Smart Beta 7: ETFs From The Inside Out

Dispensing with 'smart beta' leads to a clearer understanding of the ETF landscape.

Director of Research
Reviewed by: Elisabeth Kashner
Edited by: Elisabeth Kashner

Dispensing with 'smart beta' leads to a clearer understanding of the ETF landscape.

This blog is the seventh and final installment of a series transforming our ideas about "smart beta." Part 1 started with the proposition that defining smart beta in an ETF context is essentially impossible. Part 2 laid out the ground rules to prove the point; Part 3 sunk noncap weighting as a method to categorize smart beta; and Part 4 took a wrecking ball to the notion that factor-focused tilts were synonymous with smart beta. Part 5 examined the dearth of real risk-adjusted outperformance in the realm of smart beta. Part 6 introduced a strategy rubric that looks through labels to describe what a fund actually does. This final blog caps the effort by tracking the economic principles behind each fund's index construction.

Before I killed off the term "smart beta" and introduced's strategy description, I was hounded for a count of "smart-beta funds."

That's how I opened this blog series, remember? "Everyone wants facts and figures about so-called smart-beta funds: How many are there? What's the biggest one? What's the total number of assets under management in smart-beta funds? How many launched in 2013?"

I refused to answer these questions, because I couldn't define "smart beta." Then I killed the term "smart beta." Out of its ashes, I proposed that we talk about a fund's strategy—what a fund does, and how it's constructed—rather than about its name or marketing.

Now, when a journalist at the Wall Street Journal calls, stubbornly asking, "What's the biggest 'smart beta' fund?" I can use our new strategy field to answer. recently identified 27 strategies in U.S.-listed ETFs. Strategy tags like "vanilla," "dividends" or "optimized commodity" explain how an index sculpts its opportunity set.

My best answer to the question, "How many 'smart-beta' funds are there in the U.S.?" is that I think investors are better off talking about investment strategies. Which ones seem smart to you, and for which purpose? But deciding which strategies are smart takes some effort.

More importantly, "smart" is very context-sensitive. What's smart for, say, my retired father-in-law might not be so smart for my 22-year-old niece.

My Second-Best Answer

Marketers and the press—the folks who want "smart-beta" statistics—aren't managing portfolios. They want to understand and ride trends, and they need tools for measuring investor behavior. And they ask us all the time—we've gotten a call every day about this last week.

For them, I've sorted the strategies into four groups: "vanilla," "active," "strategic" and "idiosyncratic."

Four Strategy Groups

I formed my groups by classifying the criteria that an index uses to select its investment universe, to choose securities within that universe, and then to weight the chosen securities. I focused on three key questions:


  1. Is the portfolio construction process rules-based? Any active fund, or any index with an activist selection committee like Cohen & Steers', which trolls the REIT market for "experienced management teams and high quality properties," is not rules-based. It's therefore "active."
  2. Does the portfolio reflect the complete, unaltered market segment? If yes, the fund is "vanilla."
  3. Does the index apply economic concepts at the constituent level to select and/or weight its portfolio securities? If yes, the fund is "strategic;" if no, it's "idiosyncratic."

More formally, here are my definitions of the four strategy types, plus a few key statistics:

  • Active: portfolios crafted by humans. This includes funds designated as active by the Securities and Exchange Commission.
  1. Total AUM as of May 15, 2014: $14.9 billion
  2. Fund count as of May 15, 2014: 71 (excluding asset allocation and alternatives)
  3. Example fund: Pimco Total Return ETF (BOND | B)
  • Vanilla: funds/indexes that represent the entire investable opportunity set in a market segment. A vanilla index reflects as broad a pool as possible for a given market segment, weighted at its consensus market valuation.
  1. Total AUM as of May 15, 2014: $1.275 trillion
  2. Fund count as of May 15, 2014: 805
  3. Example fund: iShares MSCI EAFE ETF (EFA | A-91)
  • Strategic: funds/indexes that apply research-based, economic concepts for security analysis to select and/or weight positions from an investable, representative, vanilla security universe. These economically based indexes can use bottom-up techniques, technical factors such as momentum, and macroeconomic analysis to select and weight their constituents.
  1. Total AUM: $353.4 billion
  2. Fund count as of May 15, 2014: 456
  3. Example fund: Vanguard Dividend Appreciation (VIG | A-61)
  • Idiosyncratic: Rules-based portfolios applying selection and/or weighting rules that are unrelated to the economic characteristics of its component securities.
  1. Total AUM: $90.2 billion
  2. Fund count as of May 15, 2014: 129
  3. Example fund: PowerShares QQQ ETF (QQQ | A-49)
  • Not included: asset allocation and alternatives, plus 27 funds with activist index committees. My bad. I couldn't face parsing the strategies for these complex, largely unloved funds.
  1. Total AUM: $13.3 billion
  2. Fund count as of May 15, 2014: 121
  3. Example fund: Guggenheim Multi-Asset Income ETF (CVY)

That was easy. Well, OK, it wasn't that easy.

