The fundamentals of fundamental indexing

Market-cap weighting reflects all market participants' views, allowing passive investors a free ride on the collective wisdom of all active investors. But there is a risk that it could also lead a portfolio to overweight the most expensive areas of the market. Fundamental index funds attempt to circumvent this potential problem by severing the link between a stock's portfolio weight and its price.

Market-cap-weighted indexes are inherently efficient and inexpensive to implement. A stock's weight in a market-cap-weighted index moves with its share price. Investors save on transaction costs because the index's largest positions are naturally the biggest companies with the most-liquid shares.

Unlike market-cap-weighted indexes, fundamental indexes size their holdings by metrics such as company revenue, earnings or dividends instead of share prices. "Fundamental" here is a bit of a misnomer because these metrics are only "fundamental" in that they're usually metrics that traditional fundamental stock-pickers use to evaluate companies.

But because fundamental indexes use the absolute level (such as total revenue) rather than a price-relative ratio (such as price/sales) to size positions, fundamental indexes maintain a similar large-cap orientation to market-cap-weighted indexes, while breaking the link to share price. The largest companies generally have the most revenue/earnings/dividends and receive the largest weightings in the index. Fundamental indexing is an intuitive weighting approach, but what does it really bring to the table?

Value and size rebranded?

The primary argument against fundamental indexing is that it is simply a value strategy repackaged as something new. Most of the time when a portfolio deviates from market-cap weighting, the alternatively weighted portfolio increases the weight of smaller-cap stocks and decreases the weightings of larger-cap stocks. The value tilt inherent to fundamental indexing stems from two sources. First, by weighting its holdings by a fundamental metric (such as revenue), the strategy overweights stocks trading at a lower multiple of that metric. Second, when the strategy rebalances, it trims stocks that have increased in price relative to that metric and adds to those names that lagged the market relative to the fundamental metric. While this weighting approach favours value stocks, it has less consistent exposure to value than market-cap-weighted value index alternatives because it does not explicitly target value stocks.

In a paper titled Why Fundamental Indexation Might—or Might Not—Work, Morningstar's Paul Kaplan highlights that fundamental indexing implicitly assumes that all companies should trade at the same valuation multiples. Efficient market supporters counter that stocks should, and empirically do, trade at different multiples based on their risk and growth profiles. Kaplan also asserts that without knowing a stock's fair value or fair valuation multiple beforehand, it is impossible to conclude in advance that a stock's current market price is too high or low, and whether stocks that command high valuations are overpriced.

Size and value everywhere

The first fundamentally weighted exchange-traded funds offered in Canada were launched more than a decade ago by what was then the Claymore family of ETFs, which is now part of iShares. These ETFs leverage Research Affiliates' RAFI indexes, which combine several fundamental metrics such as cash flows, revenue and dividends rather than a single metric to weight holdings. More recently, Invesco launched several PowerShares ETFs that also use RAFI Indexes.

There are several suites of fundamentally weighted ETFs listed in the United States that have live track records nearly a decade long. WisdomTree offers both dividend- and earnings-weighted ETF suites for the large-, mid- and small-cap U.S. market segments. Oppenheimer provides ETFs that weight their holdings by revenue. And PowerShares offers fundamentally weighted ETFs whose underlying indexes use the same RAFI methodology as their Canadian-listed counterparts.

I used Portfolio Visualizer, a free web-based tool, to test the idea that fundamental indexing benefits from small-size and value tilts by regressing several U.S.-listed funds' monthly returns against monthly small-size, value, momentum and profitability factor returns per the French Data Library. This analysis used monthly return data from April 2008 (the earliest common inception date for the selected funds) through May 2017. The table below displays the regression coefficients (or factor loadings) of each fund's returns against selected factor returns.

