In creating an index, the traditional approach has been to weight a component according to market capitalization. But more recently, companies have applied the term “smart beta” to an index that is created using alternate approaches.

Among the “smart” (or, in some cases, “strategic”) beta products are equally weighted indexes, fundamentally weighted indexes (which focus on a particular financial aspect of a security, such as its dividends or earnings), factor-weighted indexes (such as value or momentum) and volatility-weighted indexes.

The big question with any new approach or idea is whether it performs better than the traditional approach.

But what is the accepted benchmark?

When it comes to index construction, the benchmark is the traditional approach — so that points to an index weighted by market cap. From that starting point, we will be able to establish if the smart beta indexes (or rather, smart beta funds that mimic them) have added any value above a market-capitalization approach over the past three- and five-year periods.


To conduct this test, I created four portfolios based on a broadly diversified model of 12 asset classes, rebalanced annually.

The first scenario is the multi-asset model built with index-based ETFs weighted by market capitalization. This scenario might be thought of as a “raw beta” model — and represents the benchmark from which added value can be measured.

See the “12-Asset Model Portfolio” chart below for the portfolio components. Note that I used well-known and broadly used ETFs for this scenario: SPDR S&P 500 Index, SPDR S&P MidCap 400 ETF, iShares MSCI EAFE and the like.

The second scenario is a multi-asset model that uses actively managed funds. There were several choices of fund families for this; ultimately, I chose T. Rowe Price. This represents what some might call an alpha-seeking approach, inasmuch as the T. Rowe Price funds are actively managed.

The last two scenarios involve the use of funds from families that embrace smart beta approaches: WisdomTree and PowerShares.

WisdomTree uses a fundamentally weighted approach that focuses on dividend weighting in creating its smart beta products. PowerShares uses both single-factor strategies (like low volatility or high beta) and multiple factor strategies (such as size and style) in its smart beta product construction.

Many of the PowerShares ETFs are based on Research Affiliates fundamentally weighted indexes, where the weighting is based on such factors as sales, dividends, cash flow or book value rather than on market cap.

One note about the smart beta scenarios. The multi-asset model uses 12 funds, weighted equally — but because neither WisdomTree nor PowerShares have smart beta products for all 12 assets, I subbed market-cap weighted ETFs from the benchmark scenario wherever a smart beta version was missing.

In the WisdomTree version of the model, I used WisdomTree smart beta funds for the six categories in which a smart beta version was available: large-cap U.S. stocks (ticker DLN), mid-cap U.S. stocks (DON), small-cap U.S. stocks (DES), non-U.S. stocks (DWM), emerging-market stocks (DEM) and non-U.S. real estate (DRW), which is the only smart beta REIT fund WisdomTree offers. The remaining six categories — natural resources, commodities, U.S. bonds, TIPS, non-U.S. bonds and cash — were populated with traditional index ETFs from the benchmark scenario.

The PowerShares model, meanwhile, also used six smart beta (or “intelligent beta,” as PowerShares refers to it) ETFs: for large-cap U.S. stocks (PXLC), mid-cap U.S. stocks (PXMC), small-cap U.S. stocks (PRFZ), non-U.S. stocks (PID), emerging stocks (PXH) and natural resources (PXI). Again, I used the same ETFs as in the passive index scenario for the remaining six categories: real estate, commodities, U.S. bonds, TIPS, non-U.S. bonds and cash.


A performance summary of the four scenarios is shown in the “Battle of the Betas” chart below; the analysis fully accounts for expense ratios.

First, the benchmark: The passive index model using market-cap-weighted ETFs generated a three-year annualized return of 6.33% for 2011-2013 and a five-year annualized return of 11.45% for 2009-2013.

The second model, the alpha-seeking option using actively managed T. Rowe Price funds, had a three-year annualized return of 6.41% — nearly identical to the passive benchmark — but turned up a five-year return of 12.81%: a 136-basis-point performance advantage over the past five years.

The smart beta models were next. The portfolio using WisdomTree ETFs generated a 6.4% three-year annualized return and a 10.89% five-year return. Again, this performance was very similar to the passive index model over the past three years, but with a slight underperformance over the past five years — and a significant underperformance (of nearly 200 bps) compared with the T. Rowe Price model.

The best performer over the past three years, it turns out, was the smart beta portfolio using PowerShares ETFs, with a 7.23% annualized return. That portfolio also generated a 12.53% annualized return over the past five years — trailing the T. Rowe Price model but better than both the benchmark passive model and the WisdomTree portfolio.

In general, outperformance tends to come at a price: Note the higher volatility indicated by the large standard deviation of return for the alpha-seeking model. Yet it should be noted that the higher standard deviation was primarily caused by the portfolio’s larger positive return in 2009. Higher volatility on the upside is unlikely to bother investors.


The performance differences among individual funds — based on whether they are a raw beta design, a smart beta design or an alpha-seeking design — may be significant from year to year. But when the three types of funds are considered within the context of a broadly diversified 12-asset model, much of the variation between the various methodologies can be balanced out.

One beta approach will be superior during one time frame and then another approach will win during a different time frame. It is exceptionally difficult, approaching impossible, to anticipate which approach will be superior going forward.

And one specific concern about smart beta funds — even acknowledging the investing truism that past performance does not guarantee future returns — is that many do not have a long enough track record to even offer a good indication of how they’d fare in differing market conditions.

For investors who review and measure each component in their portfolio as a separate stand-alone element, the performance variation between raw beta and smart beta funds will be noticeable at the individual level.

Likewise, clients who do not choose a broadly diversified portfolio will be more sensitive to the performance variation among investment products; their return is solely dependent on a small number of funds.

But for clients who recognize the value of diversification, the returns of the various asset classes within their overall portfolio will be less relevant. Their focus is on the overall return of the portfolio rather than the performance of the separate elements.

Said differently: If an overall portfolio has acceptable performance, your clients won’t really care if it was raw beta, smart beta or alpha-seeking ingredients that produced the outcome.


As always when building a portfolio, a good allocation model can balance out weaknesses in any individual portfolio element. That goes not only for the asset class’ model, but also the fund components the advisor uses to structure the mix.

When we match up smart beta funds against other passive or active funds, we will likely see significant performance variation at a fund-to-fund level. But, at the broad portfolio level, the differences between portfolios will likely be more muted — simply because we are combining moving parts that will take turns adding or subtracting value from the overall portfolio.

In the end, the focus should be on building smart portfolios. And while ideas about intelligent portfolio design may vary based on a client’s needs and goals, smart portfolios will have several common characteristics: low cost, diversification across a number of asset classes, an overall equity-to-fixed income split that is risk appropriate for the specific investor, a resistance to overmanagement and a logical basis that is straightforward and easy to understand.

Smart portfolios will also be adjustable — adaptable enough to accommodate the needs and constraints of different investors. Building smart portfolios is the real goal, and they can be built with raw beta, alpha-seeking or smart beta ingredients.

Craig L. Israelsen, a Financial Planning contributing writer in Springville, Utah, is an executive in residence in the personal financial planning program at Utah Valley University’s Woodbury School of Business. He is also the developer of the 7Twelve portfolio.

Read more: