Pre-hedging is a hot topic at the moment, not least because of the Mark Johnson trial and the possible ramifications of the jury’s guilty verdict, but what happens when pre-hedging goes wrong?
This question was raised during a very interesting Insights call on Thursday (my thanks, as always, to those who joined in to share their views and ask some pertinent questions), and is something I’d like to go into in more depth here.
The initial question was asked, ‘if pre-hedging a Fix order, should the bank pass on any price improvement?’
My reaction was immediate – the bank is pre-hedging the order on behalf of the client, therefore any improvement should be offered to the client. I also noted how many index tracking asset managers (disgracefully in my opinion – they are there to maximise investor returns) refuse to accept any improvement offered because it causes Tracking Error. Aside from my eternal pursuit of managers pursuing such an unprofessional path I added that I firmly believe a client should be offered the improvement because, after all, the executing party would not have been buying/selling were it not for that order.
As always, however, it’s not that simple. As several attendees on the call were quick to ask, ‘what happens if the pre-hedging goes wrong? If another, larger, order comes in the other direction?’
This is a little trickier I accept, but I think the answer should be the client accepts the loss. We are really only talking very large orders being pre-hedged, therefore the number of occasions a larger order takes it out is probably pretty low (it does happen though) and as such – from the client’s perspective – the overall experience will be positive.
Of course, if the client only has one or two orders a year that require pre-hedging ahead of a Fix that is less certain, but again, the probability is pre-hedging will improve the execution, not least through reduced market impact.
I understand that in the new world of the Fix the five minute window and the approach taken by the banks to use a TWAP algo has rendered this argument relatively meaningless, but I am still told by dealers of larger Fix orders that require pre-hedging because even in a five minute window they are likely to cause market disruption and huge slippage (I am told of one asset manager who is still nursing the wounds of a 15 point drift on a very large order just a few months ago).
So, in spite of the reformed benchmark Fix guidelines helping improve matters, there still seems a need for smoothing the impact of some orders at the Fix and it is certainly the case for orders outside of the Fix – but is that even pre-hedging in the case of the latter? And what happens when it goes wrong?
To answer the first question, yes it is if the order is at a fixed point in the market. As for the second, I think it depends on the relationship between customer and bank. Given the need for transparency with the client, the latter will be aware that pre-hedging is taking place and would have agreed to it. If that is the case then the customer has given directions that pre-hedging is acceptable and should accept the consequences – they will take an improvement, so they should, in all fairness, suffer any impairment.
I suppose a spanner can be thrown in the works if the pre-hedging decision is taken resulting from advice from the bank. What happens then? There are some who believe this changes things a little and that the bank should wear the loss, but I don’t agree. Banks often give poor advice – not deliberately I should stress – because they are not omniscient. Their advice constitutes the best estimates of their experts at any given time. I can’t think of a field where experts are never wrong – but it may exist, latency arbitraging for example (sorry, couldn’t resist it!).
If a firm produces research on a company that prompts a client to buy stock and it turns out to be wrong, is there a comeback? Absolutely not, because that is the nature of a market. In that example, of course, there are plenty of disclaimers at work and the same will probably become the norm in foreign exchange around pre-hedging. They will say something like, ‘pre-hedging advice represents our best available assessment of market conditions but may be completely and utterly wrong’ – I am sure the lawyers will, for a sizeable fee, make that read much more smoothly.
Notwithstanding that, greater clarity on this issue would be welcomed. Part of the problem that led to the Cartel’s activities allegedly crossing a line was that the dealers felt they were put in harm’s way by their institution around the Fix in particular and their answer was pre-hedging. I have spoken to people on the periphery of this episode and they are all of the same opinion, broadly expressed as, ‘we are running the risk of the market going against us by pre-hedging, not the customer’ and it is this that I think probably needs to change.
Just as I found it reprehensible that parties on all sides wanted to re-paper some trades but not others during the SNB event, I find it equally distasteful that people on both sides want the other to assume the loss but pocket any profit from pre-hedging themselves.
There is no room for an asymmetric approach to this, the client and the bank dealer have to be treated fairly and in exactly the same fashion.
I don’t think it was pushed enough by the Johnson defence team but market impact has to be considered more than it is. The client should be given all the information and the best advice possible ahead of a large order and can then decide whether to assume modest risk by allowing pre-hedging (which can be TCA-ed independently of course), or accepting probable slippage during the Fix window.
Pre-trade TCA tools are available to help make that decision and they are getting better by the month almost as data quality continues to improve – and we should, as an industry, be using them more.
Ultimately, however, what we are talking about here is giving the client the choice between having the order worked by a professional dealer over an established window; or having the order executed in the five minute window at the Fix. I think we have enough data to show the latter often can result in costly and expensive FX hedging, but sometimes it will work.
Given our predictive powers regarding the flow likely at the Fix (especially at month end), if a client has a large order than will increase the total net at the Fix, recommend the client uses that mechanism. If it doesn’t, or worse still, is likely to increase the net imbalance at the Fix, execute it the old-fashioned way. It’s not like we can’t get independent verification of execution quality is it?
Generally speaking, I think the easiest way is simply to state that if the executing party is trading for the customer ahead of the order to minimise market impact the improvement or otherwise should be passed onto the client unless the client explicitly states, in writing, they do not want it. This can be called pre-hedging and should, as the Global Code states, be done with the client’s acknowledgement and acceptance.
If the trading is done with knowledge of the order but not as part of it, that should constitute front running and inappropriate use of client information – and punishments already exist for that. Either way, the executing businesses need to show definitively into which book each trade went.
Two final observations on this. Firstly, I think this ends up with more business going to the agency execution teams at the banks, and that means algos. As such, if a bank doesn’t have the appropriate suite of products in place they will, inevitably I believe, start to lose out on business.
Secondly, these are current issues because of events outside the industry’s control, but houses have been cleaned up. I noted earlier most of this is covered in the Global Code and these challenges (from the pre-Code era) only serve to increase my esteem for those that created the document. Yes, it could maybe do more in a few small areas, and yes, it may need more examples including to help people navigate grey areas, but overall the creators have done a fine job of taking out a lot of the conduct risk around day-to-day operations – pre-hedging being one of them.
But if I could ask one small favour? I touched upon it earlier and can’t resist it I know but it needs to be said – can we apply the same rules to last look? That is, unless agreed by all parties beforehand, any trading in the last look window has to be disclosed and if on any occasion it is not on behalf of the client, it is defined as front running?
My sense is the industry is also cleaning up its act on last look and more explicit disclosures that put all the information in the hands of the customer – as in the case of pre-hedging – could take a lot of angst out of the industry.
The reform process is going well, but no one should think for a minute it is complete.