London 4pm fix: the 60-second window that gets quoted all the time

[Shadow Trader]

The WM/Reuters 4pm London fix is one of the most discussed windows in FX, mainly because a huge amount of benchmark-linked client business is priced off the average rate during a short window around it. Asset managers who need to rebalance currency exposures place orders at the fix, and the banks who accept them need to buy or sell in the market around the same moment.

A few things that scalpers notice: spreads on EURUSD and GBPUSD tend to tighten a bit going into it, then widen immediately afterward. Price can accelerate in the final minutes and then fade right after. And on month-end days, the effect gets amplified, because rebalancing flows are bigger.

The structure of the fix has changed. Regulators clamped down on banks after the manipulation scandals, and the calculation window was widened from one minute to five. That spread the flow over a longer period, which reduced the sharp spikes. But the tendency for flows to cluster remains.

For short-term trading this means that a move between 3:55 and 4:05 London is not a normal market signal. It's a benchmark process. Entering a breakout there and expecting follow-through often leads to disappointment.

One more complication is that the "fix" isn't one number from one moment. The benchmark is computed from trades and quotes across a window, so the market has no single instant at which the number is struck. Traders who aim for the fix are aiming for an average.

For that reason, the price pressure doesn't appear at a precise second. It spreads over a few minutes on both sides, with the fade after the window often sharper than the build-up.

A scalper looking for edge around the fix should avoid assuming a specific second. Looking at the five-minute segments from 15:45 to 16:15 London time across several months shows whether any directional tendency exists. On many ordinary days, it doesn't. On month-end days, it appears more often.

Whether that's tradable after spreads is a separate question, because EURUSD spreads on retail accounts often widen right after the window closes.

Has anyone logged the average move in the 10 minutes after the fix on month-end days versus normal days?

[Amanhasnoname]

I kept a log of the window from 15:45 to 16:15 London for most of last year on EURUSD, split into normal days and the last business day of each month. On ordinary days I couldn't find anything worth trading. The ten minutes after the fix moved a couple of pips either way with no consistent direction, and the spread widening afterwards would have eaten any small edge.

Month end was different, though not as cleanly as people make out. In a little over half of those sessions, the move from 15:50 into the window reversed by a meaningful part within ten minutes of it closing. The rest of the time it simply kept going, and those were the days that would have hurt a fade.

In the end I stopped trying to trade it and use it as a filter instead. I won't open anything new from about 15:50 to 16:10, and on month-end days I don't trust any breakout that starts inside that window until it holds into the next half hour.