Need to know

• Concerns about costs and predatory trading have prompted Refinitiv to launch a consultation on how the WM/Refinitiv 4pm fixing window might be reformed.
• Increased trading volumes during the window have resulted in predictable moves that critics say can lead to worse pricing for investors. Critics also claim that the actions of high-frequency traders have had a distorting effect on the fix.
• A European Central Bank-backed working group has suggested extending the window, and market participants agree that a longer period could be beneficial.
• However, if the window were too long it could lead to the kind of manipulation that was uncovered in 2013.
• Although, there is demand for an alternative, the 4pm fix is so entrenched that the cost and effort of changing benchmarks would be prohibitive for most firms.

For 27 years, the WM/Refinitiv 4pm fixing has enabled buy-side traders to value their portfolios and benchmark their foreign exchange trades.

But now, for the first time in its history, the operator of the rate – formerly known as the WM/Reuters fixing – has launched a public consultation on the way the benchmark operates. It comes amid concerns about high trading costs and the negative impact of predatory trading activities during the fixing window, which can lead to worse prices for investors.

Refinitiv is asking for feedback on the length of the calculation window, how frequently the rate should be calculated, whether additional sources of market data can be incorporated, and whether the benchmark could be extended to other currencies and tenors.

Andrea Maechler, chair of the Global Foreign Exchange Committee, welcomes the request for feedback. “It encourages market participants to consciously look at the associated risks and procedures of fixings,” she says. “Possible improvements should aim at promoting a robust, fair and appropriately transparent market for a diverse set of market participants.”

In the absence of an official closing price for the FX market, the 4pm fix was created in 1994 to give asset managers and index providers a rate that would enable their assets to be more easily valued and compared with those of their peers.

As both the number and size of passive index-tracking funds have grown over the years, so has the use of the fix to re-value assets at the end of a month or a quarter. Asset managers use the benchmark for their FX swaps and forwards hedges in order to reduce tracking errors and adjust the sizes of these positions in line with periodic valuation changes. They do so via fixing orders, in which the dealing bank is responsible for delivering on the fixing price in exchange for a pre-agreed spread. However, the fix is now also used as the settlement rate in a variety of derivatives trades, and as a general fair and transparent price on which to execute FX trades.

The fix originally comprised the median of all trades carried out within a one-minute window: 30 seconds either side of 4pm UK time. However, in 2013, several banks were accused of manipulating the benchmark by pushing through a large number of trades during its calculation window. Seven banks were fined hundreds of millions of dollars and more than 20 traders were fired as a result.

The episode prompted the benchmark’s operator, then known as Thomson Financial, to lengthen the observation window of trades from one to five minutes in 2015, and instead use the median of snapshots of trades taken at 15-second intervals to guard against manipulation. Yet criticism of the fixing continued, and it is this that has led Refinitiv to launch its consultation.

“User engagement and consultations are an important part of the WM/R benchmark review process,” says a Refinitiv spokesperson. “Receiving valuable insight from users enables us to respond effectively and continue providing WM/R benchmarks that are appropriate.”

One concern among users has been that the fixing orders that dealers need to hedge during the window are very directional. A presentation by Refinitiv to the European Central Bank in February showed that trading volumes across most currency pairs increase during the window. The volume spikes are also more pronounced at the end of each month, in all likelihood because of automated hedging flows.

Many asset managers assume that the impact on the market will be lowest when liquidity is at its deepest and trading volumes are their highest – during the window around the 4pm fix – and so they continue to use the benchmark regardless. Most of the FX trades that take place during the window involve swapping non-US currencies back into dollars.

The effect of a mass of asset managers placing one-directional orders was apparent at the start of the Covid-19 crisis. A study by Australian asset manager QIC showed that the AUD/USD rate on February 28, 2020 was around 0.655 at 2pm UK time, but dipped sharply in the half hour before the fixing window before eventually fixing at 0.645 at 4pm. It then rebounded to 0.655 at around 4:30pm. This implied that the fix was lower than would have been the case under normal market conditions. US investors would have lost out had they held AUD-denominated assets that were converted into dollars.

