Benchmark fixes have been immersed in controversy for the past five years, but anecdotal evidence sees no shift in asset manager attitudes to them. Colin Lambert asks, will these firms ever desert the Fix?
If there has been one lightning rod for controversy in what has been a pretty turbulent period for the foreign exchange industry it has been benchmark fixes. Banks have been fined, traders and managers have been dismissed, and some are facing legal sanctions, including jail, thanks to various activities all of which were centred on the WM and European Central Bank fixes.
Throughout all this, the end users of the fixes, typically corporates and particularly asset managers, have averted their eyes from the controversy, supported the changes to the methodology, ensured their bank service providers had ring-fenced their fixing business, agreed (grudgingly perhaps) to start paying a fee for the Fix again – and carried on regardless.
Most asset managers are aware of the issues surrounding the Fix, but they are reluctant to read too much into it, as highlighted by one European based manager who speaks for many when saying, “It was a few bad apples that took things too far – the fixing mechanism works fine, it just needed more controls and structure around how orders were handled.”
In some quarters this view is seen as being wilfully negligent, in others as a pragmatic approach to a temporary situation – which has now been resolved. The critics of the Fix say that imbalances create slippage and execution methods create signalling risk, both of which mean the cost of execution rises. Proponents, as noted earlier, think the Fix was temporarily vulnerable to bad behaviour, which has now been eradicated.
At the heart of the issue lies another tricky subject for the foreign exchange industry – pre-hedging – which makes it impossible to ignore when discussing the WM fixes. “It is difficult to judge the impact of pre-hedging,” explains a senior FX salesperson with experience of handling large fixing orders. “I would argue that it helps smooth the market impact of a large enough order and as such gives a better outcome for the client, but the only way you can measure the benefit is in volatility because every Fix is different.
“Pre-hedging creates a move ahead of the order but means the window itself is less volatile – and that’s when most eyes are on the market,” the salesperson continues. “Alternatively, you ram the whole order through in five minutes, according to the methodology of the Fix, and get a sharper, possibly more violent move in the market. The question to ask is, ‘do you want a longer, smoother execution path or a quicker, sharper one?’ The outcomes are likely to be broadly the same, it’s just with the window-only method you run the risk of the market not absorbing the order well and the market impact being exacerbated.”
Speaking to Profit & Loss recently, one of the former bank traders caught up in the Fix scandal explained, “We had to prehedge the Fix, you simply could not buy the amounts in question in a one-minute window and, frankly, it is often a silly thing to try in a five-minute window.
“Did we ask questions about the profits from pre-hedging? And I want to stress it did not always go well. We did at first, but after you are told ‘they’re not interested’ by your sales desk when you mention an improvement of a pip or two you kind of give up asking.”
Perhaps naturally given their proximity to the issue and the impact upon their career, the former bank trader also believes that when the revelations about activity around the Fix finally came out, the world picked the wrong target. “So someone wakes the world up to the money being left on the table by the customers, but rather than look at the asset managers and ask, ‘why are you not looking at this?’ they looked at the bank desks and found an easy target.
“The chat rooms were not a good idea, I accept that, but you get sucked into it like a good trade idea,” the former trader continues. “There is a fear of missing out which drags you in. The chat rooms were no different, you felt you had to take a look – often your manager is demanding that you do – and once you do that, you’re in.”
Putting aside the allegations of collusion surrounding the bank traders, it does seem a valid question to ask of the asset management community in particular – if you know this type of behaviour is going on, and that the Fix is still vulnerable to (legal) front running by independent third parties, why use it?
The answer is almost universal. As one Asia-based manager says succinctly, “It’s the legals.” Pushed further, the manager explains that asset owners and trustees have embedded in their legal documents that currency hedging of certain assets has to be via a transparent and public fixing process – predominantly the WM 4pm Fix.
This is a common refrain from the asset management community, many of whom believe trading away from the Fix would improve returns and, into the bargain, help them demonstrate their value. “You think I don’t want to show my value with this business?” asks an execution trader at a European fund. “I think a good execution desk can make a difference and the Fix doesn’t allow us to do that, so I would love to get this stuff away from it and into my hands – it’s not going to happen though because there are risks attached that passive investors in particular don’t want to take.”
There is also one other issue that managers seem universally agreed upon – it would cost too much to change the “legals” on all the accounts held by their firms. “How would an investor take it if we spent a good proportion of their already volatile and fragile returns on changing legal documentation on an administration trade?” asks the Asia-based manager. “That’s a hard sell at the best of times, especially when they don’t know how much the Fix costs – no one does.”
There, as they say, may be the rub. Empirical analysis exists of how the market behaves at the Fix and how high signalling risk can be, but little exists around the likely impact of executing at another time of day – the argument put forward by critics of the Fix. As with all analysis, there is a basis of assumption that could break down in the first instance, or failing that at any time in the future, with potentially disastrous consequences for those who made what would be seen as a brave decision to plough a different furrow.
