The O’Doul’s of Value Investing

Note: Please also read part two which corrects for a few outliers in the below results

Traditional index funds are gaining record market share. The weight of each stock in these index funds is determined by its market capitalization. But bigger is not better.

Enter fundamental index. These improved indexes weight stocks by variables “smarter” than size, like book value and sales. But these changes are just the first step towards true value investing. Fundamental index is the O’Doul’s of value investing*.

I don’t really get why anyone would prefer fundamental index to more concentrated, pure-play bets on value or shareholder yield. Let’s see why.

Here are a number of ways you could weight stocks in an index like fund. Key point: every one of these strategies owns (mostly) the same stocks (all large U.S. domiciled stocks), but weights them according to different variables. The only reason a stock would be included in one but not another is because it has a negative value for the variable (say, a company with negative earnings or cash flow, which screw up the math).

smart beta weightingMarket cap is the worst variable by which to weight stocks.

“Fundamental” weights do better with similar volatility and thus improved Sharpe ratios.

Value weighting (using “yields” like earnings yield (earnings/price)) to do the weighting delivers the best returns. Using shareholder yield to weight stocks works great: solid returns with lower than market volatility.

Think of it this way: the benefit of fundamental index is that it removes price from the equation and replaces it with some sort of fundamental value (sales, earnings, book value). This removes a persistent bias towards more expensive stocks. But these results show that for earning better returns, you want to re-introduce a price-bias, just in the other direction: towards cheap stocks.

Here’s the thing, I ran this in 20 minutes–yet market cap weighted index funds are garnering record inflows.

I realize that the future may be different: maybe expensive stocks, or those paying no dividends and issuing shares, will dominate for decades. I doubt it.

There are no costs (fees, transaction costs) baked into these results, but at the very least, these simple results expose a key flaw of market cap weighting.


*I had a more vulgar joke in here but thought the better of it, thanks to Tobias Carlisle for talking me out of it and giving me a replacement