DAX additions lost ground before rebalancing. Deletions gained.
Across 38 DAX membership changes from 2015 to 2023, additions underperformed and deletions outperformed from the next open after announcement to the rebalancing close. Those moves followed months of relative strength in additions and weakness in deletions.
When a stock joins the DAX, funds tracking the index need to add it to their portfolios. When a stock leaves, they need to remove it. That gives a straightforward expectation: buying should support additions and selling should weigh on deletions before funds rebalance.
The sample showed the opposite. Most additions lost ground against the DAX, while most deletions gained. The stocks also arrived at the announcement with very different price histories: additions had already outperformed for months, and deletions had underperformed.
Prices moved against the expected effect
We reconstructed 24 additions and 14 deletions from archived Deutsche Börse composition records. For each stock, we summed its daily return minus the DAX return to measure performance relative to the index.
The measurement starts at the next Xetra open after the announcement and ends at the rebalancing close. This is the “chaining close” in the DAX rules: the closing prices used to rebalance before the new composition takes effect. For scheduled reviews, it falls on the third Friday of March, June, September, or December.
| Cohort | To rebalancing close | Next 5 sessions | Next 20 sessions |
|---|---|---|---|
| Additions (24) | −3.98% | +0.74% | −3.00% |
| Deletions (14) | +3.04% | −4.02% | −3.12% |
All returns are relative to the DAX. A positive return for a deletion means the stock outperformed the index.
The result extended across both groups. Only 5 of 24 additions outperformed before rebalancing, compared with 11 of 14 deletions. After the rebalancing close, deletions gave back their gains. Additions recovered a little over five sessions but underperformed again over twenty.
Additions had already risen, and deletions had already fallen
Over the 60 sessions before the announcement, additions outperformed the DAX by an average of 12.64%, while deletions underperformed by 19.41%. That history was available for 21 additions and 13 deletions. The remaining stocks had listed too recently.
Among additions with a full history, 19 of 21 had outperformed before the announcement. Their losses on the way to rebalancing followed a broad period of relative strength.
That sequence matters because index membership follows movements in market capitalization. Stocks enter the announcement with a price history tied to their selection. An unwinding of earlier gains and losses, or positions built ahead of the announcement, could help explain the subsequent moves. The return data alone can't distinguish those explanations from other causes or isolate the effect of tracker demand.
The expansion and Wirecard change the size of the result
The 38 changes came from 16 review events, with ten additions sharing the September 2021 announcement that expanded the DAX. Those ten underperformed by 1.00% before rebalancing, compared with 6.11% for the other fourteen additions. Both groups moved in the same direction, but the expansion softened the overall decline. The 24 additions aren't 24 independent observations.
Wirecard has a different effect on the deletion results. Its August 2020 removal followed insolvency. Excluding it, deletions still outperformed before rebalancing, by 3.43%, but their subsequent losses were much smaller: 1.18% over five sessions and 1.85% over twenty. Their average pre-announcement decline also shrank from 19.41% to 7.02%, using the twelve remaining stocks with a full history.
Neither exclusion changes the direction of the pre-rebalancing result. Wirecard does account for much of the size of the deletions' earlier decline and their losses after rebalancing.
The announcement reaction went the expected way
The announcements came after the market closed, so the next-open starting point leaves out the overnight reaction. During that gap, additions outperformed by 0.62%, with 20 of 24 positive, while deletions underperformed by 0.25%. The initial reaction followed the expected direction, but it was too small to offset the moves that followed.
The opposite pattern lasted through the rebalancing session itself: additions underperformed by 0.99% and deletions outperformed by 1.44%. On the next session, when the new composition took effect, additions gained 1.04% relative to the DAX and deletions lost 0.97%. Those two sessions nearly canceled for additions and left deletions ahead by 0.47%, so the broader result wasn't confined to that final pair of sessions.
The finding is clear within this sample: buying additions and selling deletions at the next open after announcement would have faced adverse relative price moves before rebalancing. Whether the opposite trade would be profitable remains untested. The study doesn't establish statistical significance or account for trading costs, market impact, borrow availability, or capacity.
The companion DAX rebalance flow: methods note contains the full tables, sample construction, sensitivity checks, and DAX rulebook reference.