Stats › Currencies and commodities › The 4pm fix
The 4pm fix and what happened to it
At four o'clock every afternoon in London a number gets published for about 157 currencies, and a very large share of the world's index funds, pension schemes and custodians settle their foreign exchange against it. It is called the WM/Reuters fix, or the WMR 4pm fix, or the London close. It is the single most referenced FX benchmark on earth.
For years it was calculated from a one minute window, thirty seconds either side of four o'clock. Traders at the largest dealers sat in private chat rooms called The Cartel, the Sterling Lads, the 3 musketeers and the A-team, shared their clients' orders with each other, and pushed the number. The bill so far, across seven regulators on two continents, runs to billions in five different currencies.
In February 2015 the window was widened from one minute to five. This page is about all three parts: how the thing is built, what was actually found and fined, and what the research says about whether the fix works better now. The last part is the least satisfying and the most interesting.
The WM/Reuters 4pm London fix was calculated from a one minute window, thirty seconds either side of 16:00, using 61 one-per-second snapshots from a single platform. On 15 Feb 2015 the window was widened to five minutes, 2 minutes 30 seconds either side. Regulators fined the banks in waves: the FCA £1,114,918,000 across five banks on 12 Nov 2014 and £284,432,000 on Barclays on 20 May 2015; the CFTC over $1.4bn then $400m; the OCC $950m; FINMA CHF 134m from UBS; the DOJ $2.783bn in criminal fines on 20 May 2015 with parent level guilty pleas; the Federal Reserve $1.8bn; and the European Commission EUR 1.068bn in 2019 and EUR 344m in 2021. On whether widening helped, the answer is mixed: the volume spike went away, but an FCA occasional paper found the benchmark's robustness gain was under 1 basis point while tracking error rose more than fivefold.
Members only
The rest of this page is for members
Below this point there are 9 sections, 2 tables and 12 named sources, roughly 3250 words of it. Every figure carries the source it came from and the date the data is from.
You can keep browsing every statistic in the library for free. The intro and the summary are always open.