Month-end rebalancing flows and the pairs they push around
Posted: Thu Oct 29, 2026 4:09 am
Fund managers who hold global portfolios need to keep currency exposure within targets. When equities rally in one region and fall in another, the value of foreign holdings shifts, and managers often rebalance by selling or buying currency near the end of the month.
The effect is easiest to see in the last few trading days, especially around the London fix on the final business day. Directional pressure can appear in the dollar against the euro, the pound, and the yen without any matching news.
The messy part is that nobody knows the exact size or direction in advance. Some analysts estimate it from relative equity performance during the month. If US stocks outperformed global stocks, a common argument is that dollar selling follows, to bring the hedge ratio back in line.
That estimate is imperfect and often wrong, but it provides a reason for some days to have a skew that others do not.
One caution: month-end effects vary from month to month, and the direction is not reliable enough to trade blindly. Treating the last few days as higher-uncertainty periods, with smaller size, is a more defensible response than betting on a direction.
Has anyone tested a simple bias, such as fading the month's trend into the final-day fix, on historical data for the pairs they trade?
The effect is easiest to see in the last few trading days, especially around the London fix on the final business day. Directional pressure can appear in the dollar against the euro, the pound, and the yen without any matching news.
The messy part is that nobody knows the exact size or direction in advance. Some analysts estimate it from relative equity performance during the month. If US stocks outperformed global stocks, a common argument is that dollar selling follows, to bring the hedge ratio back in line.
That estimate is imperfect and often wrong, but it provides a reason for some days to have a skew that others do not.
One caution: month-end effects vary from month to month, and the direction is not reliable enough to trade blindly. Treating the last few days as higher-uncertainty periods, with smaller size, is a more defensible response than betting on a direction.
Has anyone tested a simple bias, such as fading the month's trend into the final-day fix, on historical data for the pairs they trade?