But it is pretty easy to pick out the active funds and the vanilla indexes.

The Securities and Exchange Commission (SEC) has taken care of the Active ones.

For finding the vanilla funds, I built myself a series of flow charts for using's ETF Classification System's selection and weighting fields to identify vanilla funds in each asset class. Email me at [email protected] if you want a copy.

The hard part is making sense of the nonvanilla 49 percent of the U.S. ETF landscape. As we saw in blog 3 of this series, it's a full-time job for ETF curators to understand, describe and sort these funds.

A little evolutionary explanation will go a long way in helping us sort through the chaos of the 49 percent.


Adaptive Vs. Adoptive

Some indexes are essentially adaptive—they're solving a problem of a particular place, time or social group. Others are adoptive, applying research and economic thinking to value and analyze securities.

The adaptive funds, which I'm calling idiosyncratic, feature either an arbitrary weighting scheme or restricted universes. Neither their selection nor their weighting have to do with their components' economic properties. Some have adapted to harsh conditions including illiquidity, competition among exchanges, inaccessible primary markets and, according to some, evil corporations.

Others idiosyncratics evolved in primitive environments, including the unmeasured market of the 19th century and today's over-the-counter bond market. The equal-weighted funds, adaptations to SEC requirements—or sometimes touted as an antidote to price-bubbles in cap-weighted indexes—fail to use their resources well. They incorporate no information about their component securities.

 Idiosyncratic Funds

In contrast stand the adoptive funds. These funds have their origins not in the wilderness, but in a land of abundance. These indexes are essentially rearranging a vanilla universe according to a theory about what drives returns. They're looking to thrive, not just survive. They seek out and emphasize specific economic properties in their selection universes.

The adoptive funds make up my strategic category. They're my second-best answer to trend-watchers who want to count "smart-beta" funds.

The main categories of strategic funds are bottom-up; technical; and macroeconomic. Bottom-up indexes use income statements and balance-sheet data. Technical or quantitative indexes incorporate patterns of security prices. Macroeconomic index construction involves assessments of the prospects of entire sectors, countries and political systems.

When issuers or the press call, we'll provide this list of strategic fund types:



StrategyClusterExample FundStrategy AUM ($)
as of 5/15/2014
Currency Hedged FundamentalBottom-upDXJ12,020,119,001
Currency Hedged DividendsBottom-upDXJS73,910,070
Low VolatilityTechnicalSPLV1,282,894,074
Optimized commodityTechnicalDBC9,999,195,916
Target DurationTechnicalTDTT2,419,212,750
Volatility hedgedTechnicalSPXH1,503,889,632
High BetaTechnicalSPHB243,405,913
Duration HedgedTechnicalHYHG235,540,370
Total AUM  353,359,219,003

"Smart-beta" seekers will appreciate the elegance of the strategic group. It includes all asset classes. Its focus on security-level economic relevance allows established approaches like value or dividend investing to share the tent with newer methods like low volatility and commodity tenor optimization. It makes nonvanilla selection and/or weighting a necessary, but not sufficient, condition for inclusion.

Nor is it outcome-sensitive. It requires neither risk-adjusted outperformance nor a specific level of factor exposure. Instead, it focuses on construction methodologies. Performance and future factor exposure cannot be analyzed ex-ante, but construction methodology can.

For those who prefer the term "quasi-active," the strategic group's requirement of research-based portfolio formation links the practices of many active managers to passive, rules-based strategies.

Moreover, you can parse the strategic group however you like.

If you only want newfangled strategies, because value and growth are too established to seem "smart" to you, just cut out the dividends, buy-writes, fundamentals, value and growth funds, and you'll have a really cutting-edge list:



StrategyClusterExample FundStrategy AUM ($)
as of 5/15/14
CopycatRestricted UniverseGURU848,967,656
Optimized CommodityTechnicalDBC10,033,009,556
Target DurationTechnicalTDTT2,419,212,750
Volatility HedgedTechnicalVQT1,503,889,632
Duration HedgedTechnicalHYHG235,540,370
Currency Hedged DividendsBottom-upDXJS73,910,070
Currency Hedged FundamentalBottom-upDXJ12,020,119,001
High BetaTechnicalSPHB243,405,913
Low VolatilityTechnicalSPLV11,282,894,074
Total AUM  71,531,038,207

But remember, some well-established strategies are relative newcomers to the ETF space.

You won't find, for example, Rob Arnott's Fundamental funds on the new-school list, because, while his indexing techniques are new, fundamental analysis is as old as the hills. So is dividend investing. Also, "buy-write" strategies have been around for years, even though the first buy-write ETF launched in 2007.

Parse these lists in any way that makes sense to you. I'd love to hear your suggestions for ways to split up the ETF landscape—because the press keeps calling, and I like to have solid, consistent, well-thought-out answers for them.



I bet you have some questions. So did my colleagues. So I'll end this blog with a FAQ section.

Q: Why do you call equal weighting "idiosyncratic," and not "strategic"? Don't all studies of alternative indexing reference equal weighting?

A: Equal weighting is an agnostic position, implying/applying no informational advantage. We're looking for actual security analysis to inform index construction. If there's anything "smart" about our strategic funds, it's the grounding in research.

Also, in practice, more often than not, equal weighting is an adaptation—a way to get around RIC compliance rules. It's a classic adaptation.

Q: Why include macroeconomic factors in the strategic group? They have nothing to do with security valuation.

A: Macroeconomic factors, such as GDP growth or fiscal policy, are well-researched determinants of security returns. While differing from bottom-up factors in the level of specificity, top-down factors can swamp security-specific valuations during periods of political or economic uncertainty.

Q: Why do you consider value and growth to be part of the strategic group? These are well-established style-box funds, not part of the growing "smart indexing" trend.

A: Value and growth are classic, mostly bottom-up security valuation techniques. The most common value and growth metrics—such as internal growth rates, price/book ratios, price/sales ratios, dividend yields and sales growth trends—often serve as inputs in fundamental indexation. (Some growth funds also include momentum or technical indicators.)

In truth, it's sometimes hard to know where value ends and fundamental or dividends begin. There's a case for lumping them all together in "bottom up," but it wouldn't help investors who actually want to find dividend funds.

Q: What about the small-caps? In blog 1 of this series, you walloped us with the proposition that the iShares Russell 2000 (IWM | A-83) accessed the small-cap factor. Shouldn't the strategic funds include the small-caps and other factor funds?

A: In a sense, every fund that's not a total-market fund like the Vanguard Total World Stock fund (VT | B) is slicing, dicing and applying some sort of selection criteria. There's a case to be made that 99 percent of all funds are strategic. But that's not a useful distinction.'s strategic group is meant to isolate funds by construction methodology, not by exposure. Market-cap-selected and -weighted small-cap funds are, by definition, vanilla, because they select and weight constituents based on market cap. If they don't, they're not vanilla; they're a strategic small-cap like the Schwab Fundamental US Small Cap ETF (FNDA | B-84).

Q: Don't the copycat funds select securities from active managers using well-researched techniques?

A: Maybe, but there's no way to know for sure. Who knows what those active managers are up to?

Q: Can committee-selected funds be counted in the strategic group?

A: No, because they're not fully rules-based. These funds do have a strategy assigned to them, but we won't count them in response to any "smart beta" requests or as part of the strategic group. At present, there are only 12 funds that suffer this fate, mostly Rogers Commodity funds. Who could replicate Jimmy Rogers?

Note that committee-selected funds can count as vanilla, so long as the committee doesn't alter the overall representativeness of the index (see the S&P 500 as the best-known example). We admit this is fuzzy territory, and we'll dig in hard on any new committee-based funds we see. Luckily, we don't see much development here.

At the time of publication, the author held long positions in EFA and IWM. Contact Elisabeth Kashner, CFA, at [email protected].


Elisabeth Kashner is FactSet's director of ETF research. She is responsible for the methodology powering FactSet's Analytics system, providing leadership in data quality, investment analysis and ETF classification. Kashner also serves as co-head of the San Francisco chapter of Women in ETFs.