Month-end flows: the unseen hand

[Fairman]

Month-end flows: the unseen hand

Fund managers who hold international assets rebalance at the end of the month. If US stocks outperformed European ones, for instance, they sell some of the winners and buy the laggards, and the currency leg of that trade shows up in the market on the last few days.

The pattern is not reliable on any single day. It is more like a small nudge that shows up in aggregate. But scalpers sometimes notice that the last trading day of the month has a strange character, with unexplained pushes around the London fix and a preference for sudden reversals.

A few things to check on a chart:

Does the 4pm London move on month-end days differ from normal days?

Do trend-following setups perform worse in the last two days?

Is the daily range wider than usual with less follow-through?

Anyone who tracks this over six months will probably spot the pattern. It is a good example of a flow that has nothing to do with technical analysis but still affects every technical setup.

One interesting wrinkle is that the direction of the month-end flow depends on how stocks performed that month, which is public information. A month when US equities beat European ones by a wide margin leaves funds with too much US exposure, and the rebalancing leans toward selling dollars. Estimating it is imprecise and nobody should trade it blindly, but it gives the strange day a plausible explanation.

There is also a calendar trap: if the month ends on a weekend, the flows are concentrated on the preceding Friday, and the odd behaviour shows up there instead of on the last calendar day.

What does the last day of the month look like in your journal compared with the rest of it?