Month-end and quarter-end rebalancing: flows without a release time

Where the flows come from

Multi-asset portfolios hold target weights, for example a fixed split between equities and bonds. When one asset class outperforms during the month, the portfolio drifts away from its targets, and rebalancing sells some of the winner and buys the loser. Many funds do this at month-end or quarter-end, which concentrates the flows in time.

Other month-end mechanics add to it: foreign investors resetting currency hedges on the value of their holdings, and bond index changes at month-end that can shift demand for longer maturities. None of these flows are published in advance, and their size is estimated, not known.

Why estimates are unreliable guides

• Different models give different sizes and sometimes different directions.
• Funds can rebalance earlier, later or partially, and some use futures rather than the cash market.
• When estimates are widely publicised, other participants can position ahead, which changes the effect at the close.
• A month-end without visible flows is common, and not a failure of the concept.

Where they show up

In equities, rebalancing orders often go into the closing auction of the cash market. In the final minutes before the 4:00 p.m. Eastern close, exchanges publish information on closing-auction imbalances, and index futures can react as participants trade against or alongside them. In currencies, many large transactions use the 4 p.m. London WM/Reuters benchmark fix, which concentrates FX futures volume in the minutes around it.

Order flow near the close

On a month-end day, the footprint of ES in the last 10 to 15 minutes before 3:00 p.m. Chicago can show much more volume than on a normal day, sometimes with large two-sided trade and little net movement, sometimes with a directional burst. Big trade bubbles cluster in that window. For 6E and other FX futures, a similar concentration can appear around 10:00 a.m. Chicago, when the London fix window occurs while clocks are aligned.

Because nothing is scheduled at a precise second, the question is proportion: comparing the final minutes with the same minutes on previous sessions separates genuine month-end activity from normal closing volume.

In Senzoukria

Rebalancing is not an economic release, so it does not appear in the calendar. Time-of-Day Volume, which divides each bar's volume by the average of the same minute over previous sessions, is the direct tool: a closing bar at several times its usual volume on the last trading day of the month stands out.

The Big Trades overlay, with its 100 ms merge window and size filter, shows whether the closing volume came in large prints, and the DOM ladder's session columns show at which prices it traded.