Our 2018 article “Buy High and Sell Low with Index Funds!” suggests several ways index funds could add a modest excess return in most years. The industry, however, has trained clients to view any tracking error relative to the published index (including positive tracking error!) as evidence of inferior, sloppy risk management. Index fund managers who seek to take advantage of these inefficiencies lose business because their positive tracking error is taken to be evidence of sloppiness. The saving grace for indexers is that the stocks added at too high a price and sold at too low a price represent only about 2–4% of annual turnover. Even with an avoidable 1,000 bp performance drag from the buy-high, sell-low behavior, the annual cost to clients is roughly 20–40 bps, which the clients typically do not notice because the published index itself suffers from the same drag!
As of year-end 2019, approximately $4.6 trillion of investment capital was directly tracking the S&P 500 and another $6.6 trillion was benchmarked to the index. When the composition of the index changes, the resulting trades drive up the price of the addition and pull down the price of the deletion. Only 35% of additions to the index will have a higher share price on the day the Index Committee announces the stock will be added than the effective closing price on the day they are actually added. The prices of the remaining 65% of added stocks rally between announcement date and effective date of the addition. Reciprocally, at the open on the announcement date, 60% of discretionary deletions are above the effective closing price on the day they are removed from the index. Thus, the share price of 60% of discretionary deletions falls between announcement and effective date. Over the last two decades, on average the deletions underperformed the additions by 6.2%, from the market open on the announcement date to the market close on the rebalance date.6
Typically, an additional 1.0% of underperformance follows on the first day after the addition is made as some index trackers play catch-up on trades they did not complete on the effective date. This additional 1.0% is primarily due to buying pressure from the liquidity providers who traded the stocks ahead of the rebalance as well as from trading activity on the rebalance date. Most index-tracking strategies trade en masse at the exact closing price the day before the addition officially occurs.
Following the rebalance event, prices of the two stocks that were added and deleted will start to move in the opposite direction. Over the 12 months following the rebalance, our research shows only 43% of additions finished ahead of their effective closing price relative to the market. Because stock prices can rise much more than they fall, this translates to only an average 1% loss relative to the market in the 12 months following the rebalance. Reciprocally, fully half of discretionary deletions finish the subsequent 12-month period ahead of their effective closing price relative to the market. Because the gains tend to be significantly larger than the losses, the discretionary deletions beat the market by nearly 20% over the next 12 months, on average. As a result, in the first six months following the rebalance, the additions tend to lag the deletions by 14%, and by month 12, the additions lag by 20%.
The index rebalance is a great opportunity… to do the opposite of what the index does: buy the deletion and sell the addition. Providing index investors liquidity and benefitting from the mean reversion of the price changes has historically proven to be an excellent investment idea.7