With a number of new approaches to weighting the individual stock holdings in index mutual funds, I though it worthwhile to discuss the topic.
The term "weighting" in a stock index applies to the percentage of the portfolio that is committed to each individual stock.
Stock index fund weightings are determined by the type of index the fund is following, which has rules for the weightings. In addition, an index, even if newly created and unusual, has certain characteristics not necessary in a non-indexed portfolio. An index fund must include all the stocks in the underlying index (there are some exceptions for very small-cap indexes), and the stock components can change only gradually.
There are three basic approaches to stock index weightings (although there are some arbitrary approaches-the Dow Jones averages, which are based on share price alone, being the most noteworthy):
A Closer Look
Which of these approaches is best?
Clearly the last approach would be superior if anyone was able to predict future returns with any accuracy, even on a probability basis.
In fact, if you throw a "zero weighting" into the approach mix—making decisions not to hold certain stocks along with decisions to hold others—this is the approach used by most actively managed mutual funds, as well as individuals managing their own portfolios.
These approaches are based on the belief that there are certain characteristics that make a specific equity more likely to have excess returns.
And it is likely that this is, in fact, the case. Historically, capitalization size and the price-to-book ratio have been successful overall predictors of excess returns. In general, however, methods for making return predictions with any degree of accuracy on a long-term portfolio basis have not been successful.
Market-Cap Weightings
Weighting based on market capitalization has been the dominant approach, and was the only approach for index funds until a few years ago.
If one’s mindset is that an index is meant to convey information about how well the stock market is doing, then a measure of the change in total market value of all stocks certainly makes sense. It is not a perfect measure because the price of the last hundred shares transferred is not necessarily an indication of the value of all shares if the company was bought out or liquidated, but it is probably the best approach for a simple system of measuring market value changes.
But wait! Most of us are not macro economists, and while we have a curiosity about the market in general, our real goal is to improve market returns. If we are going to invest in index funds, then our concern is an approach to stock weightings that will provide higher returns than the market, not necessarily be an accurate economic measure.
Equal Weightings
Equally weighting all stocks in an index has some significant advantages. Since small-cap stocks have outperformed over the long run, an index portfolio using equal weightings should outperform capitalization-weighted indexes, because smaller-cap stocks will be a greater percentage of the equal-weighted index than the capitalization-weighted indexes.
In addition, value stocks are underweighted in a capitalization-weighted portfolio, even though historically they outperform growth stocks, because their share prices are typically depressed (and remember, a stock’s market capitalization is based on the number of shares outstanding times its share price). So, equal weightings of all stocks will increase the weight of value stocks (relative to a capitalization-weighted portfolio) and thus the return of the portfolio.
While there are few equal-weight index funds, there are simulated returns for equal weighting as compared to cap weighting. You can view the comparison over various time periods by going to the Wilshire Web site.
Particularly illuminating is that, while the cap-weighted NASDAQ composite index was losing 30% of its value during 2000-2002, the average NASDAQ stock was up—and an equally weighted NASDAQ index was up 50% during the same period.
However, these comparisons do not include transaction costs, which may be significant. The major problem with equal weighting is that rebalancing is required; otherwise price changes will slowly move the portfolio toward cap weightings.
Experience with equal weighting indicates that rebalancing should not be done frequently. My opinion is that rebalancing should be done no more than once a year, and much of it can be accomplished when regular portfolio changes occur.
We have not rebalanced the Model Fund Portfolio as yet, and we have never made transactions just for rebalancing in the Model Shadow Stock Portfolio.
The only equally weighted index with any history (five years) is the Rydex S&P Equal-Weighted Fund (RSP), which is an exchange-traded fund based on the equally weighted S&P 500. It can be compared to the iShares S&P 500 ETF (IVV), which is an exchange-traded fund based on the cap-weighted S&P 500.
Five years is very short, and while over five years the Rydex ETF has outperformed iShares S&P 500 9.42% to 7.48% annually (market returns as of June 30), the last year and a half has been one of the few periods in which small-cap and value stocks have underperformed, so these results may not be indicative of the long-term differences.
There is some additional information that provides insight into the rebalancing problem. The Rydex Fund averaged 25% turnover a year, as compared to 5% for the iShares S&P 500. My estimate is that for very large cap, active stocks this would reduce return less than 0.1%, but for small-cap, illiquid stocks it could be significant, particularly with frequent rebalancing. Both cap weightings and equal weightings, to the extent they are widespread, will tend to make stock prices "stickier"—that is, make prices slower to respond to company performance.
Fundamental Weightings
Recently there have been a number of new funds, mostly exchange-traded funds, that use "fundamental weightings." This is weighting based on the fundamental values of the company (price-earnings ratio, price-to-book-value ratio, dividends, etc.). [I use "fundamental weighting" here to mean any weighting based on stock fundamentals; however, "Fundamental Index" is a trademark of Research Associates, LLC.]
This, of course, is the approach taken by most non-index mutual funds, except they use these fundamentals to determine whether to buy a stock or not (zero weightings), as well as to determine the weight.
When these fundamental-weighted funds were first introduced, there was a claim that there was an inherent mathematical advantage in such weightings because stocks deviate from fair value, and fundamental indexing takes advantage of this deviation. This would mean that fundamental indexing could increase returns even under efficient market assumptions.
After dozens of mathematical articles in academic journals, it seems doubtful that this is a valid claim.
Even though such an advantage does not exist, it is still possible that fundamentally weighted index funds may outperform their cap-weighted equivalents. In fact, unless rebalancing is too frequent and expensive, they are almost bound to outperform because they are closer to an equally weighted index and thus give greater weight to small-cap and value stocks.
In addition, if price relative to book value is one of the fundamentals used, then they will obtain some of the excess return of value stocks.
One can’t help but wonder, though: If you are a long-term investor and want higher returns, why not just select the value stocks of whatever market segment is of interest and equally weight them? Of course, short-term volatility would likely increase.
Get updates about the portfolio that has outperformed the market by 211.9%
since inception!
