Rethinking Index Investing

There are two general different ways to play the equities game. You can try to pick individual stocks yourself in the hopes of someday replicating the success of Warren Buffett, or you can leave such aspirations to the more ambitious folks in our society and simply try to earn an average return.

As I have argued before, although I do believe it is possible to beat the market consistently as Warren Buffett has done, this does not mean that just anyone can hope to do so.

It takes a unique person. Of course, where does that leave the rest of us?
Mr. Buffett has said that the ordinary investor is better off investing in a passively managed index fund than an actively managed mutual fund. While the proponents and salesmen of index funds or fund families that try to capture the benefits of index funds love to use this quotation for their own self-serving benefit, it is hard to disagree with Mr. Buffett. He calls such ordinary investors “know-nothing investors,” although he doesn’t mean it as an insult. He just means that if you don't know what you are doing, it is safest to simply accept your mediocrity and guarantee yourself average returns. This begs the question, though, what is “average? “

This argument is also consistent with the efficient market theorist’s investment strategy. Since we can't expect to beat the market anyway, why not simply arrange to earn the relatively high returns equities offer us while minimizing fees through passive management? Note the underlying assumption behind all of this, that "the market" is best proxied by a value-weighted index, and this is where we get in to trouble.

What is “the market?” Professor Richard Roll’s famous critique of the Capital Asset Pricing Model has stumped theorists for nearly thirty years now. He contended that we will never be able to truly quantify risk because the true “market” portfolio is unobservable – it includes all assets that contribute to wealth, everything from stocks to baseball cards to human capital.

So let me ask two simple questions. Why are market-value weighted indexes such as the Nasdaq Composite or S&P 500 the most generally accepted benchmarks by which to compare investment results? Is there any particular reason for such value-weighting other than convention?

Some fund companies have made a big business out of creating alternative indexes that try to outperform or by taking the components of existing indexes and weighting them equally, and yet others weight them according to the Fama-French Three Factor Model. Yet we continue to measure performance relative to value-weighted benchmarks. Perhaps, since as individual investors we likely never have sufficient wealth to affect the value of one stock purely through our individual trades, we should have no limit to how to we (passively) allocate our wealth. Hence, by benchmarking ourselves to the S&P 500 we are actually overstating our relative performance.

Exchange traded funds, ETFs, are very much in vogue at the moment. The original ETFs were broad-market index funds such as the SPDRs, which mimic the performance of the S&P 500, and appealed to two types of investors. The first type is the delusional folks that think they can actually predict the direction of the market. We can thank those geniuses, I suppose, for providing liquidity to the second type of investor who uses the SPDRs for true investment purposes.

But by purchasing shares of SPDRs or any other broad-market index fund, investors are inadvertently investing most heavily in those companies that are most likely to be overvalued. This is the crux of my argument. Since the S&P 500 (SPY) is a value-weighted index, the value of one share of a SPDR will be dominated by the largest companies and those that have most recently appreciated. So while you may rest assured that your share does in fact represent the collective performance of five hundred different companies, why should you prefer that, among those five hundred companies, you are weighted towards the ones that are quite possibly the most overvalued?

For illustration, consider the three largest components of the S&P 500; ExxonMobil (XOM), GE (GE) and Microsoft (MSFT). This trio is clearly not surprising - these companies are the three biggest in the economy. But what makes me think that these three companies deserve the largest portion of my money? Do I really want $5.88 of the $146.00 I pay for my SPDR to be allocated to ExxonMobil? Moreover, the mere fact ExxonMobil is so large (as measured by market value) may be because investors are paying too much for it.

These are philosophical questions. The index fund advocate will argue that investing in the largest companies is the point. That is, because ExxonMobil is so dominant in the economy, so too should it be dominate in my index fund. But to me this seems unnecessarily limiting, since both larger companies and those that have recently appreciated are likely to underperform their counterparts - provided we believe the evidence provided by Fama and French (1988, 1992) of long-term mean reversion in stock prices and the outperformance of small firms. (I do.)

As I have a habit of doing, I have raised some questions without suggesting any real answers. I will leave that up to anyone wishing to comment.

Comments

[THofler]

I have a modest portion of my portfolio in the IVV S&P 500 ETF. About 1 year ago I had considered the Rydex Equal Weight (ticker RSP) ETF, which has approx. equal dollar weights for the 500 components of the S&P.

While I liked the general idea of the Rydex fund, it seemed to me at the time that the market trend was towards large and mega cap stocks. So that while the fund performed very well after 2002, it would perform less well in 2007 & beyond.

Any thoughts?

[DougM]

Well the problem is, if you try some other weighting than mcap, you'll be forced to reconfigure your positions constantly as stocks rise and fall. This creates turnover, costs the index transaction fees and throws off tax effects that may be undesirable. It's not clear that stocks with the highest mcaps relative to peers are in fact overvalued, since mcap says nothing about the earnings, dividend yield, or growth rate. As for small stocks outperforming, this is a matter of index selection; simply select the right index for the asset class as your benchmark, e.g. the Russell 2000. Another problem is that if all investors heed the results of the research and overweight "small cap" stocks, by definition their prices would rise until they became "large" and you ended up with a market weighting. Obviously not all investors will, but what share of the overall market is now allocated based on F&F's research (counting institutional investors such as pension funds, higher-ed trust funds, etc.)? In the limit, it becomes the market weighting.

F&F's work showing that "value" stocks tend to outperform the average is interesting but the issue of what constitutes "value" is IMO open to question. F&F used P/B, which may be flawed, outdated, or both. Look at how stocks from the financial sector dominate value indexes today (e.g. IWN, IVE, EFV). Book value is a dubious concept at best. At times, I would argue that growth stocks actually represent better "value", in that the price of 5 years of above-market earnings growth is reasonable versus the market averages. In fact, I believe this to be the case now, if you look at the above-average growth prospects for many technology companies trading at P/E multiples not much higher than the index averages. This is the end result of 6 years of "value" investing being in favor. The market is well aware of F&F's research and it's entirely possible the value premium is being reduced until it exactly matches the risk premium associated with e.g. beaten-down stocks (like said financials).

Another problem for F&F - world markets are becoming increasing correlated, such that the differences between US and international equities are blurring. It may be time for a fresh look at what constitute distinct asset classes. For example, dividing the market by sector might result in less correlation among assets now. Rebalancing across sectors could work just as well in MPT, and if proven, would lend some credence to the Wall Street notion of "sector rotation".

[Pathboy]

What is the point? Just go to sharebuilder.com and buy Brkb, LUK, BAM, GE, PG, MO, DEO, JNJ, WFC, KO, VZ or ATT and CSCO each month for the next twenty years and forget about it. You can do this for about $20 dollars a month. Your total mangement fee will be $4,800 for the twenty year period. What will your management fee be when you get to $1,000,000 in the SPY? If it is 20 basis points, $2000 a year. Avoid the trap of ETF/Mutual Funds/Wealth Managers etc.