Allocating money to an index-based investment fund is popular because it’s cheap. And the easiest and most transparent way of doing so is to invest in a capitalisation-based index fund, where the weighting of each stock or bond reflects the market footprint (capitalisation) of each company.
But a cap-weighted portfolio can become heavily exposed to overvalued sectors during market bubbles—think internet and telecoms stocks in 1999/2000, banks in 2006/07 and Apple shares in 2011/2012.
Index providers struggle with problems of excessive concentration, and often in an arbitrary and unsatisfactory way.
Last April Nasdaq cut Apple’s weighting in its 100-stock cap-weighted index from over 20 percent to around 12 percent, to assuage concerns that the computer maker’s index weighting had become outsized. The move provided only temporary relief. By summer 2012 Apple’s share of the Nasdaq-100 was back up at 20 percent, reflecting a further sharp rise in the company’s shares.
When Volkswagen’s shares hit a weighting of 27 percent in Germany’s DAX index in 2008 as the result of a massive short squeeze, the German Stock Exchange cut its weighting without warning to 10 percent. STOXX advertises a “rules-based” approach to compiling indices, but here it made clear that ignoring the rulebook to deal with perceived market distortions was more important.
If you buy a cap-weighted index fund or ETF, you are exposed to such skews, which are unfortunately a lot more difficult to exploit at the time than to recognise in hindsight.
Firms like Vanguard, who stick rigidly to capitalisation-weighting in their tracker funds, make the valid point that market timing is difficult. They argue that any departure from the market portfolio is a bet against the consensus and should be recognised as such. Cap-weighted index approaches also generate the lowest turnover and ETFs based on such indices are usually by far the most liquid tracker funds to trade.
But if you accept that the market can be inefficient, one of the prime reasons for distortions in valuations is that a group of people with very large sums of money to throw around may have other considerations than investors’ best interests at heart.
For those involved in managing other people’s money, the potential rewards can be astronomical. According to Chris Newlands of the FT, the pay of the heads of several publicly quoted UK fund management businesses now reaches £5-10 million a year. An unnamed executive at Pimco Europe, the European arm of the world’s largest fixed-income manager, took home £30 million last year.
And while regulators are cracking down on bankers’ pay, fund managers remain relatively untouched. Cynical observers might draw a connection between the current focus on bankers’ bonuses and the recent fashion for investment banks to reinvent themselves as asset and wealth managers.
With such fortunes on offer, it’s unsurprising that fund managers may put their own career risk ahead of the interests of their investors. And as the main risk to fund managers’ careers comes from taking big bets, getting them wrong, underperforming the competition and getting fired, most of them copy each other, following the same trends and inflating bubbles.
Fund managers suffer from “benchmark tyranny”, putting it another way. As their performance is often measured quarterly or even monthly against a capitalisation-weighted index, many track those benchmarks much more closely than they will admit in public.
So a big problem with cap-weighting has nothing to do with the abstruse arguments about efficient markets that we hear regularly at conferences on ETFs and indexing. Instead, it concerns the way active managers are rewarded.
As long as fund managers are incentivised by option-like pay awards for chasing and exceeding benchmark performance, rather than taking a dispassionate, long-term view of market valuations, bubbles in market indices are likely to be much more prevalent than otherwise.
And while such perverse reward schemes exist, as an index investor you’re better off following another weighting scheme than the traditional approach. Equal or fundamental weighting, low- or minimum-volatility, almost anything will serve you better than cap-weighting.