The Last Look…

The dust has settled after IOSCO’s report on pre-hedging, so it’s probably a good time for a new perspective – after all, things could have changed in the seven months since publication…they haven’t of course, but perhaps the FX industry should now be thinking about how it goes forward with what is still one of the biggest reputational risks it is running.

I still talk to people in FX – and I want to stress this is only about FX, I still don’t care about fixed income! – that think there is nothing wrong in the pre-hedging arena; equally I talk to many who truly don’t understand why it’s not called front-running. The reality is, both views sit towards the extremities, when the solution is likely to be closer to the middle. There have been proven occasions when pre-hedging has led to better execution, my problem is that the benefit of that better execution only goes to one party.

Those that argue that pre-hedging is a risky proposition for the executing party are deluding themselves, yes, I can recall times when large orders didn’t move the market, in fact it went the other way, but that was in an era in which the entire fill was passed to the client, not just the last 40-60%. Oh, and don’t even get me started on the utter nonsense that is “hedging ahead of the Fix” – the window for things to go wrong there is even shorter, and the executing party is helped by all the spec accounts joining in to make sure the end investor pays maximum dollar!

The reality is that those pre-hedging are taking on minimal risk and extracting an outsized reward when compared to that amount of risk, and too many in the buy side oversight world are either wilfully ignorant of what is going on, or genuinely have no understanding – in which case they shouldn’t be in the that nice cushy seat.

The challenge with pre-hedging now is who takes it forward as an issue? Best practice bodies were waiting for IOSCO, which pretty much kicked it back into their court

But…that is not to say that what happens is 100% a bad thing. The reality is that when it comes to large orders, common sense dictates that they are executed over a longer window, and if there is a fixed price to be met, the dealer should not have to hold too much residual risk after that point in time. This sounds very much like pre-hedging I hear you cry – and you would be right, although where we might differ is on where the benefit would go. Here, I have an idea…

Previously I have argued that if a client order needs to be executed over a period of time, then that client should get a TCA report – and a fill – that reflects the total execution cost, not just the last 15 minutes or the point-in-time price delivered by the dealer. I still tend to think this is the way to go, after all, as I have also previously argued, pre-hedging is in reality the same as “working an order”, it’s just in the latter the customer gets the real price, not a partial, or made up one. That said, perhaps we need to think about establishing some sort of loose criteria around these orders that can benefit both dealer and customer, but in a more balanced way?

The key phrase in all the pre-hedging guidance seems to be “for the benefit of the client”, and this is where the conundrum lies. On one hand, how can it be to the client’s benefit when the pre-hedging moves the market 5-10 pips or more? It can’t. On the other, however, does the client get a better fill than they would have through risk transfer? Often.

How do we balance this? How about using all of the pre-trade analytics available? I could pick up any of a dozen pre-trade TCA tools available (there are more I know) and it will give me a decent idea of where the final rate is likely to be. Throw in a couple of pips to cover the tracking error risk, and you have a solution where the client will get a guaranteed rate, but one inside the likely risk transfer rate.

“Aha”, I hear you say, surely this is just the same as picking a dealer’s algo, you pay a couple of pips, get good execution, but you have to run the market risk? Well, no it’s not, because I used the phrase “guaranteed rate”. In other words, the TCA gives a price, the dealer commits to accepting that (plus a couple of pips) and therefore retains the market risk. It is then really a question of how well that dealer exits the risk – if it is better, then the dealer gets a bigger reward, if it is worse, the dealer loses out (but still has a cushion of an extra couple of pips.

There are risks involved in larger transactions, but I don’t think they are being allocated fairly

I understand this is not a perfect solution – does one exist? – but it would at least balance the field a little, and by using an auditable pre-trade TCA, it would avoid the chances of dealers putting extremely cautious rates on bigger deals. Perhaps they would use their own TCA – then we would find out which dealers really do think their analysis and execution tools are good! Of course, this would then lead to dealers inevitably trimming the mark-up, but if they are accepting a guaranteed rate then they must be happy to win the business and the client gets an even-better fill.

This model clearly favours the big internalisers and franchises, but the sad reality is that unless they are willing to take on sizeable risk (which few are nowadays), smaller players can’t handle big orders anyway, unless they too pre-hedge, and probably more aggressively given the need to externalise more of the flow.

The challenge with pre-hedging now is who takes it forward as an issue? Best practice bodies were waiting for IOSCO, which pretty much kicked it back into their court with minimal guidance (beyond clients being able to explicitly rule the practice out when executing their order). It is hard to establish criteria for the practice, because every trade is different, but perhaps we can consider establishing a process?

If best practice suggested that a client (and perhaps the dealer at the same time) access a pre-trade TCA tool to establish a “likely” final price for the trade being considered, this at least gives the parties a baseline projection, from there it is a question of how confident the dealer is in the TCA and what extra spread they are going to put on it?

For those that argue this would not be best execution because of the mark-up, I would simply reply, fine, don’t use this method, either use an algo and take on the market risk yourself, or transfer the risk – at what is likely to be a worse price. It is impossible to define “best execution” when every deal is different and executed in different market conditions, but one aspect of the process should be how much risk is being taken by each party, so that the reward can be allocated thus.

At the moment, the reward is all in the executing party’s favour, and that is partly the fault of some on the buy side demanding two yards gets done at top-of-book, which is nonsensical. There are risks involved in larger transactions, but I don’t think they are being allocated fairly. I am sure there are instances where pre-hedging has cost the dealer in the past 20 years, but they are very few and far between and miniscule compared to the times they got the benefit.

Dealers are running the risk, yes, but at a level that would make any insurance salesperson very happy. Perhaps we need to better reflect the level of risk being taken and into the bargain help protect the industry from future trouble?

Oh, and the really big question for the Pre-Hedger? Of course it is…(IYKYK)

colin@thefullfx.com