Fundamentalists at the Gates: The Trouble With Cap-Weighted Indexes

There are fundamentalists roaming the financial landscape, but unless you're John Bogle or Burton Malkiel, you'd hardly describe them as terrorists.

The fundamentals embraced by these folk are those that Benjamin Graham made into a religion for value-minded investors: sales, income, book value and dividends. Their insistence on weighting the equities universe by these metrics instead of market capitalization has overturned the apple cart from which asset allocations have been dealt on Wall Street for the past three decades.

A couple of years ago, Robert Arnott, chairman of California-based money manager Research Affiliates, together with collaborators Jason Hsu and Philip Moore, published an article in the Financial Analysts Journal entitled, "Fundamental Indexation." Apples started to tumble as Arnott and his colleagues described a way to combat the tendency of market-capitalization indexes to overweight large growth stocks.

Weighting large U.S. companies according to their "economic footprint," said Arnott et al, produces a benchmark yielding higher average returns than the capitalization-weighted Russell 1000 and S&P 500 indexes. To boot, a fundamentally weighted market would be less likely to develop "bubbles" such as the gut-wrenching Tech Wreck of 2000.

Arnott and his merry band of fundamentalists offered some tantalizing data in their assertion. In backtesting between 1962 and 2004, the composite Research Affiliates Fundamental Index US 1000 [RAFI] annualized returns outpaced the S&P 500 by 197 basis points (1.97 percent). No small change, that.

With the "Fundamental Indexation" white paper in hand, wholesalers started scouring the land in search of potential users for an ETF tracking the RAFI--the NYSE-listed PowerShares FTSE RAFI US 1000 ETF (PRF)--in December 2005.

The fundamental indexing notion has apparently developed some traction: PRF has amassed a shade less than $800 million in assets in its first year, putting the fund 35th in the ETFZone size ranking of 301 portfolios.

It's telling that PRF's size lands it between Vanguard's Small Cap ETF (VB) and BGI's iShares Russell 3000 Value Fund (IWW).

"Arnott's findings imply that, in aggregate, investors have been systematically undervaluing large companies relative to small companies," says The Index Investor's senior editor Tom Coyne.

In this case, "large companies" refers to companies with high "fundamental" values (sales, income, book value and dividends), not market capitalization.

"However," notes Coyne, "if one believes that financial markets are generally efficient, this shouldn't happen, as market capitalization-based weighting should provide the best estimate of future returns."

Arnott admits to value and, to a lesser extent, small stock exposure in his RAFI benchmark. Whether that's a "tilt" can be debated, but tongues in the index world have been wagging, asking just how much of the excess return is due to these factors.

One line of recent research indicates that nearly two-thirds of the RAFI's excess return may, in fact, be attributable to value and small stock exposure; only a third seems to be inherent to the indexing technique itself.

There's a definite benefit to tilting toward value and small stock exposure in the domestic U.S. market, at least as seen through a filter devised by economists Eugene Fama and Kenneth French. The pay-off--measured by excess return over large stocks--has averaged 8.4 percent per year since 1957 (see Chart 1).

Clearly, a leaning towards value- and small-stock investing could have produced outsized returns in the past, though there could times when it would be very uncomfortable to defend such a strategy. In 14 years, or 28 percent of the 50-year history tracked in Chart 1, the net risk premium was negative.

There's considerable variance in the size of these premia as illustrated in Chart 2 For the past five decades, the median small-cap premium has been 2.0 percent per year while the median value premium amounted to 7.3 percent.

"The interesting thing, however, is that these premia are negatively correlated," says Portfolio Solutions CEO Rick Ferri from Troy, MI. "Correlation tests to see if the size and value premium are the same risk. The inconsistent and mostly negative correlations between size and value show that they are separate risks. Having both small stocks and value stocks in a portfolio (or using a small value fund to kill two birds with one stone) is good diversification."

That said, portfolio runners are asking to what extent RAFI US 1000 mines for these premia. "It's a known fact that the RAFI US 1000 has a value tilt which explains most of its outperformance over the S&P 500," says Michael Markov, CEO and director of research at the Markov Processes, International consultancy. Backing up Markov's contention is the RAFI's concentration of returns to the right and downward of the S&P 500's in the style map shown as Chart 3.

For Tom Coyne, the implication is obvious:

"While it is clear that Bob Arnott, with hindsight, has discovered a theoretically profitable anomaly," he says, "but what basis is there for assuming it will continue in the future now that it has been publicized?"

