Author shares new approach to weighing stocks

We take for granted the notion that stock market indexes represent the stock market. The notion is a pillar of modern investment theory. A nifty thing called the Capital Asset Pricing Model led to a fistful of Nobel Prizes for the string of economists that helped assemble it.

Never mind that the model has long since been shown to have such deep flaws as to render it almost useless. There have been more academic papers written to tear it down than were written to build it up. To my continuing consternation, the model serves as the heart of all financial adviser training programs, including the Certified Financial Planner designation, which I once held.

The Capital Asset Pricing Model assumes that our stock portfolios contain a proportional share of all the stocks available. It further assumes that these proportions are weighted by the true value of the companies. In the real world, we cannot know the true and genuine value of any stock. The best we can do is scratch out some numbers, make an educated guess and bid the stocks up or down out in the market. We then use the market’s determination of value as a stand-in for the true value.

We end up with a portfolio that is weighted by the market capitalization of each stock. This is how the S&P 500 and other indexes are constructed. If General Electric’s market value is twice that of Intel, then your index fund investment will have twice as much of your money in GE as in Intel. Never mind what GE and Intel might really be worth.

The Capital Asset Pricing Model requires a cap-weighted assumption to hold itself together. Further, index funds require this assumption in order to manage your money. A neat feature of a cap-weighted index is that the fund manager doesn’t need to make daily trades as market movements change the relative market values of the companies. The adjustments to the portfolio are automatic.

Since the investment theorists and the index fund companies both require a cap-weighted portfolio, this approach has become part of investment dogma.

The problem is this: It is a bad way to invest in the stock market.

In a groundbreaking paper released this month, RobertArnott, chairman of Research Affiliates and editor of the FinancialAnalysts Journal, lays bare a problem that has gnawed at me for years. A capitalization-weighted stock market index places undue emphasis on stocks whose prices have been lifted to irrationally exuberant levels by the market.

As I noted above, we can never know the true value of a stock; all stocks trade either too high or too low. Our job as investors is to buy the ones that are too low. Buying the other side of the coin is a road to disappointment and ruin. Since it is almost impossible to distinguish the high from the low, we buy a diversified fund that invests in, or targets, the entire index.

Arnott makes the logical deduction that a capitalization-weighted stock index or fund means that we are always over-exposed to expensive stocks, and always under-exposed to cheap ones. Read that again for a moment. While we might not know which stocks are which, an index ensures that we will own too much of the bad ones, and not enough of the good ones. I implore you to read that one more time.

Arnott goes on to examine what other measures of company size might make good weightings for an index. For example, most of us would judge the size of a company by something like its total sales, or how many employees it has or the value of the assets on its books.

Arnott found that any of these, and other, alternate measures of company size produced an index whose investment return beat the S&P 500 by an astounding 2 percent per year, on average, since 1963. Since 1999, the average alternate index would have beaten the S&P 500 by 40 percentage points. Arnott shows that it works in bull and bear markets; inexpansions and recessions.

Again, these are the same stocks, just weighted differently. Rather than weight them according to over- and under-valuation, like the S&P 500, he weights them based on how big the companies actually are.

These alternative size-weighted indexes help to ensure that we don’t end up over-weighted in overpriced stocks and under-weighted in cheap ones. As of today, there are no indexes or funds available that follow this new weighting scheme, but I am confident that Arnott is working on it.

When more news develops, I’ll write about it here.

Rick Ashburn manages investments for private clients. Write to him at or 7777 Fay Ave., Suite 230, La Jolla, 92037.