Fundamental indexing: A magic formula?

Passive investing does not preclude value investing.

If you’ve read this blog for any period of time, you know that I have some paradoxical views on how we should invest. I think markets are mostly efficient, but that periods of inefficiency (a la the housing and dot-com bubbles) can last for extremely long periods of time. I think it doesn’t make sense to buy a company at any price, but that actually attempting to buy companies at “bargain” prices can lead to getting severely burned.

Basically, I believe in value investing, but that I (and most people) would suck at it.

Enter Rob Arnott, chairman of Research Affiliates. Arnott and his company have designed a series of mutual funds and ETFs that seek to follow value investing principles without having an active manager at the helm. In your typical value mutual fund, you’d have a value manager and team of analysts who ran stock screens, studied companies, and made judgment calls on what to buy and what to pass up on. But Arnott’s funds, detailed well in this Fortune article, remove the human element. They simply buy the whole index, but weight how much they buy of each company based on value metrics, like price-to-sales ratios and dividend yields. The higher the dividend yield, for example, the more of a certain company they would buy.

The flaw with index funds.

When you buy a share of the Fidelity Spartan Total Market Index Fund, you get a portfolio of companies weighted entirely on how big those companies are by market capitalization. ExxonMobil is worth $270.4 billion. Microsoft is worth an even $217 billion. So when you buy a share of Fidelity’s fund, you automatically buy a slightly larger chunk of Exxon than you do of Microsoft.

That might not sound crazy, until you start to think about how that’s ended up in practice.

Dot-com bubble. At their peak in 1999 and 2000, technology companies made up 30% of the S&P 500. If you bought an S&P 500 index fund at that time, you would have effectively bought two-and-a-half times more of eToys.com than you would have bought of Toys R’ Us.

Financial bubble. In 2006, financial companies made up more than 22% of the S&P 500. Banks minted money with leverage and clever ways to move debt off their balance sheets. And yet, when you bought shares of your favorite index fund, your biggest purchases were of the AIGs, Citigroups, and Wachovias that we know and love today.

When I started to invest, I had the admittedly mushy idea that buying an index fund was like buying a slice of the U.S. economy. If the U.S. did well overall, my fund would do well. I was kind of right, but why not design a fund tried to weight money in each company based on their economic contributions to the economy as a whole rather than to how high-priced they were?

Ok, this is where Arnott really steps in.

That’s exactly what Arnott’s boss asked him to do a decade ago. And a couple bubbles later, Arnott’s company manages more than $52 billion and growing.

The idea is a tough sell for a number of reasons. Let’s start with the institutional problems. If you have a financial advisor or you read a personal finance magazine or you read a personal finance blog (ahem), you’ve read for the last few decades that the only cheap, foolproof way to invest is to put your money in market-cap weighted index funds, like the Fidelity Total Market Stock Fund I mentioned above. The first of those institutions to adopt fundamental index investing will be going out on a limb, and their customers (i.e. you) will judge them harshly for that decision if it falters in the first few years (even though, as with any investing strategy, it’s really designed to work over decades).

Pension fund boards, which control the real money, don’t want to walk onto that limb either. In the end, their performance will be judged by how they do against the traditional, market-cap weighted indexes. So if they invest in the index and do poorly, at least they’re doing poorly with everyone else. If they invest in the fundamental index and do poorly, their situation is a bit more lonely.

And finally you. When you analyze pretty much any mutual fund, you compare its performance to the market as a whole. That’s how charts are displayed on Morningstar, in magazines, and even in mutual fund prospectuses. The “default” will always be a market-cap weighted index until that changes.

How it works

The Fortune story glibly calls it a magic indexing formula, but the reality is a bit more mundane. This is no black box: Arnott and his team detail how they devised their investment strategy here.

In short, they found that the best measures of a company’s economic contribution to an index are its sales, dividend, book value, and cash flow. Lacking a superior alternative, they weighted each factor equally. (Seriously, I can’t find anything with a less arbitrary explanation.) So now, when you buy a share of the PowerShares FTSE RAFI US 1000 ETF you won’t be buying the biggest companies by market value, but the most superior in terms of dividend yield and those other three factors.

ExxonMobil, for example, still happens to be the biggest holding of the RAFI ETF just mentioned, but that’s followed closely by General Electric and Bank of America, the latter of which isn’t even in the top five of a traditional Russell 1000 fund.

If you backtest Arnott’s strategy, you basically find that it would have outperformed traditional indexing by about 2 percentage points per year. That’s huge. That’s like turning $100,000 into $387,000 over 20 years instead of $265,000. Backtesting is no measure of how a strategy will perform in the future. Critics of formula investing have gleefully devised backtests that showed companies with arbitrary attributes, like beginning with the letter “J”, performed better over time. But given that the funds have only existed for about a decade, there’s not much in the way of actual historical performance to look at.

Avoiding some of the pitfalls of active value investing

And fundamental indexing also sidesteps some inconvenient truths of actively-managed mutual funds. For one, they’re actively managed by humans—who have all the same fears of loss, panic attacks, and bouts of greed that you do.

Partly because it avoids those humans and all their analyst human sidekicks, fundamental indexing also avoids the massive fees that active funds command. The RAFI ETF mentioned above, for example, carries a paltry expense ratio of 0.39%. That’s more than twice as high as a comparable index ETF (which might only be 0.15%) but a fraction of the 1% to 2% that a typical active fund would charge.

Part of me also thinks the ratio is so high because there’s such a lack of competition. The only other fundamental indexing type shop that I know of is WisdomTree, and they don’t really offer products that you could compare directly to those of Research Affiliates. As they did with those of regular index funds, I bet expense ratios on the fundamental indexes will drop over time.

So what should you do?

I don’t have a definitive answer. I think one of the strengths of this blog is its wishy-washiness. Unlike most other areas of personal finance, investing has few clear-cut answers. I can’t say, “Invest only in market-cap based index funds,” in the same way I can write “Avoid high-interest credit card debt.”

But I do know that buying more of a company simply because it’s more expensive than the next company makes no sense. Until recently, we had to live with it or pay high fees to an active fund manager who suffered from the same behavioral biases that make the rest of us crummy as investors. This strategy doesn’t seem to suffer from that problem.

Anyone else have any ideas what might make this kind of strategy stumble or make sense?

Tagged as: fundamental indexing, index funds, Investing, stock market