Destroying Smart Beta 2: Ground Rules

April 11, 2014

How is defining smart beta tricky? Let us count the ways.

This blog is the second installment of a series transforming our ideas about smart beta. Part I set the stage, arguing that defining smart beta in an ETF context is essentially impossible.


In my previous blog, “Destroying Smart Beta: Part I,” I challenged readers’ perceptions about what defines smart-beta funds.

By taking a broad look at the way indexers solve problems and adapt to their universe, I was able to show that many of the features of smart-beta funds show up widely, while some smart funds don’t fit preconceptions.

I promised to demolish your hopes of defining smart beta.

Today the work starts in earnest, beginning with the ground rules. We need to understand what a successful definition of smart beta would look like, and what the possible candidates are. We can even test a few of the definition candidates.

Like any type of ETF classification, a successful definition of smart-beta ETFs must satisfy basic ground rules. It must:

  1. Apply to all funds consistently
  2. Classify funds according to how they are constructed, rather than by their names or marketing: Many investors incorrectly believe that a fund’s name tells the whole story. They fail to understand that a fund like the IQ Global Resources ETF (GRES | D-44) focuses on momentum in selecting its constituents, or that the Market Vectors Africa ETF (AFK | D-28) selects and weights securities based on country GDP rather than market cap.
  3. Make meaningful groupings: Criteria that create overly broad groups, perhaps including 95 percent or more of US ETFs, will not help us separate out the funds we want to discuss.
  4. Produce results that are widely acceptable to the ETF community: The S&P 1500 is a curated index shaped by profitability screens and an selection committee. Its smallest security, AutoNation, ranked No. 954 in the Russell 1000 as of April 7, 2014. In the U.S., for all but a few index nerds, the S&P 500 Index represents the stock market. In other words, any definition of smart beta must make sense to the people most likely to use it.

These ground rules will allow us to test smart-beta definitions. Any criterion that can satisfy all four ground rules is a winner—a bona fide working definition of smart beta.

Conversely, if a criterion fails any of them, it’s not workable.

I am not alone in my pursuit to define smart beta. Fund sponsors, consultants, journalists and marketers have staked out plenty of criteria they claim define smart beta.

At the end of this blog, I’ve included links of smart-beta definitions from the Financial Times, Cogent, Towers Watson, Russell Investments, Morningstar and Vanguard.

All these experts converge on the following seven smart-beta criteria:

  1. Transparency
  2. Rules-based/quantitative
  3. Thematic/specific segments or objectives
  4. Noncap-weighting
  5. Captures risk premia/factor exposure
  6. Superior risk-adjusted returns
  7. Improves portfolio diversification

I’ll examine these criteria one by one.

Today I’ll take on criteria 1-3, and next time I’ll examine “noncap-weighting” as a criterion for defining smart beta. That’s where folks tend to take dangerous short cuts, so I’ll need an entire blog to set the record straight.

From there, I’ll turn my attention to “risk premia/factor exposure”—also in a stand-alone blog. Then I’ll address the final two criteria, “risk-adjusted returns” and “diversification,” together in the fourth blog of this series.


Find your next ETF