively using my information – and it could be argued it is to my detriment because they are impacting the market and sending signals left, right and centre to other players. More pertinently, three of the four are likely to make money out of my order merely because they took a position after they were privy to the information. Are these three going to buy back their pre-hedge immediately I tell them they didn’t win the deal? Not until turkeys start voting for Christmas and I stop detesting a certain London football club (which will never happen FYG) it won’t.

In these circumstances what do I do as a client? Personally I believe they should adjust their best execution policy to enable them to engage with one LP and hold that one LP accountable through a very transparent liaison before, during and after execution. One thing I most certainly should not do is ask multiple LPs.

But what if I had certain nefarious tendencies and sought to turn this to my advantage? What is to stop me asking for an indicative offer from four LPs, watch them pre-hedge, and then execute my sell order quietly with a fifth LP? Is this poor conduct? I believe it is, spoofing is still a bad word, but spotting this is pretty difficult, because each of the four LPs won’t know what is going on – they will think another LP won the trade and did a good job of executing the balance.

The above example is pretty simple, I accept, but it highlights the moral morass the industry potentially faces. Taking it back (sorry) to times past, it’s like a trader being asked to sell $800 million for a customer and that trader saying “the first 200 is for me”. It went on and was frowned upon – why is it OK now?

If we are to allow hedging in front of an order (and apparently sometimes it can be called pre-hedging and sometimes it can’t) then at least there should be an acknowledgement of the economics involved.

To me, the GFXC paper needed to better address the issue of the Fix, or perhaps more generally, what is to be done over the impact on the end price. Again, the executing parties are on a hiding to nothing with the Fix, because I am sure some would be very keen to pass on some of the benefits of the pre-hedging to the client, but the client won’t accept them because they want to Fix rate! This seriously has the makings of a comedy show about life in the markets, forget Industry (a BBC drama based on graduates in the trading room), it’s more The Office.

If we are to allow hedging in front of an order (and apparently sometimes it can be called pre-hedging and sometimes it can’t) then at least there should be an acknowledgement of the economics involved. Unless the price is adjusted for the trading ahead of the order it is hard to see a benefit for the client – which is at the core of the Code’s rationale for pre-hedging.

If clients choose not to accept some of the economic benefits of this practice that is up to them, but they should indemnify the executing party appropriately. Then, and only then, will we have some legal security. If a client won’t accept pre-hedging on a stop loss then they need to accept slippage – the execution can be TCA-ed so it should be handled in the correct manner – this should not be a one-way street.

Whichever way I look at it, the GFXC paper has failed in one significant area – rather than clarifying the issue, it seems to have clouded it even further. That is the last thing the FX industry needs.

colin@thefullfx.com

@colinlambertFX