PsyFi Search


Monday, 12 March 2012

Hubble, Bubble, Index Trouble

Consequences, Consequences

In a world in which our behavioral biases are continually encouraging us to stray into temptation, or at least into the grasping hands of the securities industry, we have to look for whatever small mercies we can find. Not the least amongst these is the humble index tracking fund, an asset type specifically created to minimise the opportunity for self-inflicted mischief.

Yet such is the nature of the financial business that even in the world of the behaviorally inert, passive index tracker we can’t avoid the dead hand of unintended consequences and cognitively induced misadventure. In becoming popular index tracking funds have simultaneously created their own behavioral problems: because whenever too much money imbued with too little intelligence chases too few assets the only possible outcome is a bubble. And we all know what happens with bubbles.

Fundamentally Flawed

As the role of index tracking securities has grown so has their impact, to the point where it now appears that the most popular index trackers are no longer passive and dispassionate observers of markets, simply tracking their appointed indexes, but are having a material impact on their performance. By one estimate the effect of this on S&P 500 stocks may be to cause an overvaluation, relative to the rest of the market, of as much as 40%.

What this suggests is that the valuation of a stock may no longer simply depend upon the normal suspects – their valuation fundamentals or crazy, demented investor sentiment – but can be seriously affected by whatever index it happens to belong to. In fact, investing in large capitalisation indexes may be simply helping pump up a index driven bubble. This should be no surprise: this is the world of finance, where no good deed goes unpunished.


The idea that an increased reliance on index tracking should actually start to impact the stocks in the indexes being tracked is just the latest example in a long line of clever financial innovations being pushed beyond their limits by an industry that’s never heard of the concept of "too much of a good thing". Unusually, this time, the main problems aren’t deliberate, but are unintended consequences of the success of the index trackers.

In The Mysterious Growing Value of S&P 500 Membership Randall Morck and Fan Yang point out that the potential growth of S&P 500 index tracking leads, eventually, to a ridiculous outcome:
“This is easy to see in reductio ad absurdum. If the amount of money indexed to the S&P 500 grows without bound, index funds will come to buy and hold virtually all the shares in the firms in the index. Obviously, if still more money is pumped into index funds, investors squatting on the last few shares of each index member firm can demand exorbitant prices.”

Perhaps this can be seen most clearly when we look at the impact of stocks joining the S&P 500. Firstly the stock will tend to jump around 9% in value at or around the time of the announcement. However, as Jeffrey Wurgler relates, in On The Economic Consequences of Indexing:
“The inclusion effect is just the beginning. The return pattern of the newly included S&P 500 member changes magically and quickly. It begins to move more closely with its 499 new neighbors and less closely with the rest of the market. It is as if it has joined a new school of fish”
As this effect takes hold the valuations of the members of the most popular index start to sidle away from “real” valuations. When index constituent stocks are benchmarked against similar stocks outside the S&P 500 researchers find an overvaluation of between 30% and 40%. This may be an overestimate, but even if you scale this back a bit the implications are still startling, and potentially still growing.


Technically you’d expect this indexing premium to lead to arbitrage, where rational investors exploit the overvaluation of the popular indexed stocks. This doesn’t happen, which is probably because smart investors don’t tend to trade against the money – and the amount of money being invested in index funds is huge: over a trillion dollars directly, and more indirectly – and because of the difficulties and dangers in shorting such trends (see: Save Our Short Sellers).

Indirect indexing is even more important than the direct funds attributed to it imply, because active fund managers increasingly tend to be benchmarked against relevant indexes and end up being closet trackers themselves, albeit while charging rather more for the experience. This itself poses a limit to the arbitrage effect, as the very investors who might seek to exploit the overvaluation of the most popular indexes are often handicapped by needing to closely mimic them.

Here’s a rich irony. The ideal of the Efficient Market Hypothesis, in which we track the market, is the one that the index tracker aspires to, helping us to avoid all of the self-inflicted pain that behavioral biases create. Yet the impact of the index tracker, as it becomes popular, in part due to investors realising that they suffer from behavioral bias, is to drive constituent stock valuations away from fundamental valuations.

Bubbling Under

Does this matter? Well, probably. If stocks within the most popular indexes are experiencing an index tracking induced bubble then there is only one ultimate end possible. Something that can’t go on forever, won’t, and eventually the bubble will pop. Too much money chasing too few assets never, ever, ends happily ever after.

Markets just don’t ever offer up free lunches forever and the index tracker is no exception to this. If you happen to be heavily invested in an index that’s overvalued due to too much index tracker money you’re running a far greater risk than you probably appreciate. This isn’t the old argument that investing in market capitalisation weighted indexes means you’re investing in overvalued stocks, it's that the index itself is overvalued, and its constituent stocks are also overvalued - because they’re in the index. These are not equivalent arguments.

