Last Friday, I had the pleasure of attending a talk by Rob Arnott, the creator of the Research Affiliates Fundamental Indexes (RAFI).
If you’re not familiar with fundamental indexing, here’s a primer. Most traditional indexes are weighted by market capitalization, which means that a company’s influence is determined by its size (as measured by the number of shares outstanding, multiplied by the price per share). If a stock doubles in price, its influence in a cap-weighted index will also double. Similarly, if a company’s stock price declines, so does its weight in the index.
This methodology makes perfect sense if the goal of an index is to be a barometer of the market. Indeed, the cap-weighted S&P 500 was created in 1957 to measure the performance of the U.S. stock market, and other traditional indexes (including the S&P/TSX Composite in Canada) perform the same role.
However, proponents of fundamental indexing point out that these benchmarks were never designed to be the basis of an investment strategy. And in that respect, they argue, cap-weighted indexes have a major flaw: they give the most weight to overvalued companies and the least to undervalued companies.
The trouble with bubbles
The most glaring example of this problem occurred during the dot-com bubble. As the price of technology companies inflated to absurd levels, so did their influence in cap-weighted indexes. Meanwhile, profitable businesses that fell out of favour saw their stock prices and their index weight plummet. When the bubble finally burst in March 2000, the S&P 500 was decimated.
The RAFI fundamental indexes remove stock prices from the equation. They assign each company a weight based on four factors: total sales, book value, cash flow and dividends. What’s more, they use an average of these factors over five-year periods to smooth out any short-term aberrations. The result is an index that measures companies based on their footprint in the economy, rather than their footprint in the market.
Nortel is the most obvious Canadian example of how big a difference this can make: in 2000, the company made up more than 29% of the S&P/TSX Composite, but would have been less than 4% of a RAFI index based on its fundamentals.
A fundamentally different method
As the ETF marketplace grows, all manner of new indexes are cropping up in attempt to address the supposed flaws of cap-weighting. Many have more to do with marketing than anything else. But the RAFI indexes are in a different category. While they don’t have universal support in the academic community, they are based on peer-reviewed research and have been endorsed by none other than Harry Markowitz, the Nobel laureate and creator of Modern Portfolio Theory, the rock upon which index investing is built.
As Arnott described during his talk, indexes weighted by company fundamentals would have outperformed cap-weighting by an average 2.8% annually in 22 of 23 developed countries since the 1980s. His book, The Fundamental Index, goes in to much more detail and makes a compelling case for the strategy. It’s a must-read for anyone thinking about building a portfolio with the RAFI-based funds.
The promise of fundamental indexing is enormous. However, it’s important to remember that the long-term performance of the strategy is hypothetical. The data assume that funds based on the RAFI indexes would have tracked their benchmarks perfectly—and that’s always a big assumption. The first ETFs and mutual funds based on the RAFI indexes have been live for only five or six years, with mixed results. Later this week I’ll look at how some of them have stood up in the real world.
I am looking *really* forward to reading the next few posts. Keep ’em coming!
I’d like see a comparison of the various market weighted ETFs vs the fundamentally weighted etfs. For example XIN vs CIE.
@Couchpotato: What happens to TSX based ETFs if the LSE and TSX merge?
“…outperformed cap-weighting by an average 2.8% annually.” What if they chose the 1990’s as a start. Note how non-specific 1980’s and 1990’s sounds and how specific 2.8% is.
Higher MER’s and poor tracking (probably due to smaller less liquid companies in the ETF given our small market plus currency fluctuations for foreign markets) would reduce this advantage. Also many cap-weighted ETF’s have a cap of 10% of assets in any one security – still too much in my opinion.
Will be looking forward to what you conclude as I have been tracking for almost a year now two complete portfolios, one that uses cap-weight indexing and the other fundamental weighting. The on-going evolution can be seen at the bottom of my blog. As of the instant I write this, the fundamental portfolio is ahead by about $1000 in total on a $100,000 initial investment amount.
Great post! I’m looking forward to your followup and comparisons of Fundamental Indexing (RAFI) ETFs to the conventional market cap-weighted ETFs and Index Funds. This will be a very useful comparison, withvery useful information for investors. Well done!
Be nice if TD e-series comes out with asome RAFI Index Funds, wouldn’t it ?
@CanadianInvestor: I have been following your real-world comparison, which will be useful, though it will take along time before you can draw any meaningful conclusion. I’m more interested in comparing the real-world RAFI returns to the index returns.
@Ninja: I would be very surprised if TD ever expanded its e-Series offerings. Besides, I would guess that Pro-Financial has exclusive licensing rights to the RAFI indexes for Canadian mutual funds: http://www.pro-index.ca
Did anyone (cough *you* cough) bring up Claymore’s considerable tracking errors during this talk? :)
Not sold on fundamental indexes especially for those of us who hold the majority of our assets in non registered accounts. The increased turnover is bound to have substantial tax implications.
Looking forward to the rest of this series.
@gsp: I didn’t bring up the tracking errors at the talk, though I have reported them on the blog in the past and will certainly get into that in my next post.
Can I say Fundamental Indexing = Active Indexing?
By applying these 4 factors it looks like they are doing stock picking.
@Invetsor2011: I wouldn’t go so far as to describe fundamental indexing as active. The apply a rules-based screen to select the stocks and then weight them according to a formula. There is no manager making individual selections or second-guessing the strategy based on a view of where the market might be headed.
Remember, cap-weighted indexes have their own set of rules for including and exclusion, so they’re not entirely free of management.
If you were looking to overweight value within a passive indexing portfolio, would fundamental index funds be a suitable substitute for value index funds? Or are they better served as a market-cap weighted index substitute with a “value tilt”?
@MKF: Fundamental indexes will certainly give your portfolio a value tilt and would indeed be a good substitute for cap-weighted value funds.
Thank you for that article, I’m working on my education in ETFs before I take the step of using real money and your articles are a windfall of gems.
I have a question on cap-weighted funds vs. value indexed funds. Would this be another way to position yourself for bull vs. bear markets? If you think a market is veering from fundamentals into that strange area of bubbles I’d think the cap-weighted funds would capture more of that run. On the other hand if you think the bubble has run its course positioning in value indexed funds should do better.
@inglish: Thanks for the comment. The problem with your reasoning is that you’re assuming you can identify a bull or bear market in advance. Certainly, if it were possible to switch to cap-weighted strategy during periods when growth stocks outperform, and then to a fundamental strategy when value outperforms, you would do extremely well. But these periods are only identifiable in hindsight.
What are your thoughts on the new Powershares RAFI ETFs, specifically PXLV, PXMV, PXSV?
One additional question, do you happen to know of any REIT ETFs that uses an alternative weighting strategy for US REITs?
@IndexInvestor: I believe that the PowerShares ETFs you mention just relaunched this week with new index benchmarks, so it’s hard to say anything definitive about them at this point.
There is a PowerShares REIT ETF, though it does not follow a RAFI index:
Hi Dan, now that 3 years have passed since this was written, how have RAFI based efts compared to traditional cap weighted ETFs. If you have already commented on this, I apologize in advance for missing it.
@Lawrence: The returns of the RAFI funds have generally been good over the last three years. I haven’t done any thorough investigation, but the returns are all published on the ETF websites so they would be easy enough to compare if you’re interested.