	Fundamental index ETF French-Fama factor loadings
	Name	Ticker	Small size	Value	Momentum	Profitability
	WisdomTree US LargeCap Dividend	DLN	-0.28	0.22	-0.01	0.15
	WisdomTree US Earnings 500	EPS	-0.13	0.03	-0.03	0.07
	Oppenheimer Large Cap Revenue	RWL	-0.07	0.12	-0.06	0.13
	PowerShares FTSE RAFI US 1000	PRF	-0.05	0.23	-0.15	-0.02
	iShares Russell 1000	IWB	-0.08	-0.01	-0.01	0.02
	iShares Russell 1000 Value	IWD	-0.12	0.24	0.00	-0.04
	WisdomTree US MidCap Dividend	DON	0.21	0.29	-0.07	0.25
	WisdomTree US MidCap Earnings	EZM	0.57	-0.01	-0.16	0.18
	Oppenheimer Mid Cap Revenue	RWK	0.56	0.03	-0.15	0.14
	iShares Russell Mid-Cap	IWR	0.27	-0.09	-0.06	0.01
	iShares Russell Mid-Cap Value	IWS	0.24	0.17	-0.05	0.08
	WisdomTree US SmallCap Dividend	DES	0.71	0.56	-0.12	0.40
	WisdomTree US SmallCap Earnings	EES	0.91	0.34	-0.22	0.32
	Oppenheimer Small Cap Revenue	RWJ	0.98	0.33	-0.17	0.37
	PowerShares FTSE RAFI US 1500 Sm-Mid	PRFZ	0.86	0.23	-0.19	0.11
	iShares Russell 2000	IWM	0.84	0.15	0.02	-0.05
	iShares Russell 2000 Value	IWN	0.78	0.49	0.02	0.12
	Data per Portfolio Visualizer using monthly return data from April 2008 through May 2017. Bolded values are statistically significant at the 5% confidence level.

Nearly all of the fundamentally weighted ETFs have value factor loadings that are statistically significant, or most likely not attributable to chance. This indicates that a portion of the fundamentally weighted ETFs' performance can be attributed to exposure to the value factor. The exceptions are

WisdomTree US Earnings 500 ETF ( EPS ),

WisdomTree US MidCap Earnings ETF ( EZM ), and

Oppenheimer Mid Cap Revenue ETF ( RWK ). And all of the fundamentally weighted ETFs, except

WisdomTree US LargeCap Dividend ETF ( DLN ), have statistically significant negative momentum loadings. This is likely due to their contrarian rebalance. Because these strategies are segmented by market capitalization, the small-size factor exposure is not surprising. The large-cap fundamentally weighted ETFs show negative exposure to the small-size exposure, although only EPS and DLN's size exposures are statistically significant. All mid- and small-cap fundamentally and market-cap-weighted strategies show statistically significant positive size factor loadings. Value and small-size loadings contribute to investment return of this sample of fundamentally weighted strategies.

Research Affiliates, pioneers in fundamental indexing, agree with the assertion that fundamental indexing strategies largely benefit from value and small-size factor tilts. Rob Arnott, CEO of Research Affiliates, and colleagues found that many non-market-cap-weighted strategies, including fundamental weighting, outperformed the market-cap-weighted benchmark. In a paper titled The Surprising Alpha from Malkiel's Monkey and Upside-Down Strategies, they show the performance results when they flip the weightings of the fundamentally weighted portfolios around so that the lowest-ranked holdings receive the largest weights. These inverted portfolios also outperformed the market-cap benchmark and, in many cases, the original strategies. The authors argue that the success of both the original strategies and their inverses is attributable to their implicit tilts toward small-cap and value stocks.

Research Affiliates has also argued that fundamental index strategies benefit from their inconsistent value loading. The strategy adds to stocks as they become cheaper to a fundamental metric, rather than bifurcating the market based on value and growth stocks. Research Affiliates goes on to assert that fundamental index strategies increase value loading as value becomes cheaper, and decreases it as value becomes more expensive. Thus, fundamental indexing offers a dynamic value tilt that can benefit investors.

The dynamic value orientation could be worth the higher fee that these funds levy compared with market-cap-weighted value funds. But timing the value factor is notoriously difficult, and the dynamic approach still does not offer a unique exposure. That is part of the reason why the fundamental index funds in this article that Morningstar rates earn Morningstar Analyst Ratings of Bronze, one peg lower than the cheapest and best run market-cap-weighted value funds, which offer more consistent value exposure at a fraction of the price of the fundamentally weighted options.