The rough outcome of this is that not all index trackers are equal. The more popular the index is – and the S&P 500 is the most popular of all – then the more likely it is to be suffering from this index tracking bias. Looking for less popular indexes, albeit ones that are still representative of the real market, is a safer bet. Of course, even a wider general US index will still be heavily exposed to the S&P 500, which makes up nearly 80% of total US market capitalization (when it should probably be closer to 60%, if you reverse out the index bubble effect).

Foreign indexes or specialist trackers, especially where the market liquidity or the range of stocks is limited and the geography or asset type is wildly popular due to some fad or mania, are also likely to fall foul of this type of problem. The general equation is simple: too much money chasing too few outlets always equals too high a valuation. Fashion never equates to value for money.


This offers up some interesting opportunities for those investors possessed of a mildly adventurous nature. I wouldn’t want to suggest that passive investors should throw caution to the winds and start day trading in frontier market small cap stocks or yak futures, but some careful diversification away from the most heavily impacted indexes might be in order. Including fundamental indexers, particularly those that aren’t market capitalisation weighted or which are value based, or even (ahem) active funds that are explicitly not benchmarked to indexes might be an option and periodic rebalancing between asset classes should help as well, although you’d need to get your asset allocations right; another area fraught with bias.

There’s one final hypothesis we can derive from the index bubble effect. If this theory is correct it would mean that active fund managers have been faced with an almost impossible task over the past twenty years, because they’ve been battling the wall of money cascading into index trackers. In such an environment it would mean that the average quality of active managers has declined because it’s impossible to discern who’s good and who’s not and would also suggest that all of the careful and clever studies designed to show how active management is hopeless are so much wastepaper.

How ironic would that be: genuinely good active fund managers driven to the margins by index trackers, researchers wasting their time proving a general theory of active management failure undermined by an environment biased by flow of funds to index trackers while passive funds themselves are the subject of a behaviorally induced bubble, caused by investors seeking to escape the underperformance of said active managers. It’s reflexivity writ large. Hubble bubble: modern finance is a witches’ brew indeed.

Related articles:


Anonymous said...

Really quickly - I think this sentence "and the amount of money being invested in index funds is huge: over a billion dollars directly, and more indirectly" should be a trillion.

Other than that - an excellent indicator of where the next bubble may be inflating; rather worryingly the breadth of the problem suggests it'll be difficult to avoid the burst though...

cig said...

The idea that there is a finite supply of index shares is factually incorrect: all 500 cos of the SP500 can issue as many new shares as they wish. Basically their cost of funding will be cheaper than non index cos, if the hypothesis is true. They can also arb the issue by changing the balance between equity and debt, which isn't fixed either.

I don't think the idea that individual cos get dominated by index correlation is that material, correlation can increase volatility and make convergence with fundamentals more jerky than it would otherwise be, but it's far fetched to extrapolate that to eternal divergence.

Finally the idea that the top 500 "should be" 40% of the market defies basic common sense. Company size distribution tends to be power law, indexing or not, so you're hardly ever going with a huge small cap sector unless the gov actively controls company size.

timarr said...

Trillions is right, not billions, now fixed.

cig's point about the companies themselves arbitraging this effect is interesting. I can think of a few reasons they might not: for instance, the amount of cash currently held by corporations is "irrational" anyway when debt is so cheap, which is probably a psychological hangover from the banking crisis. But it's an intriguing thought.

The numbers aren't quite right there, though. The papers state that the S&P 500 accounts for 80% of total market cap, compared to a "fair value" 60% - so roughly a (maximum) 40% over-valuation. Plenty to argue with there, though.

Stephen Brooks said...

1. What if too many invest in an "all-share" index that tracks nearly the whole equity market? Does that mean there's just a bubble in equities in general?

2. If the shares in the index are overvalued, the shares outside the S&P 500 should have better dividend yield. Not much good for traders but according to this theory, long-term investors who accumulate dividends from the American small-caps should do disproportionately well.

cig said...

I agree current cash levels may be a bit funny but on a longer term view I think financial gravity will win. Another supply side arb source is new entrants making extra effort to get in popular indices.

Top 500 being 80% of market cap is totally normal (sorry got the numbers upside down in my last comment) and I'd be surprised if it wasn't true before indexing became popular and in many other markets. You get power law distributions anywhere there's economies of scale and no big external constraint on size (book sales, website popularity, blog readership, etc are probably similar).

Anonymous said...

I really appreciate articles written by Psy-Fi. One of the most useful and innovative information sources I have ever come across.
Sincere Gratitude and Best Wishes !