This was not a new phenomenon. A 2018 study by the UK’s Financial Conduct Authority found dealers had directionality of about 57% outside the window and 92% within it.

This has resulted in what some see as predictable moves during the window that can lead to worse pricing for investors – particularly at the end of a month or a quarter, when large passive index funds are rolling their swaps and forwards hedges.

These swings have also attracted high-frequency traders, which seek to take advantage of predictable movements during the window by selling into a falling market and buying as it comes back to normal after the fix is done. The 2018 FCA study showed HFTs only had a directionality score of 63% during the window and concluded that their trading patterns were consistent with different strategies, “such as market-making, going after short-term profit opportunities or high-frequency arbitrage”.

Critics of the fix claim this activity can exacerbate the swings seen during the window. Nathan Vurgest, head of trading at Record Currency Management, says: “The large and erratic market moves both in the lead-up and during the window… [are] due to activity which is not necessarily using the benchmark due to an underlying asset valuation reason and is merely trying to profit from the market moves.”

Christian Schoeppe, consultant and former head of trading at German asset manager DWS, agrees that, at certain crowded times, fast money is essentially trying to profit off the hedging flow of traditional asset managers, and that this is affecting the fixing.

“To traditional asset managers, they wonder, ‘is this still the 4pm rate?’, because the market rate in those times was different,” he says.

Record’s Vurgest says this has a double-whammy effect: not only do dealers charge more for fixing orders, given the extra volatility; there is also the danger that investors will end up with a worse rate. “The explicit cost may be X, but if the market has been moved by speculators trying to profit from WM-based currency moves, then the true overall costs could be much higher,” he says. “But that is very hard to measure.”

Regulators have repeatedly raised concerns about the fix’s impact on the normal functioning of markets, particularly at the start of the pandemic.

In the minutes of its February meeting, the European Central Bank’s FX committee – which comprises representatives of the largest liquidity providers – outlined its own concerns: “Owing to the large fluctuations in FX markets, end-of-month flows related to index rebalancing were expected to be very large and imbalanced at a time when banks had less operational and risk capacity. Since these flows would be concentrated at FX fixings, this was likely to create additional volatility and pressure on market functioning around the time of these fixings.”

The committee also floated the possibility of “extending the WM/R fixing window to ensure proper market functioning” – an idea that is being considered as part of the consultation. Refinitiv has not said that any particular window is under consideration, but market participants agree that a longer period could be beneficial.

Record’s Vurgest says that spacing out the snapshots and potentially making it harder to front run might help to resolve problems arising from increased speculation. “The benefit of any change would be that the WM rate may become harder to predict, and therefore dampen interest for speculators trying to profit from the window,” he says. “In theory, if it becomes harder to predict then less speculation will take place, and therefore less volume could lead to fewer erratic market moves and less market impact.”

How wide that new window should be is hard to say. One FX executive at a European-based dealer says that if it were too long, it would be harder to automate and more intervention from traders would be required, which could create opportunities for the kind of abusive behaviour seen in the past: “You’d go back to having a manual hedging process instead of an automated algorithmic process. That could result in the issues we had in the first place, such as fix manipulation.”

Furthermore, if the window were longer, firms would have to submit execution requests much earlier. This could pose operational challenges for asset managers in creating, calculating and submitting baskets of trades ahead of the deadline.

“If they [Refinitiv] use four hours as a window, then you have to submit the orders at noon,” says Schoeppe. “For a lot of asset managers, that is a massive operational change and effort which they don’t want to do. So, they can’t widen too much from an operational perspective.”

He also notes that although doubling the length of the window would make things twice as hard for speculators, it would not eliminate the existing challenges: “They will watch and monitor how the time and volume distribution would be during this new, say, 10-minute window. Then they will find mechanisms to speculate again, and you can’t prevent them from doing that.”

Needs mustn’t

Traders suggest any changes should tackle the issue of asset managers using the fix as an execution price not because they need to, but because they perceive it to be the most representative price for the day.

“The London 4pm fix appears to be a problem to those using it for a purpose it wasn’t necessarily designed for,” says Stuart Simmons, head of currency at QIC. “As a reference rate for marking to market foreign assets and exchange rate instruments, it performs well. As a central point of execution, investors should look elsewhere.”

Simmons says any trader asking a benchmark provider to change its methodology because the trader’s execution outcomes are poor should instead be examining its own execution processes. “Systematically rebalancing at London 4pm is the antithesis of best execution for any investor concerned about execution costs,” he says. “No amount of changes to the window will address the natural imbalance from the order flow.”

Record’s Vurgest says measures that made WM/R harder to predict could, in theory, cause banks to stop offering the rate as an execution service. This in turn would cause the FX industry to “move into a culture of not being focused on achieving a benchmark rate, and therefore… comfortable with accepting tracking error against the benchmark, as I understand to be the case in other asset classes”.

This, in turn, might also make the costs of execution more explicit and transaction cost analysis easier for asset managers – instead of the managers having small explicit costs and potentially large implicit market impact costs, which are difficult to measure.

Switching providers

According to the FX executive at the European-based dealer, there is demand for an alternative benchmark, but this is mainly coming from the banks rather than the buy side: “The banks know that there are still people out there who can estimate what the fixing would do and try to prey on that, so they are keen for ways to improve this.”

In theory, the rationale for switching FX benchmarks is clear and can be compared to the reasons for changing energy suppliers. In practice, however, it is not quite the same. The 4pm fix is so entrenched in investor contracts and derivatives that the operational cost and effort of changing benchmarks would be significant.

“The buy side could use other fixing times, and there are a lot of alternatives that they can use,” the FX executive says. “But they don’t, as the re-papering for their existing processes is too massive. They may be aware that they could save some money if they move to an alternative, but it’s too big for them to be able to take it.”

The fact that MSCI, the primary equities index used for so many passive index funds, and other asset valuations are tied to the 4pm rate means that many asset managers are tied to it too, whether they like it or not. Furthermore, others may have to accept that using another benchmark will run the risk of a tracking error.

“Until then, a lot of people will struggle to move away,” says Vurgest. “The specific WM/R rate is the most important factor to execution for many people, rather than the cost of execution.”

Rival in the wings

As Refinitiv looks to address users’ concerns, other benchmark providers are hoping to capitalise by offering alternatives. One of these is Siren, which claims buy-side firms could save millions in execution costs by switching to it.

Its administrator, Raidne, has called for an investigation into the execution costs of the 4pm fix after it completed an FX benchmark audit, which it says revealed potential savings from switching to Siren. Raidne says the savings came from the rate achieved during the window – in essence, that Siren does not move as far from the prevailing rate as the WM/R fix can.

According to the audit, using Siren instead of WM/R would have achieved savings of $1,491 per$1 million traded for AUD/USD and $1,164 per$1 million traded for USD/JPY at the end of April.

Raidne’s co-founder Jim Wyatt says that UK local government pension schemes cannot ignore the lower execution costs they might achieve by using a different benchmark: “If they do not manage this in a fiduciary responsible manner, for everyone who lives in that borough and pays council tax, their money is being wasted.”

Siren differs from WM/R by using a 20-minute calculation window. This has a twofold effect, its backers claim: the benchmark will be harder for HFTs to arbitrage; and it will be less susceptible to market impact, which will mean tighter prices for users such as pensions.

“Given the fact that Refinitiv have said, ‘we’re not sure it is right, we need feedback’, now is the time to just pause and take a moment to check, ‘am I losing more money than I should on FX execution?’” Wyatt says. “There are some people that definitely should stick with WM/R and there are some companies that ought to move to Siren. We’re talking with some of the major organisations with trillions of dollars under management, and their view is an optimised benchmark choice is the best for the client.”

Editing by Daniel Blackburn

Only users who have a paid subscription or are part of a corporate subscription are able to print or copy content.

To access these options, along with all other subscription benefits, please contact customer services - www.fx-markets.com/static/contact-us, or view our subscription options here: https://subscriptions.fx-markets.com/subscribe

Nomura hires Leng for dual global FX trading roles

Choose to follow a range of topics and receive a regular roundup of the latest coverage.