Arnott: The RAFI Five-Year Scorecard

January 31, 2011

RAFI strategies may play an important role as a low-cost 'core' part of investors' equity portfolios.


When the Fundamental Index® concept was introduced, it was met with fierce attacks. Critics decried its backtested results as data-mining or said the approach was just a repackaged value investment process. Five years after the first RAFI indices went live, the proof is in: The methodology has generated superior performance during a period when value has lagged growth all over the world.

In a relatively short span of time, the Fundamental Index approach has revolutionized passive investing and has served as an important milestone in the evolution of our investment thinking, helping spawn a new field of investing—that of Alternative Beta or Strategy Indices. In this issue we show that live RAFI results support our earlier research and the robustness of the Fundamental Index methodology.

RAFI’s Better Beta

We published our initial research on the Fundamental Index concept in the March/April 2005 issue of the Financial Analysts Journal, showing how the methodology outperformed in U.S. markets during a 43-year span ending December 2004, based on a simulation.1

That initial research was extended by Nomura Securities to cover the 23 developed markets in the FTSE Developed Index.2 The results are summarized in Table 1. Impressively, the Fundamental Index approach generated excess returns over cap weighting in 21 of the 23 developed market applications over a 21-year period. The average outperformance in developed large company markets was 2.7%. With these results in hand, we concluded that an investor could achieve a return of 2–4% over cap-weighted indexes in developed markets, over a complete market cycle.

For a larger view, please click on the image above.

Since FTSE launched its FTSE RAFI family of indexes in November 2005, we’ve lived through two bull markets sandwiched around the biggest global bear market since the 1930s. If this isn’t a full market cycle, we are not sure what is. Value stocks won handily in 2006 and the middle quarters of 2009, flanked by years of mostly growth equity outperformance. So how has the Fundamental Index approach fared over this stretch?

Using live results, 19 of 23 countries added value via the Fundamental Index methodology. The average excess return per country has been 2.0%, as shown in Table 2. The FTSE RAFI US 1000 Index produced an excess return of 2.3% per annum, just a bit better than our original study showed.3




Table2 Live Annualized Returns Since Inception of FTSE RAFI


The live results fall into in the 2–4% range we discovered in our initial research. But admittedly they are at the lower end. While an excess return of 2% is respectable, why didn’t we fully match the long-term backtested results? For that, let’s turn our attention to the other major criticism of the Fundamental Index approach—the value effect.

Positive Results in a Growth-Oriented Market

Critics have contended that the Fundamental Index methodology derives its benefit from a value tilt. We don’t disagree with that view as they’re half right. Let’s explain. Suppose there are two stocks with identical sales, book values, cash flow, and dividends. The first stock, Growthy Shares Inc., trades at twice the market multiple due to its outstanding recent operating results and the investors’ resultant high expectations for future growth. Meanwhile, Unloved Value Co., with a stream of recent bad news, sells at half the market multiple. Cap weighting doubles the weight of Growthy Shares and halves the weight in Unloved Value Co., relative to their economic scale, despite the companies being the exact same size. Repeat this exercise across the whole market and the result is a strong growth tilt for a cap-weighted index portfolio. Meanwhile, a fundamentally weighted index portfolio doesn’t share that tilt to growth; it matches the look and composition of the economy. From a cap-centric point of view, the Fundamental Index portfolio does have a value tilt, but it’s a special value tilt—an exact mirror image of the market’s willingness to pay up for perceived future growth opportunities.

It is well documented that value has historically outperformed growth.4 Our own analysis has found roughly one-fourth of the RAFI method’s excess return is attributed to a static value tilt as outlined above. A majority of the excess return stems from contra-trading against the fads, bubbles, crashes, and constantly shifting expectations and speculations at work in the capital markets.

For the 21-year period ending December 2004, the global developed equity market witnessed a value premium of 1.3%, thus acting as a tailwind for the typically value-oriented RAFI performance.5



In opposite fashion, value acted as a headwind during the live period. In the United States, large-cap value stocks returned +1.4% from November 30, 2005, through December 31, 2010, while their growth counterparts were up +3.6%.6 Thus, the negative premium for value stocks has been about half of this 2.2% difference per year in the United States. Similarly, in developed markets, the value premium was –1.1% per year for the past five-plus years (See Figure 1).


Figure1 RAFI Excess Returns & The Value Premium


Our simulated results showed an average RAFI excess return of 2.7% per year vis-à-vis cap weighting in a period of 1.3% value outperformance, so clearly RAFI performance is more than just value. The RAFI live index results have been even more impressive, achieving 2.0% excess returns during a time when value indexes lagged the broad markets by more than 1% per year! The RAFI approach beat the cap-weighted value benchmark by well over 300 bps, winning while value was losing, compounded over five years.