“We have presented repeatedly to trustees and asset owners on the projected benefits of moving away from the Fix,” says a senior manager at an asset manager in the US. “They don’t want to know. The problem is we can talk about market impact and the expected execution cost of trading at, for example, 10am New York, but we are dealing in theoreticals. To win such an argument you need hard data and while we can show the 4pm London Fix in particular is open to higher execution costs, we can’t prove that any other time of day is better. If we could execute the orders for one or two days away from the 4pm that might help sway the argument but the mandate doesn’t allow it – it’s a Catch-22.”
So is signalling risk actually a factor in the London 4pm Fix window? In the months after the methodology was changed in 2015, research from Pragma Trading highlighted how trading patterns during the new five-minute window followed a “predictable pattern”, thanks to the changes implemented by the banks following the 2014 report from the Financial Stability Board on reforming FX benchmarks.
These changes involved banks taking Fix business away from the trading desks and either giving it to a dedicated – and separate – team; taking dealers off the desk to execute the orders; or, most often, handing responsibility to the algo execution teams. The problem this created, as highlighted by Pragma’s research at the time, was that the banks all rolled out a “Fix Algo” strategy, which basically mimicked the WM process. This meant that at 3.57.30pm London time, anything upwards of eight TWAP (time weighted average price) algos entered the market at the same time and sought to execute in the same fashion.
In normal market conditions this could work, however at the Fix, with so much attention being focused on that fiveminute window, and thanks to the daily imbalance at the Fix, which can often be counted in the billions of dollars, it only served to signal to the outside world what was happening – leading to a group of traders effectively and legally front running the market.
“We created a strategy especially for the Fix, especially month and quarter ends,” reveals the principal of a hedge fund in Asia. “Our models could pick up the direction in the first 10-to-15 seconds and we would then take what could be a sizeable position for the next two-and-ahalf to three minutes. The information was always there about the Fix if you looked hard enough, but in the days when we used to see pre-hedging, you had to hold the position for much longer. In 2015, it became a no-brainer, four-minute trade.”
The hedge fund principal also notes that the strategy became less profitable as the market bedded the new Fix down, mainly, they believe, because some banks were already outperforming the benchmark and as such were executing “out of pattern”. This is again confirmed by Pragma which, one year later, published further research noting that although the pattern was by then of “no value”; however, there was still a discernible pattern, especially at month ends when trading volumes tend to be larger.
So while there does appear to be signalling risk around the Fix, there are now questions as to whether it is enough to force managers away from using the mechanism, and a straw poll taken amongst interviewees for this article would suggest that it is not. “If they didn’t care before when they were giving occasionally big money away, they sure won’t now when the amounts are smaller,” observes one banker.
There is also the question of whether the Fix is attracting more volume in the right fashion – and therefore whether this makes it even more attractive to the prospective users. Sources say that netting ratios have gone up in the past three years, but not by much, with one suggesting their institution’s netting ratio at the 4pm Fix had risen from 55% to 60%.
This may be a small improvement, but it is an improvement and highlights how confidence in the new methodology has attracted more flow, or as the execution dealer in Europe puts it, “the right flow”.
The flow in question is counter-trend activity, from traders likely to trade against the Fix flows. “We can provide data to clients on which way we think the Fix is going and this is leading some of them, corporates and regional bank customers mainly, to also leave orders at the Fix,” explains the European execution dealer. “It makes sense because they know their order won’t have market impact and they generally will see the Fix shift the price in their direction.”
This is not the only way that price moves are being dampened around the Fix, however, because although the outcome is the same, several customers are apparently looking to trade in the Fix window, but without submitting orders to an executing broker, according to other sources.
The European-based execution dealer adds, “There are customers who don’t want to pay the fee for the Fix and like the flexibility of being able to execute their trade ahead of time if the level suits. These clients, however, do have the information on what way the Fix is likely to push the market, and they have observable patterns of behaviour during the window and so they act accordingly. If they normally see a reversal 200-250 seconds into the Fix window they trade at the start of that reversal period. They get to execute their deal knowing the market has been working for them but without giving up information, which they hope will push it another half or one point, at no cost.
One regional bank trader who handles Fix flow on behalf of clients confirms this strategy. “We often encourage our antitrend customers not to give the order up to the Fix, but to execute it at some stage during the window. We think it gives them a better outcome because you can always see how fast the market is moving. If nothing is happening, trade just after the top of the hour; if it is moving then delay that by a minute. The differences are small, but the risk levels are also lower. A pip here or there can add up over a year.”
This is not to say, though, that challenges do not still remain. Occasionally things can go wrong, although not thus far – thankfully – to the degree warned about by former Bank of England FX head Chris Salmon when he hypothesised on the risk of a flash event during the 4pm window. That said, if a firm sees 10 pips slippage during the window that can be expensive.
