The Last Look…
Posted by Colin Lambert. Last updated: August 2, 2022
ESMA may not work quickly, but it is persistent and thorough – and for that reason the FX industry needs to follow the course of the regulator’s review of pre-hedging closely.
The review asks a lot of relevant and pertinent questions, hopefully all responses will be published, for as the authority itself recognises, this is a divisive issue and one fraught with legal risk.
If some clarity comes out of this review that can only be a good thing, however with the global nature of FX markets it is likely that any laws or regulation emerging will be treated as guidance outside of ESMA’s sphere of authority.
Whilst I am firmly of the belief that pre-hedging is rarely, if ever, to the benefit of the client, I do accept that often it is the client’s behaviour that pushes the LP to pre-hedge in the first place by refusing to accept either market risk or what is often a reasonable spread for the transaction.
The bottom line for me is simple: How it is in the interest of the client to push the price, albeit gently and without intent to harm, against them? We can argue all we like, but the fact is, the more trades you execute ahead of the transaction, unless it is as a passive maker on dark (bi-lateral) venues, the more you will influence the price – it’s economic logic.
Perhaps clients need to better understand this and meet the LPs halfway by, for example, accepting the need, on very large transactions, to work with one party only? It doesn’t have to be the same party every time, and a good job can be rewarded with the next large trade, and a bad job with a removal from the list of LPs engaged with on these transactions. One way of reducing the spread on a large transaction is not to spread the information around, thereby protecting the “winner” of the trade.
Clients need to actually accept that the LPs are taking on a huge risk quoting for these tickets, especially in competition (a secret shared is a secret squared), and need protection in the form of the spread – and a substantial one at that. Risk warehousing is not a free service.
This is, generally speaking, the crux of the pre-hedging issue; the client doesn’t want to take on any market risk but equally doesn’t want to pay too much in spread. Cake and eat it anyone?
In terms of what the cost to the client is, however, here is an example. I ask an LP what their bid would be for GBP 3 billion in Cable. Mid-market is 1.2230 and the spread is, for example, 100 pips, the LP offers me a bid of 1.2180. I say I will be back.
The quoting party tells me they have to pre-hedge and immediately skews to help build inventory. It also, very gently, starts to sell. I come back to the quoting party 15 minutes later and ask for a price. The spread is still 100 points and the mid is now 1.2210 – I sell at 1.2160.
Pre-hedging – and I have argued this before – is a potential cancer at the centre of the FX market – one that leaves too many people open to legal risk
That pre-hedging has meant I have lost a potential 20 pips in execution cost savings or, more prosaically, $6 million. How is that better for the client? Of course, the LP could say to the client upfront that they will offer a narrower spread if they can pre-hedge, but isn’t this just saying “if I work the order from now, I will get a better fill”? If so, would the client not then merely hand the order over to the LP for execution?
Before everyone starts shouting down my hypothetical example, today we publish the latest in our series of analyses of the month-end London 4pm Fix and, using Cable, the evidence is that executing over a 20 minute window (to a slightly different methodology it has to be acknowledged), I could have sold my sterling at 1.21461, whereas in a five minute window the price was 1.2169 – just about 23 pips difference.
I understand the mechanics of the Fix make this a special case, but if this was reflected elsewhere during the trading day, I am being shown that if I trade with the five-minute window the LP is taking the benefit of the (pre) hedging; whereas in the 20-minute execution I, the client, am getting the benefit. Remind me again why the former is beneficial to the client?
Away from what seems to be an obvious example of why there needs to be clarity and change on this issue, the ESMA paper makes a couple of points that I too have raised over the years, the most pertinent of which is what happens when clients who are opposed to pre-hedging start asking for prices in size with only occasional interest to deal? It’s not efficient and it weakens the relationship, but it would keep the LPs “honest” according to critics of pre-hedging – but then is this, in itself, ethical, or even legal? I would argue it is spoofing, but if ESMA decides pre-hedging is lawful what is left to protect the client beyond the decency of the executing party?
I happen to believe that the banks in particular have largely put their house in order on the treatment of orders, but the one burning issue is pre-hedging. It is a complex and tricky issue because of a lack of movement on both sides of the trade – the LP is enjoying the benefits (they call it risk, I respectfully disagree), and the client can point to tighter spreads or a benchmarked execution.
To me, both sides need to shift ground. First, use a Fix, and before everyone’s jaws hit the floor, I mean a Fix that is calculated over a length of time commensurate with the size of the order. In other words, let’s actually use some of this choice we have over calculation windows.
Alternatively, and this is where I think we have the reasonable foundation of a solution of some sort, share the market risk. For the client, actually talk to the LP about the order you are looking to execute, use pre-trade analytics and agree a reasonable level (based on what can be independent analysis) at which the order should be filled. The LP then goes and works the order. If the LP executes at a better price than the agreed target then the spoils are shared, if slippage is greater, the cost too, is shared.
This would enable clients to get the benefit of the LP’s skill in executing large tickets, and the LP will of course, get paid in a fee (algo or no algo). The LP also has the incentive of doing a better job if possible and the client has the incentive to keep their mouth shut to other parties (I refer you to the HP-Barclays fiasco a few years back – another instance that lost a dealer their job and facing potential legal action).
I honestly believe that we need an alternative process to pre-hedging if the industry is not to fall into yet another legal hell-hole at some stage in the future. ESMA seems to have a similar view because while the paper rarely mentions FX specifically, it talks of “last look” and a trading mechanism where “a trade has been agreed by the client and the liquidity provider, but some elements of the trade (e.g. price) will be specified at a later stage”.
Perhaps we need to embrace a new era in FX markets, where participants “work an order”, rather than “pre-hedge” it?
The fact is, pre-hedging – and I have argued this before – is a potential cancer at the centre of the FX market – one that leaves too many people open to legal risk. Things go wrong, it is inevitable, but when it’s pre-hedging that goes wrong either the LP compensates quickly, or runs the risk of legal action. More worryingly from my point of view, is that too often in these cases, individuals from the dealing/execution team are held accountable, which is unfair because they are between a rock and a hard place.
I actually don’t recall the first time I heard the phrase “pre-hedging” but I understood it was going on as a practice when I was still trading, which is a long while ago. The experience of it back then turned me against the practice, and I remain in that camp.
Clients need to re-think how or where they execute these large flows. They can use independent parties to check execution quality and share the market risk, or they can accept what seems a wide spread but probably isn’t, especially in current conditions.
If I am right and the banks especially, as the parties that most often execute these sizes of tickets, have put their house well in order, then perhaps we need to embrace a new era, where participants “work an order”, rather than “pre-hedge” it.
I think we need this change, because whichever way you look at it, the word “pre” means “ahead of”, or “before”. And from there it’s a short journey to “front running” as a host of legal sharks are no doubt keen to remind us.