This article is part of a regular series of thought leadership pieces from some of the more influential ETF strategists in the money management industry. Today's article is by John Eckstein, chief investment officer and director of research at Astor Investment Management.
As I threatened to do in my last column published on ETF.com, I return with a second round of smart beta analysis. Last time I focused on dividend strategies; this time I’ll look at nondividend strategies. I found the best-performing fund surprising, and I have a story that might explain why it has done the best.
Recalling The Lessons On Backtests
First, a quick refresher on what I did last time. I chose to use the funds’ backtests from before their launches. A backtest is produced with the advantage of history shaping the index rules. With that in mind, if a “smart beta” fund were to launch today, you can bet it would have a simulated history.
The history is a careful application of the indexes' rules to the past, but in looking at index returns before and after they go live, I found a statistically significant relationship in the direction of increased beta in real time and a statistically marginal relationship between alpha before and after launch.
In other words, the typical fund in my analysis has performance closer to the market with lower idiosyncratic returns than in the simulation.
But even considering that, I’ll still look at the longer history that the backtests offer because they have more than double the amount of data we can study. Also, using backtests allows us to look at performance around the financial crisis.
As I did last time, I’m going to define “smart beta” by taking both words seriously.
Defining Smart Beta
I take "beta" to mean that the strategy has to be a modification of some major index. The "smart" in “smart beta” means abandoning the cap-weighted standard and finding some other sort of way to weigh constituents.
The new weighting techniques I’m considering here are mainly based on one or more factors.
These factors have been identified in the academic literature as being some well-defined characteristic of stocks that can be used to sort stocks from highest to lowest.
An example is value, which may be defined as price-to-book ratio. You could rank every stock from the one with the highest price to book ratio to the lowest and call, say, the 30 percent of stocks with the lowest price-to-book ratio value stocks.
The smart-beta strategies I’m considering here use one or more of these factors to reweight an index or select a subset of stocks.
Strategies, Not Labels
The universe of ETFs I am looking at today are the “smart beta” ETFs that—unlike in my previous article—do not weight by dividends. I choose the biggest ETFs, with at least $200 million in assets and—crucial to my purposes here—have indexes calculated back to at least the year 2000. This gives me 12 funds and their associated indexes to examine.
When I downloaded the data from ETF.com’s ETF Screener & Database, these funds have total assets of about $26 billion. That’s about a third of the assets in the dividend-focused ETFs that I examined in my previous article.
It’ll take some effort to discuss the different philosophies expressed by the various funds presented here.
What I mean to say is that no one should be looking to invest in a “smart beta” per se. That’s because the term encompasses a variety of philosophies, or strategies, really. And in some cases, these various strategies can in some sense cancel each other out.
For example, if you put an equal investment in the 12 funds I examine here, you would end up with S&P 500-like exposure (beta around 0.9) with a slight value exposure of about 15 percent.
|Compound Annual Growth Rate 2001-2014||Annualized Risk-Adjusted Returns||Worst Draw-down||Compound Annual Growth Rate 2010-2014||Annualized Risk- adjusted returns||IndexLive||Weighing Scheme|
|FEX||9%||0.54||-54%||16%||1.04||2007-04||Composite (growth & value)|
|TILT||7%||0.42||-53%||15%||0.96||2011-09||Composite (size & value)|
Nondividend smart-beta ETFs, performance of index tracked by funds, adjusted for fees 2001-2014. Source: Bloomberg, ETF.com, Astor calculations.