“Our analysis shows that the Fix window is busier and more volatile, but not excessively so,” says the senior manager at the US asset manager. “We have had a couple of troublesome episodes, though, at the end of May we struggled to buy dollars, especially against sterling and yen, and we saw slippage. Our investors neither knew, nor cared about it, that’s the mechanism we use and they understand that, but it did make us look twice and think if it became a regular event then we would have to do something different.”
Thankfully for that asset manager, extensive slippage during the Fix does seem to be a random event with most months showing normal behaviour. With significant multiples of volume going through in the five-minute window (FSB analysis showed that volumes in the oneminute window were 10 times those of the average minute) the past few months offer an insight into market behaviour.
A glance at the last three month ends suggests that it depends very much on the imbalances evident at every Fix. The asset manager who saw the slippage at the end of May is right to highlight problems in Cable and USD/JPY, for both markets were busier (as expressed by the range) in the five-minute window than they were in the previous 30 minutes – indeed in the case of USD/JPY it was a more volatile window compared to the previous hour.
Away from those two markets though, the pattern seems to be that the Fix window is about half as busy in terms of the trading range, than the total range of the previous 30 or 60 minutes. There are exceptions to the rule of course, EUR/USD in June was as busy in the window as the previous 30 minutes (as was USD/JPY again), but generally speaking it is difficult to discern a pattern that suggests that the market is behaving in anything other than a normal manner.
Time for a Change?
It is hard to argue that the benchmark reforms implemented after the FSB report have not worked – by most measures they have. There is still the issue of the signalling risk and for that to improve then more counter flow needs to be attracted to the window – the challenge for the industry would be if the flow imbalance actually worsens, although as most service providers are happy to point out, it would be a problem for the customers, not the agents (welcome to the new world of foreign exchange where the buck stops with the customer).
Overall though, most participants seem happy with how the Fix is operating, subject to the touchy issue of the US legal system’s approach to prehedging being resolved. “Banks put their traders in harm’s way with the old Fix, but they have their structures in place under the new regime and, importantly, are still making money out of the Fix,” suggests the senior FX salesperson. “Not only are they legitimately charging for the most part, but they also have algorithms that are outperforming the Fix on a regular basis and so they pick up P&L there.
“Are banks nervous about making money at the Fix? Very, but as long as they explain it to the client, what’s the issue? You could ask the client how this sits with their best execution policy, however,” the salesperson says.
The latter is a standard refrain from those unhappy with the structure of the Fix, or indeed its use for anything other that what they say was its original intention – a reference rate. Ultimately, however, most users of the Fix go back to the standard response from that quarter – “it’s what my legals tell me to do”. Basically, if the Fix is identified as the framework for that firm to achieve best execution it does not matter whether it can be improved or not – it’s all about following procedures and the correct process.
Customers also seem unfazed by the events of past years, and until they are told not to use the mechanism, they will continue to use the Fix. As to what might prompt that move away, the senior FX salesperson has one theory. “I don’t think we will ever move away from the Fix unless investors or asset owners take action – and by that I mean legal action. Banks can point to the odd point they are making out of the business and the asset owner might look at that and ask why they are not getting it – especially if they are paying a fee. And it can add up over a year. Just one tenth of a point in EUR/USD on a EUR500 million portfolio would be over a million dollars a year – how does that sit with the claims of best execution and the fees charged the end investor to properly look after their money?
“In those circumstances, where the actual owner of the money starts asking questions, then we might see a change of mood,” the salesperson continues. “I think it is unlikely, it’s not as though the banks are making outlandish amounts out of the Fix. You hear people say they are just about breaking even on the flow, others that they are making a small amount. We are doing OK – it’s not a loss leader – but there’s nothing there for people to get excited about any more, it’s all above board, executed in a transparent and fair manner for which we get paid a fair fee.”
So while there are still sceptics around the Fix process and those who believe it should return to a reference rate only, those against the mechanism have one very serious problem. The lack of an alternative. The Fix is, to paraphrase Winston Churchill’s famous citation of a quote on democracy, the worst possible method to hedge currency risk…apart from every other method that has been tried.
Until a customer is willing to partake in an experiment that sees similar orders executed at the Fix and at another, time of day, we will never know whether it is better to move away from what remains the focal point of the trading day in FX markets. What is needed is hard, empirical evidence that a better world exists.
It can be argued that one aspect of execution quality at the Fix – signalling risk – can be proven, but either asset managers have chosen to ignore this factor or they just don’t think it is important enough. As long as there is so much attention on that five-minute window, and customers continue to try to push through what many believe is too much volume in too short a space of time, then market behaviour and execution quality around the Fix will continue to be changeable, and occasionally, challenging