Indeed, for the RAFI to continue to outperform the S&P 500, says Coyne, two assumptions have to made. "First, whoever has been making the valuation errors that gave rise to the superior historical returns will continue making them and second, other investors will not arbitrage away the potential excess returns by bidding up the prices of larger companies' stocks."

Coyne thinks the odds are too long to bet on the RAFI's continuing outperformance.

Arnott, not surprisingly, takes issue with such criticism.

"The supposed small-cap bias is a myth," he declares. "Our average market cap is $76 billion, compared with $81 billion for the Russell 1000 and $1.1 billion for the Russell 2000 small-cap index."

His argument on the value tilt is a bit more nuanced.

"Is the cap-weighted market portfolio style-neutral, or does it have a growth tilt?" he asks. "After all, if a stock has twice the P/E ratio of another, its relative weight is doubled. Whether fundamental indexing has a value bias or cap-weighting has a growth bias, it's a fair statement that the RAFI has a value bias 'relative to' cap-weighting. That bias will be small when the gap between growth and value is small, and large when the gap is large. So, the value bias self-adjusts to mirror the opportunity set."

It's the "opportunity set" itself that galls some pundits, however. Burton Malkiel, for example, questions the RAFI's very classification as an index. Constructs like the RAFI, he sniffs," are not 'indexes' in the sense that all investors can own them."

To better visualize Malkiel's argument, imagine the equity market is comprised of only two stocks: the large stock has a market capitalization of $200 billion while the small stock's worth is $20 billion. The large stock would represent 91 percent of a market-cap weighted index; the small stock earns a nine-percent weighting. This $220 billion asset base represents the entire opportunity set for market participants. Thus, anyone can track the index by dropping 91 percent of his capital into the large stock and nine percent into the small stock.

It's not so tidy in a fundamentally weighted universe. If a fundamental filter weights one stock at 70 percent and the other at 30 percent, the only thing that keeps the universe from blowing itself apart is the dearth of index-minded investors. If everybody took to indexing, thirty percent of the market couldn't possibly invest in the smaller stock near its current market price.

Such fine distinctions aren't keeping portfolio runners from considering fundamentally weighted products for their clients, however.

"I think both [cap-weighted and fundamentally based indexes] have their place in the market and portfolios," says Butler Wick & Co.'s Mark Manning. "One thing you need to watch on cap-weighted funds is that they sometimes have very large positions in only a few stocks. If those are the stocks that you want to target that may be fine, but if you're buying a fund for diversification, some may not fit your needs. We use a combination of capitalization- and fundamental-weighted index funds in our managed portfolios, depending on what area of the market we're trying to target."

Ron Surz, president of California-based consultancy PPCA, Inc, offers a road map for advisors attempting to marry fundamental- and capitalization-weighted products in a portfolio. "First I’d run a returns-based style analysis on the RAFI to understand it better.:"

Returns-based style analysis, a concept introduced by Nobel laureate William Sharpe in 1988, uses an optimizer to identify a blend of indexes that behaves most like a portfolio’s investment returns.

Surz's firm specializes in providing such performance attribution to financial advisors.

"It’s a 'walks like a duck' approach," says Surz, noting that each portfolio has a unique cross-section that can be described by an amalgam of benchmarks. "The indexes that are shown to the optimizer have come to be known as a 'style palette.' The term suggests painting; we’re painting the effective style mix of the portfolio manager."

"To integrate the RAFI into an overall asset allocation, I’d ask the optimizer to match the cap-weighted broad market index with the RAFI and my other managers-–my style palette," Surz suggests.

For DTB Capital's David Krein, making a fundamentally weighted product the centerpiece of an investment strategy may mean paring back other exposures as an accommodation. "If I was to use a fundamental index as the core of a portfolio," he says, "I'd first look at its correlations to the other exposures it might crowd out to figure out the rest of my asset allocation."

Krein says the RAFI's real-time history is too short to really understand its relationship to other market measures. "It would be difficult to justify any allocation on pro-forma track record alone," he says. "It takes time to develop an intuitive 'feel' for the underlying dynamics in differing market conditions."

Still, Krein suggests that it would be acceptable to allocate a small percentage of a portfolio to the RAFI as direct substitution for an existing US mid-to-large cap equity allocation.

The Russell 1000 is a natural--but certainly not the only--target. The top six holdings by weight are nearly identical for the RAFI and Russell 1000. Collectively, these half-dozen stocks represent about 12 percent of each index's weight.

Not one to sweat the small stuff, Krein says he wouldn't spend a lot of time on formal analysis for a small allocation, but cautions that his approach "would likely change as the RAFI develops a track record for delivering alpha."

Note: This article originally appeared in the April 2007 issue of Registered Rep. Magazine.