I admit it. I am a glutton for knowledge. My daily reading includes many different websites and financial blogs. I also try to investigate new companies multiple times per week, often by running stock screens at Yahoo Finance or Morningstar. I occasionally go to the local university library and just pick up a magazine or journal and start reading.
I read a lot, yet I find very little of what I read to be truly illuminating or even interesting. Every so often I have an ‘aha!’ moment, and today I had one as I read John Mauldin’s Outside the Box newsletter, written this week by Louis-Vincent Gave and Anatole Kaletsky of GaveKal research. The text of the newsletter can be found online at Mauldin’s website. This letter is an excerpt from their new (if un-originally entitled) book, Our Brave New World.
I have yet to buy the book, but I certainly will do so because of my enthusiasm for this letter. I will not summarize the newsletter, so read it before you continue.
It may seem odd, but I disagree with much of the authors’ reasoning. I do agree with their main point, however, that indexing can only work when it is the minority strategy and when active investing is the majority strategy. To explain this, I will give a simplified analogy.
Imagine that we live in the country of Tinagra. There are only two public companies: Darmok Inc. and Jilanco. Both companies combined produce most of the goods and services. They each have revenues of $1 trillion and earnings of $100 billion. Thus, any difference in their market capitalization is due to investors’ prognoses of the companies’ respective growth prospects.
We will assume that the vast majority of investors actively deploy their funds and make their own investment decisions. Most of these investors believe that Darmok has better growth prospects, so they bid the stock up to a point where the company’s market cap is $1.5 trillion (for a P/E of 15), while Jilanco has a market cap of $1 trillion (P/E of 10). Over time the market caps of the two companies will fluctuate as the investors continually evaluate their future prospects. For the moment, let us assume that those fluctuations are small and thus irrelevant.
Suddenly, some investment advisors start to herald the benefits of indexing. They say (rightly) that any fool could have invested his money in an index fund over the last 10 years and received a return equal to that of the market while paying no fees and doing no work! Proponents of indexing form a company called Rearguard and their index mutual fund becomes very popular.
How the index fund works is simple: it buys shares of stocks according to their market capitalization: for every $2.50 invested, $1.50 is invested in Darmok and $1.00 is invested in Jilanco. Due to the simplicity and ease of this index investing, everyone switches too it and there are no more active investors. Now this would be fine if the past perfectly predicted the future. This is not the case. There is a scandal at Darmok: their newest blockbuster drug is found to have caused blindness in millions of people. The prospects for future lawsuits and lost sales means that Darmok will be a worse investment than before.
Oddly enough, nothing happens to their stock price. Since everyone is investing in a Rearguard index fund, everyone continues to buy more of Darmok than of Jilanco. The price of Darmok cannot fall, nor can the price of Jilanco rise. To any rational investor, this seems irrational, and with good reason. Circumstances have changed and the relative pricing of the two companies (1.5:1) is no longer warranted. Even if investors sell out of the market completely, the relative value of Jilanco will be greater than Darmok because their stocks will be sold off by the index funds in proportionate amounts.
Any intelligent investor would at this point sell out of his index fund, short Darmok and buy Jilanco. While the above is a bit of reductio ad absurdum, the point remains valid even if index investing is not total and even with 7,000 stocks.
As Gavekal points out, many mutual funds now behave like closet index funds because the managers are afraid to deviate too far from their benchmark. Gavekal believes that we have reached the point where this indexation has become a large enough factor to lead to serious mis-pricings in the stock market. The math to back them up is probably beyond me and I lack the data with which to do the calculations, but judging from the amount of money invested in the largest mutual and pension funds I would have to agree with them.
I mentioned before that I do not agree with all of what Gavekal say in their article. One point with which I most strongly disagree is their argument that indexing will lead to lower overall returns for the stock market and slower economic growth. I am not aware of their reasoning, for I have not yet read their book, but my first reaction when reading that is to argue that the stock market has very little to do with the conversion of savings into investment anymore. For a full explanation of why that is the case, consult John Burr Williams’ classic book, The Theory of Investment Value.
Williams’ basic argument is that since buying old stock is simply a transfer of ownership and is rarely used to finance new capital goods (except in the case of IPOs, secondary offerings, and rights offerings), it does not affect the interest rate nor the allocation of capital. The money paid for a share of IBM pays the seller of the share, who may then invest it in a share of a new venture; buying IBM does not allocate money to IBM. Thus, the stock market does not directly affect the economy very much.
Minor caveats aside, while indexing may not cause capital mis-allocation, it does cause investment mis-pricing, leading to exploitable changes in the prices of traded securities. Simply put, indexing helps the stock picker by reducing the price differences between companies that have different values. It remains the stock picker’s job to decide which companies are the better investments.
Are there any a priori identifiable mis-pricings? I believe so. In an efficient market, firms would not be priced differently solely because of their market cap. Index funds cannot invest enough in the smallest mini- and micro-cap stocks, because they must invest according to market cap. Therefore, even though small cap outperformance of the market is well-documented, and such well-documented findings tend to be discounted by the market (and thus disappear), small-cap outperformance of the market is likely to continue. The only thing that could prevent this would be for individual investors to increase their overweighting of small-cap stocks (note that in the last newsletter I mentioned that individual investors already overweight small-caps in their portfolios).
The second clear mis-pricing I can identify a priori is that indexing may lead to stickiness in the prices of some large-cap stocks (or any stock that has a large proportion of shares held by index funds). I must first note that this is a bit speculative. The basic premise is that index funds (and most non-index mutual funds) will own the largest stocks (and thus the majority of these companies’ shareholders will not be prone to selling except when the market cap changes significantly). When the company begins to perform better than expected, the price will not increase nearly as much as it should, because the largest holders, index funds, will not need to buy. The stock price will thus rise more slowly than it should and more slowly than if there were no index funs.
This hypothesis leads to the predictions that stocks without index fund owners will be more volatile (but in a good way—in response to news) than companies with index fund owners. Any problems with the above? Unfortunately, yes. The increasing size and clout of hedge funds should reduce the importance of index funds (and similar managed funds). I do not believe that we can count on indexers making it easy for those of us who choose to invest more actively.
The effect of index funds is not negligible. Any trading strategy that is widely used will affect the market. Unfortunately, we must try to predict how different strategies interact to affect the market. The question of whether the market is becoming more efficient (due to the prevalence of hedge funds) or less efficient (due to increased indexing) is too complex for me to answer at this time. The problem may be too complex to solve. I will certainly return to this question in the future. Even coming up with a poor approximation of a solution to this problem could be quite profitable.
I do believe that indexing could become a much larger part of the market without making the market less efficient. A brief discussion of the issue can be found in this academic paper [pdf].
Disclosure: I first penned the above piece some time ago. It does not necessarily represent my present views.
I think it’s a valid weakness for traditional index funds, but I also believe that the market is aware of it and has compensated already. For example, Powershares uses a different allocation formula with their ETFs, for some of the same reasons that you bring up above. I think that the biggest impact that indexing might have, would be to increase volatility on the stocks. I don’t know what the data looks like, but you would have to imagine that reducing the float would have some kind of an impact on price stability.