How is that possible? One of the important features of the RAFI investment process is the annual reconstitution back to a company’s fundamental scale as defined by a composite of sales, cash flow, book value, and dividends. This rebalancing, a key differentiator of RAFI indices, forces the portfolio to trim stocks whose prices recently outperformed their fundamentals and add to those stocks whose prices have underperformed the businesses’ economic footprints. The market is constantly changing its mind as to which companies are growth stocks and which ones are value stocks, and how much premium or discount each company deserves. A conventional value index responds by adding and dropping companies as they fall in or out of the value camp; but the weight is always the cap weight if it’s in the index. A Fundamental Index portfolio will adjust over- or underweights relative to the cap-weighted market to reflect the constantly changing premium or discount reflected in the share price. The size of the business is merely a convenient, and economically meaningful, anchor to use for rebalancing.

On a style basis, the RAFI approach increases its value exposure when value has recently underperformed and is cheap (i.e., investors are rewarded with a high forward-looking value premium). This phenomenon was vividly seen in 2009 when the RAFI strategies—oblivious to imminent Armageddon—rebalanced into the very finance, industrial, and consumer discretionary stocks that were ostensibly poised for extinction. Meanwhile, the RAFI approach reduces its value exposure when value has recently outperformed and is expensive. This contra-trading process is the reason why a value-tilted portfolio can win—even handily—in a secular growth led market.



The contra-trading impact is illustrated in Figure 2, which shows where RAFI excess returns are plotted given the value premium for the five years of live performance. The dashed line that starts in the lower left quadrant and rises to the right is how the Fundamental Index approach would have performed if it were a pure value strategy. Points above the dashed line are one-year periods when the RAFI strategy beats the pure value approach; points below the line are one-year periods when the RAFI approach loses to the pure value strategy.


Figure2 Impact of Contra-Trading, 2005-2010

Two things stand out. First, the RAFI approach generally plots above this dashed line…it beats the straight value strategy. Overall, the live batting average—or success rate—of RAFI wins is 80% of one-year periods. Second, the RAFI strategy outperforms as frequently when growth wins (the left half of the chart) as when value wins (the right half of the chart).


The Fundamental Index methodology does not work all the time; no investment process does. However, our research found that, over full market cycles, the RAFI methodology is strikingly effective at adding value over the cap-weighted benchmarks. A live five-year win rate of 83% in developed markets—during a span of substantial value underperformance when critics said the Fundamental Index methodology should have lost—is powerful evidence

Any new scientific hypothesis—or backtested analysis—warrants a skeptical look. The Fundamental Index idea is no exception. We think the live results, however, provide strong empirical evidence that the RAFI idea holds merit and will begin to open the minds of many of the early disbelievers. Further, these results suggest that RAFI strategies can play an important role as a low-cost “core” part of your equity portfolio—a better beta achieved with all of the advantages of traditional cap-weighted indexes.



1. Robert D. Arnott, Jason Hsu, and Philip Moore, 2005, “Fundamental Indexation,” Financial Analysts Journal, vol. 61, no. 2 (March/April): 83–99.

2. H. Tamura and Y. Shimizu, 2005, “Global Fundamental Indices: Do They Outperform Market-Cap Weighted Indices on a Global Basis?” Nomura Global Quantitative Research (October 28).

3. The statistician in me must admit this is fortuitous. While it is dead on with our historical average, the range of all five-year excess returns from the 1960s is considerably wider and investors should not assume all or even most five-year holding periods would produce such a similar outcome.

4. Eugene F. Fama and Kenneth R. French, 1992, “The Cross-Section of Expected Stock Returns,” Journal of Finance, vol. 47, no. 2: 427–465.

5. Calculated by MSCI World Value Index (13.0%) – MSCI World Index (11.7%) from January 1984 through December 2004.

6. As measured by the Russell 1000 Value and Growth indexes.

© Research Affiliates®, LLC 2011. The material contained in this newsletter is for information purposes only. This material is not intended as an offer or solicitation for the purchase or sale of any security or financial instrument, nor is it advice or a recommendation to enter into any securities transaction. The information contained herein should not be construed as financial or investment advice on any subject matter. Neither Robert D. Arnott nor Research Affiliates and its related entities warrants the accuracy of the information provided herein, either expressed or implied, for any particular purpose. Nothing contained in this newsletter is intended to constitute legal, tax, securities, or investment advice, nor an opinion regarding the appropriateness of any investment, nor a solicitation of any type. The general information contained in this newsletter should not be acted upon without obtaining specific legal, tax, and investment advice from a licensed professional.

ROBERT ARNOTT, Chairman and Founder of Research Affiliates, LLC.

Find your next ETF

Reset All