Last Updated: 21 May 2021
Nasdaq’s contortions over stock weightings remind us to choose our indices carefully.
As my colleague Dave Nadig pointed out in his blog a couple of days ago, the compilers of the Nasdaq 100 index have got themselves in a muddle over stock concentration limits.
When the owners of the US stock exchange’s popular benchmark wanted to license it for use by exchange-traded fund providers back in the late 1990s — and, as Dave argues, which provider wouldn’t want to make money from his index in this way? — they faced a problem.
In order to keep an index-tracking ETF compliant with the US Internal Revenue Service’s diversification rules, which set a 25 percent upper limit for a single fund constituent, Nasdaq had to rejig its index rules, which until then had set stock weights according to company capitalisation. The reason? One stock — Microsoft — exceeded 25 percent of the benchmark back in 1998, and until they got its weighting down there was no way of launching an ETF on the Nasdaq 100 that wouldn’t be tax-disadvantaged.
The exchange/index compiler got around the problem by imposing “adjustment factors”. These cut Microsoft’s weighting to below 20 percent, reduced the weighting of other large-caps (defined as those representing more than 1 percent of the index at the time) by the same proportion, and also boosted the weighting of the smaller companies in the benchmark.
Now, thirteen years later, Nasdaq’s index committee faces a new problem. As a result of the stock’s underperformance, Microsoft’s index weighting has happily shrunk over the years, so the company no longer threatens to break the 25 percent limit. But a new tech stock darling, Apple, has reached over 20 percent of the Nasdaq 100 and is heading towards the weighting ceiling.
Furthermore, the index weightings are now far from being an accurate reflection of market capitalisation. As a result of the 1998 imposition of adjustment factors, Apple’s index weight was boosted from 0.4 percent to around 0.9 percent. At the end of 1998, Microsoft’s weight had been cut from an unadjusted 22.5% to an adjusted 14.5%.
But by mid-2010, Apple and Microsoft, while having similar capitalisations overall, represented 20 percent and 4.3 percent of the index, respectively, purely as a result of the adjustment factors that they’d inherited from the late 1990s. In fact Nasdaq’s description of its index as “modified capitalisation-weighted” didn’t make sense at all.
Recognising the new imbalance in the index make-up, Nasdaq announced on Tuesday this week that it will conduct an extraordinary rebalancing in early May to bring stock weights back into line with actual market capitalisations. Apple’s weight will fall from 21 percent to 12 percent, correcting the largest weighting skew.
If all this seems arbitrary and confusing, that’s because it is. The way this particular index has been managed reminds you of a boat’s helmsman who can’t steer. A steady hand on the tiller keeps the boat moving in the same direction and at full speed; wild oversteer takes you off course and loses you ground in a race.
Nasdaq is not the only culprit in offering a benchmark that seems poorly designed for use in stock selection. If you’re new to indexing, you may be surprised to find out that the membership criteria for companies entering the world-famous S&P 500 and Dow Jones Industrial Average indices are also highly subjective.
The Dow’s components are chosen by an “averages committee” comprised of the managing editor of The Wall Street Journal, the head of Dow Jones Indexes research and the head of CME Group research.
Selection for the S&P 500 is also at the discretion of an index committee, the goal of which is “to ensure that the S&P 500 remains a leading indicator of US equities, reﬂecting the risk and return characteristics of the broader large cap universe on an ongoing basis”.
According to one well-founded analysis of the S&P 500 index committee’s stock picking record, the committee members are subject to the same style biases and drift as the average active manager. They boosted the index’s weighting in tech stocks during the bubble of the late nineties, only to remove several of the same names shortly thereafter; and they relaxed a longstanding prohibition on including holding companies in 2001, allowing lots of real estate investment trusts to be added to the index during the greatest real estate bubble in US history.
This hasn’t stopped the index from being used as the underlying benchmark for the largest ETF in the world, the SPDR S&P 500 ETF (which, at the latest count, had US$94 billion under management).
But amid all the euphoria over passive investing and the exchange-traded fund market’s growth rates, when you’re selecting a tracker product it’s worth casting a very sceptical eye over the index being used.
Index providers will license any methodology for money, and exchange-traded product issuers will sell pretty much any concept that can gain assets. But investors should stick to indices that have objective and clearly understandable rules, and above all rules that make investment sense.