WMR Shelves Plans for Cboe FX Data: Consults on Dropping Currenex Data
Posted by Colin Lambert. Last updated: May 22, 2025
Following an industry consultation, FTSE Russell, which operates the WMR FX benchmarks, including the critical London 4pm Fix, has decided not to proceed with plans to incorporate data from Cboe FX in the methodology; at the same time, it has launched a short consultation on removing Currenex data from the calculation process.
The firm conducted two consultations last year on including Cboe FX data in 10 pairs (AUD, CAD, CHF, CNH, EUR, GBP, HKD, JPY, NZD and SGD), but following feedback it will not proceed – mainly due to concerns over the potential impact on the representativeness of the benchmark due to different execution methods and the presence of curated liquidity.
The challenge for WMR in accessing market data is that its two primary sources for spot data, EBS Market and LSEG Matching, are both firm, all-to-all, central limit order books (CLOBs), whereas both Cboe and Currenex operate multiple models and a great deal of volume is executed in curated liquidity pools that are not open to everyone. The Full FX understands that the representative concerns were raised due to the smaller amounts available for trading on the latter two platforms, which could, potentially, have an outsized influence if a participant put through a large number of small transactions that impacts the market.
Such a move became more likely, when FTSE Russell conducted its second consultation in December 2024, noting at the time, “[T]he varying nature of the liquidity that is observed on existing primary platforms used by WMR, which operate as CLOBs, compared with the proposed Cboe FX platform, that is observed due to the separate and distinct trading protocols of each platform type and would require further implementation consideration.”
In its announcement, the firm explains, “Dialogue with market participants on the manner in which FX price formation occurs, and the potential subsequent impact of the proposed change on the WMR benchmarks and the wider market operations, has indicated that data sourced from Cboe, as described in the published proposal, could not be considered sufficiently equivalent to the current WMR Spot FX Benchmark primary data sources, which represent firm liquidity, using standard trading parameters, that is fully transparent and broadly available to platform participants,” it states. “These differences revealed significant practical challenges and complexities with incorporating the Cboe data into the WMR methodology, which may have diminished the utility of the benchmarks by influencing market behaviours.
“Taken together, feedback has indicated that this change may have reduced the representativeness of the WMR Spot FX Benchmarks by allowing curated transactions that are not sufficiently transparent to market participants and are subject to distinct execution methods to be included,” FTSE Russell adds.
Currenex Data in Focus
Given the outcome on Cboe FX, it seems inevitable that focus would switch to WMR use of Currenex data for three pairs as a fallback solution, CHF, EUR and JPY. Following what it says was a review of its governance, FTSE Russell is formally opening a consultation on the removal of Currenex data. The data will continue to be used during the consultation, but the firm says it is targeting 22 June 2025 as the date for removal.
“Following user feedback and a recent assessment of its operational criteria for trading platform inclusion in the WMR benchmarks calculation, FTSE Russell announces that it intends to remove the Currenex platform and transaction data from the calculation of the CHF, EUR and JPY WMR spot rate benchmarks,” the firm states. “WMR benchmark rates for CHF, EUR and JPY are currently calculated using FX input data from LSEG Matching, EBS and Currenex. FTSE Russell has undertaken a historical data analysis to assess the impact of removing Currenex data on the quality of the WMR benchmarks and concluded that there is sufficient transaction data from the remaining platforms (EBS and LSEG Matching) to ensure that the WMR benchmarks for CHF, EUR and JPY would not be adversely affected by the change and will continue to be representative of the underlying FX markets each day.”
The firm says feedback, which will be treated as confidential with possibly a summary published at a later date, can be provided to wmr.ops@lseg.com.
The Full FX View
Given the differences in trading protocols, it would appear to make sense for FTSE Russell not to include data in the Fix calculation that is subject to last look and is not for a suitable amount, but this process is highlighting a budding problem for the benchmark provider – finding a truly representative rate.
It has been well-publicised that the primary sources of data for the Fix, EBS Market and LSEG Matching, have seen their share of turnover decline sharply over the last decade and a half, and I suspect this was one of the reasons that FTSE Russell started casting its net wider in search of more volume on which to base such a critical benchmark. The results of the consultation, however, have laid bare the source of the problem – where else can the data be extracted from?
At face value, Cboe’s firm liquidity is a good starting point, but again, one of the core tenets of the Fix is that the data has to reflect the entire market, which means an all-to-all venue, which the firm venue is not. Aside from that, LMAX Exchange is a firm, no-last look venue, but again, going back to the representative requirement in the Fix methodology, it supports much smaller amounts.
It could be argued that even the smaller amounts represent a valid reflection of the prevailing market, especially given how average trade sizes have fallen over the last 20 years, but notwithstanding the surveillance efforts of the individual platforms, this raises its own problem. While some argue the current Fix is not easy to manipulate, it is very visible and this attracts, as we show every month-end, speculative activity that influences the price.
I accept this is not “manipulation” in the truest sense – these traders are using public market data and the short time horizon to their advantage – but I would argue that this activity diminishes the “representativeness” of the Fix. Equally, how representative is a calculation if anything north of 40% of what is large volume, is hedged ahead of the actual window? WMR is keen to stress that its benchmark is a reference rate, and it is correct – the problem is, as I have argued before – the market is not using it as such, it is trading it.
We find ourselves, therefore, in a situation where what is widely, and rightly, regarded as one of the critical benchmarks in our industry, is calculated off a diminishing liquidity pool. As an example, in October 2024, the seven FX committees to report FX turnover data showed that just over $2 trillion went through in spot every day that month. Using a generous interpretation of the EBS and LSEG data (neither firm breaks out CLOB volumes) at 80 yards, just 4% of spot volume goes through the two sources used to calculate the Fix.
As someone who has been, and remains, critical of both WMR and the wider industry for not looking deeper into the window length, even I have to admit that in this instance, the company is between a rock and a hard place. The data criteria established for calculating the Fix is sensible as it should be, we do not want to introduce too many variables, but is it representative? Ultimately, the answer to that has to be ‘no’, but there is no easy fix (no pun intended).
Of course, there is one way in which the representativeness could be improved – lengthen the window to allow more volume through the CLOBs to be included, but I won’t go down that rabbit hole…today!
