Has the index effect evaporated?

Can those lamenting/celebrating/gawping at the not-so-slow-motion car crash of Elon Musk’s Twitter ownership really blame the benchmarking wizards of S&P Dow Jones Indices for the whole debacle?

That’s what a lot of people reckon. When Tesla finally managed to notch up four consecutive quarters of profit in the summer of 2020 and became eligible for inclusion into the S&P 500, it helped spark a wild ride that made it one of the world’s most valuable companies. That transformed Musk into the world’s wealthiest edgelord and allowed him to buy Twitter.

It’s easy to forget that at the start of 2020 Tesla was valued at just $77bn (which even at the time felt punchy). By the end of the year Tesla’s market capitalisation had exploded to $669bn. A year ago it had reached a peak of $1.2tn, and even after the tech wreck of 2022 Tesla is still worth $565.4bn.

This is why a new NBER working paper from Robin Greenwood and Marco Sammon from Harvard Business School is so interesting (it also thanks Lloyd Blankfein, among others, for “helpful comments”). Here is its main findings:

The abnormal return associated with a stock being added to the S&P 500 has fallen from an average of 3.4% in the 1980s and 7.6% in the 1990s to 0.8% over the past decade. This has occurred despite a significant increase in the percentage of stock market assets linked to the index. A similar pattern has occurred for index deletions, with large negative abnormal returns on average during the 1980s and 1990s, but only -0.6% between 2010 and 2020.

In other words, the impact of index inclusions and deletions is pretty much statistically indistinguishable from zero. This runs completely counter to popular perception. It obviously makes intuitive sense that several trillion dollars worth of price-insensitive funds suddenly having to buy an included stock would lift it higher.

It should also be pointed out that although all perceived market evils are laid at the feet of index funds, indices have a powerful sway over traditional active funds as well. In fact, we suspect that stockpickers are on average nowadays far more “index aware” than they were back in the day. Even if there were no passive funds, benchmark changes would therefore probably have a big impact (and let’s not forget the impact of index derivatives).

Just to take the Tesla example, S&P Dow Jones itself estimated that index funds would have to dump about $51bn worth of other stocks to make way for Tesla’s inclusion. We’ve even had the first case of alleged index inclusion insider trading.

But Greenwood and Sammon corroborate similar findings by Benjamin Bennett, René Stulz and Zexi Wang in 2020, who found that the long-run impact of inclusion into the S&P 500 had actually become negative.

So what’s up? Greenwood and Sammon explore five possible explanations:

1) The fading effect is caused by different types of companies being included and excluded in recent years compared to the past.

2) The stock market is more liquid nowadays and trading costs are lower, so the impact of index changes becomes more muted.

3) Changes in net demand are much lower than they might seem because companies typically migrate from one index to another. For example, the S&P MidCap index has become a far more popular index.

4) Benchmark changes have become more predictable, and attracted arbitrageurs who front-run buying and selling by index funds.

5) The stock market has become more efficient overall, and liquidity has migrated towards dates where index changes happen, and especially at the end of the day when index funds do most of their trading.

Related to that, here is a gif showing how the “liquidity smile” has turned into a lopsided “liquidity smirk” over the past decade.

Greenwood and Sammon favour the last two explanations, with most of the emphasis on the final one. Here is their tl;dr:

Overall, the findings suggest an account along the following lines. In the 1980s, index changes were unanticipated, index funds were small, and there was mispricing in the market. As index funds grew larger, the mispricing deepened and turned into an opportunity. As a result, the market adjusted to take advantage of this opportunity, in part by better anticipating inclusions, and in part by creating arrangements where other institutions stood ready to sell to indexers upon inclusions. This worked to eliminate the anomaly on average, in spite of demand shocks that continued to grow in magnitude over the 2000s and 2010s. In this sense, the decline of the index effect is much like the evidence for other anomalies, that they decline once they are well recognized by the market.

This makes sense, even if the researchers might be underestimating the arbitrage aspect a little. We wonder if the index inclusion effect would look more meaningful if one tweaked the time parameters. Index arbitrage has become a more popular hedge fund strategy in recent years, and that implies the inclusion effect is simply getting spread out over a longer time period.

Some final thoughts on the Tesla saga though, as it would seem to undermine the whole “disappearing index effect” argument. As with everything Musk-related, this is probably just a very idiosyncratic situation, with limited read-through elsewhere.

There are so many hardcore Elon stans and Tesla fans that have repeatedly seized on any excuse to pump shares in the company, from Mars mining to Tesla becoming an insurance giant. Even aside from how owning Tesla has made many rich, Tesla stock ownership has become an extension of their identity. And probable index inclusion is as good a reason as any to buy more.

A large part of the massive ramp-up in Tesla’s market cap was therefore probably the collision of a retail investor buying frenzy with a limited free-float, given a chunk of the Tesla investor base didn’t want to sell.

But here is what Tesla has done compared to Apartment Investment and Management — the company that was ejected from the S&P 500 to make way for Elon’s carmaker — since December 21, 2020, when the change went into effect: