Nearly half of global asset owners are now investing in smart beta , with interest continuing to rise year-over-year. The term “smart beta” encompasses an increasing variety of strategies, with many new smart beta products being launched each month around the world. Smart beta products typically aim to replicate indexes that embed the strategies’ underlying methodologies.

Although some newer index approaches rely on detailed quantitative modelling, smart beta indexing doesn’t have to be associated with complexity. In this Insights, we focus on some smart beta index construction approaches that follow relatively simple, intuitive weighting schemes. And, regardless of their methodology, all FTSE Russell smart beta indexes follow transparent, consistent rules in order to achieve the stated index objectives.

Smart beta in context

An increasing number of investors worldwide are using index-based approaches to construct and manage their portfolios. The objective of a manager of an index-based (or “passive” ) investment portfolio is to replicate the index’s return, before fees and costs.

According to a 2017 report by Moody’s Investor Services, passive investments now account for US$6 trillion of assets globally and 29 percent of assets under management (AUM) in the US. Moody’s predicts that the continuing adoption of index-based investment products will lead to passive funds’ market shares in the US exceeding 50 percent by early in the next decade.

Historically, the standard approach for passive investing has been to track a capitalization-weighted index in which the index’s allocation to each constituent is determined by the constituent’s market value (its number of shares outstanding multiplied by its share price).

In the context of indexing, smart beta is a generic term for indexes that depart from the standard market capitalization weighting method in order to achieve particular objectives, such as the generation of excess index returns, the mitigation of volatility, or diversification.

But despite their change in weighting approach, smart beta indexes share important characteristics with their capitalization-weighted counterparts:

  • They share the principle of being constructed according to a consistent set of rules, rather than relying on discretionary decisions.
  • They are transparent, so that any market participant can look up and familiarize themselves with the details of the index’s construction.

A governance infrastructure also underlies any index, whether capitalization-weighted or smart beta. At FTSE Russell, the index ground rules set out the management responsibilities of the entities involved in the oversight of the index series: who acts as the benchmark administrator and the role of external advisory committees including the FTSE Russell Policy Advisory Board. Separately, the FTSE Russell Governance Board is responsible for ensuring that all the FTSE Russell indexes meet appropriate technical standards.

When choosing between standard, capitalization-weighted indexes and smart beta indexes as the basis of an investment product, market participants often bear in mind questions of simplicity and transparency, cost, liquidity, capacity and governance. The main attributes of the two types of index are set out in the table below:

Examples of smart beta index construction approaches

Although the complexity of design of smart beta indexes can vary, several indexes share the simplicity of capitalization-weighting. One category of smart beta indexes includes those constructed according to intuitive, common-sense (often called “heuristic”) rules. Among the most straightforward smart beta approaches are equal-weighting and weighting by fundamental accounting measures.

Example 1: Equal weight

Equal weight indexes aim to increase diversification, alleviating the concentration concerns inherent in market capitalization-weighted indexes.

In a capitalization-weighted index, constituents are weighted by the product of their share price and their number of outstanding shares (i.e. by their market value). By contrast, in equal weight indexes constituents in the underlying parent index receive an equal weight.

Due to stock market movements, the weights of individual constituents in equal weight indexes can become unequal over time. For FTSE Russell indexes, eligible securities are reviewed annually and rebalanced to equal weight on a quarterly basis.

As an approach, equal weighting helps address concerns about potential concentration risks in capitalization-weighted indexes. For example, during the internet bubble of 1999-2000, stocks from the technology sector gained particular prominence in capitalization-weighted indexes (see Figure 1, where the Russell 1000® Index of US large-cap stocks is used as an example).

Such periods of concentration may have a significant effect on the index performance if the sector concerned then suffers a period of relative underperformance.

At the constituent level, the equal weight methodology can result in a significant underweight position of a few large-cap stocks in comparison with the capitalization-weighted index.

For example, as can be seen in Figure 2, where the Russell 1000 Index’s Financial Services sector is used as an illustration, the Financial Services sector stocks with the highest market value, such as JP Morgan, Citigroup, US Banc Corp, American Express and BNY Mellon, receive a much smaller constituent weighting (a relative “underweight”) in an equal-weighted index than in the Russell 1000 Index. These underweighted stocks are shown at the left side of the chart.

At the same time, a long “tail” of small-cap stocks within the sector, visible on the right side of the chart, receives a small weighting boost (and an “overweight” position relative to the capitalization-weighted index) under the equal-weighted approach.

Given that an equal weight index overweights small-cap stocks by comparison with its capitalization-weighted counterpart, index designers have to bear in mind the capacity and liquidity of the resulting index in order to ensure its suitability for practical use as an underlying benchmark for index replicating financial products such as ETFs.
Index designers can address potential liquidity and capacity concerns in an equal weight index by restricting index eligibility to stocks with an adequate float-adjusted market value and to stocks whose average daily trading volume (ADTV) is above a certain threshold.

Equal weighting is not a new index approach. In fact, one of the first commercially available index funds, launched in the 1970s by Wells Fargo, was based on an equal weight index of shares on the New York Stock Exchange (NYSE). The fund was wound up because it suffered from excessive transaction costs–at the time, transactions in US shares were subject to high minimum commissions.

In more recent times, dramatic reductions in share dealing costs have made equal-weighting a viable proposition and the investment strategy is enjoying a resurgence of interest: roughly half of US financial advisors recently surveyed by FTSE Russell said they either use or are very likely to use an equal-weighted investment approach.

Example 2: Fundamental indexes

A second example of an intuitive smart beta index approach is fundamental indexation. Fundamental indexes, like equal weight indexes, determine index constituents’ weights by some other measure than their stock price. In general, fundamental indexes measure company size using fundamental measures such as sales, operating cash flow, dividends and book value.

Research Affiliates (RAFI), which popularized the strategy from 2005, notes that fundamental indexation also follows a contrarian rebalancing approach, systematically trading out of constituents whose prices have increased and increasing the allocation to securities whose prices have decreased.

Applying the fundamental index approach

The Russell RAFI Index Series ranks and weights companies by averaging three fundamental factors:

  1. Adjusted sales–constituent sales are averaged over five years and then adjusted to take into account financial leverage, decreasing the index weight of companies with significant leverage
  2. Retained operating cash flow–the five-year average cash flow from operations, less dividends and buybacks
  3. Dividends and buybacks–the five-year average of dividends paid and share buybacks

Empirical evidence shows that fundamental measures of companies’ size are relatively highly correlated with the companies’ market capitalization–in other words, a stock that scores highly on fundamental measures tends also to have a high market value. In practical terms, this means that capacity and liquidity constraints in fundamental indexes may be less of a concern than for other types of smart beta indexes.

Smart beta performance in context

Investors selecting smart beta indexes typically do so for a variety of reasons, such as the generation of excess index returns, the mitigation of volatility or diversification. The resulting performance of smart beta must therefore be seen in the context of the original objectives.

At a headline level, the total return of both a fundamental and an equal weight version of the Russell 1000 Index exceeded that of the capitalization-weighted reference index over the period from December 1999 to May 2017 (see Figure 3). However, over the most recent 1-, 3-, 5- and 7-year periods, the capitalization-weighted Russell 1000 Index provided a slightly higher per annum total return (see Figure 4).

The divergence in performance from the capitalization-weighted Russell 1000 Index is paralleled by a divergence in the smart beta indexes’ risk statistics from those of the reference index (see Figure 5). Both in terms of standard deviation of return and in terms of tracking error, the Russell 1000 Equal Weight Index showed a greater divergence from the reference Russell 1000 Index than the Russell RAFI US Index.

The deviation of the smart beta indexes’ return and risk measures from those of the reference index should be considered in terms of the stated objectives of the two different smart beta approaches.

For the Russell RAFI US Index, whose objective is to select securities based on fundamental measures, rather than stock price, the concentration level in the top ten securities is similar to that of the capitalization-weighted reference index–in this case, the Russell 3000® Index (see Figure 6).

At the same date, the Russell RAFI US Index also exhibited an improvement in fundamental scores vis-à-vis its capitalization-weighted counterpart, the Russell 1000 Index (see Figure 7). Although it had a higher price-earnings ratio, the Russell RAFI US Index had lower price-to-sales and price-to-book ratios than the reference index, as well as a higher dividend yield.

For the Russell 1000 Equal Weight Index, the effective number of securities (Effective N) in Figure 6 is markedly higher than that of the reference index, the Russell 1000, while the concentration in the top ten holdings is markedly lower than that of the reference index. These statistics, together with the relatively high active share, show that the index is meeting its primary objectives of reducing concentration and increasing diversification.


Smart beta doesn’t have to be complex. While some non-traditional index approaches rely on more involved quantitative modelling, some popular smart beta index construction approaches follow simple, intuitive methodologies. Equal-weighted and fundamental indexes are two such approaches: they use a straightforward construction methodology in order to address potential concentration or valuation concerns inherent in capitalization-weighted indexes.

When choosing between standard, capitalization-weighted indexes and smart beta indexes, market participants often bear in mind questions of simplicity and transparency, cost, liquidity, capacity and governance. Index users should also evaluate smart beta in the context of the original objectives, such as the generation of excess index returns, the mitigation of volatility, or diversification. This information is as readily available as it is for capitalization-weighted indexes and should not be an obstacle to considering smart beta alongside more traditional index options.

--Download paper--