Quant Memo
Core

FX Fixings and Benchmark Rates

Many funds and corporations are contractually required to trade at a single official exchange rate set once a day, and that concentrated demand around the fixing window has made it a distinct, watched piece of market structure.

Prerequisites: FX Quoting Conventions

An index provider needs to convert the value of thousands of foreign stocks into a single currency to publish a daily index level. It can't survey every trade all day — it needs one official number, calculated the same way every day, that everyone agrees to use. That number is an FX fixing, a benchmark rate calculated from actual trading activity in a short window, and it has become the reference price that a huge share of index funds, corporate treasuries, and asset managers are contractually obligated to trade at.

An FX fixing is an official benchmark rate calculated from trades executed in a defined window — most famously the WM/Reuters 4pm London fix. Because so much real demand is required to trade exactly at the fix, that short window sees a concentrated spike in volume and, historically, a concentrated spike in the potential for manipulation.

How a fixing actually gets calculated

The best-known benchmark, the WM/Reuters 4pm London fix, samples actual trades and quotes across multiple venues in a window centered on 4:00pm London time — typically one minute before and one minute after — and calculates a median. Any single dealer's quote is diluted by using a wide sample across the window, which is the whole design intent: no one participant should be able to move the published rate by trading alone.

time around 4:00pm London ±1 min sampling window median of sampled trades = the fix
Sampling many trades across a short window, then taking the median, is meant to prevent any single participant from steering the published rate.

Worked example

A pension fund's index mandate requires it to convert $500 million of foreign stock sale proceeds back into dollars "at the 4pm fix" — a contractual requirement, not a choice of convenience. Every bank handling similar client flow that day knows roughly how much net demand is coming through the fixing window (say, a large net need to buy dollars against several currencies), and that concentrated, predictable order flow around a known time is exactly the kind of information a trading desk can act on ahead of the print — buying the needed currency slightly early, in size, and pushing the rate against the fund's fixing order before it even executes. That gap between what a client "should" have gotten and what the concentrated, predictable flow around the fix actually delivered is precisely what regulators investigated in the 2013 FX fixing scandal, which led to billions in fines.

What this means in practice

The scandal reshaped how the fix works and how banks are allowed to handle fixing orders: the sampling window was widened from one minute to five, banks built information barriers between traders handling client fixing orders and other desks, and a global FX Global Code now sets explicit conduct standards around pre-hedging and disclosure. Asset managers who don't want single-point-in-time exposure to a fix increasingly spread their currency conversions across the day (a "TWAP" execution) instead of concentrating it all at one benchmark moment.

"Trading at the fix" does not mean a client's order and the published benchmark rate are the same thing by default — a dealer executing a client's fixing order still has to source the actual liquidity to hit that rate, and how they do that (aggressively pre-positioning versus passively working the order) is exactly where the historical abuse occurred.

Related concepts

Practice in interviews

Further reading

  • FCA, 'Fair and Effective Markets Review: FX Fixings'
ShareTwitterLinkedIn