When a company is added to a major index, its shares typically rise before the change takes effect and drift back afterwards. No analyst upgraded it. Nothing about the business changed. What changed is that several trillion dollars of index-tracking capital became obliged to own it.
Why tracking funds have no discretion
An index fund’s product is not returns. It is replication — delivering the index’s return minus a stated fee, with minimal deviation. That deviation, tracking error, is the metric the fund is judged on.
This has a consequence that is easy to state and easy to underestimate: the fund must own the index constituents in index weights, and when the index changes, it must change too. A manager who declines to buy an addition because the valuation looks unattractive is not exercising good judgment. They are failing to deliver the product.
More specifically, the fund must trade at the price the index uses to calculate the change — normally the closing price on the effective date. Buying earlier at a better price introduces tracking error in the other direction. The fund is committed to being a price-insensitive buyer at a known moment.
The predictable sequence
Index changes are announced in advance — usually days to a couple of weeks — and the interval between announcement and effect is where the price action happens.
- Announcement. The share price of an addition jumps immediately, as traders position ahead of the mandated buying.
- The interval. Continued drift upward as more participants position. Nothing forces this; it is anticipation of the flow.
- The effective date close. Enormous volume in a single auction. Index funds buy; those who positioned early sell to them.
- Afterwards. Partial reversal over the following weeks as the temporary demand disappears and the price returns toward where fundamentals put it.
Deletions run the same sequence inverted, and the reversal after deletion is often larger — forced selling into a market that knows it is forced tends to overshoot.
Why the effect has weakened
The index inclusion premium was substantially larger in the 1990s and 2000s than it is now. Three things eroded it.
It became famous. A well-documented, calendar-driven, direction-known price move attracts capital until the excess return no longer compensates for the risk. That is what an efficient market does to a known pattern.
Index providers changed their methods. Longer notice periods, phased implementation over several days, and buffer zones that prevent stocks oscillating in and out all reduce the concentration of the flow. Some providers implement large changes in tranches specifically to limit market impact.
Additions became less surprising. Index methodologies are published. For most benchmarks, the plausible candidates are known well in advance, and much of the repricing happens before any announcement.
Free float is the underappreciated mechanism
Most major indices weight by free-float market capitalisation — the shares actually available to trade, excluding blocks held by founders, governments or strategic holders.
This means the flow generated by an index change is not proportional to a company’s size. It is proportional to the shares that must be bought relative to those genuinely available. A company with a small float has a much larger effect for a given market capitalisation, because the same index demand is chasing fewer shares.
It also means that a change in float classification alone — a lock-up expiring, a strategic holder selling down, a provider reassessing what counts as free float — generates index flow with no change in index membership at all. These adjustments are less publicised and can produce larger moves than headline additions.
What this means for an ordinary investor
Three practical points.
- Do not buy an addition on the announcement expecting the premium. The move happens in minutes and the reversal is real. This is a trade for participants with low costs and fast execution, and even for them the margin is thin.
- Do not read index-driven moves as information. A stock rising eight percent on inclusion has not become a better business. The market is telling you about fund mandates.
- The cost to index investors is small but real. Funds buy additions after the price has risen and sell deletions after it has fallen. This is a genuine drag, and it is one of several reasons an index fund’s return is not exactly the index return.
That last point deserves proportion. The drag is small — well inside the fee differential between index and active management for most benchmarks — and it does not change the underlying arithmetic that makes low-cost indexing hard to beat. It is a qualification to that case, not a refutation of it, and we set out the case itself in index funds versus active management.
