Quantitative Analysis Service, Inc.

ETF Strategist - Smart and Smarter
By: Ted Theodore | 07 Jul 2015



Smart and Smarter


While there is general research that is used to show a factor in good light, the facts are that there are better and worse times to be exposed to the factor.


In the early years of ETF growth (post 1999), there were not too many variations from a simple theme.  The first equity ETFs were just replications of large index baskets, particularly the S&P 500. Those large indices were, of course, calculated by just adding up the market value of each stock.  Thus the organization of them was “market weighted.”  And soon, those baskets were sliced into their component sectors, like energy or finance and they were also market weighted. 

Recently, the financial press has been crowded with articles about “smart beta” ETFs.  As with any product, the vendors seek to distinguish their product from all others.  The term smart beta implicitly wants to make you think there are “dumb beta” ETFs.  We certainly do not claim to judge the ongoing marketing rush, but following how this story has developed over time, it is probably safe to say that the early ETFs (circa 1999-2005) were typically just simple variants on the same baskets of stocks.Even as ETFs started to divide into more separable products, the market weighting process was still dominant.  Thus, while we got a separate ETF for the Russell 2000 small caps basket, it was market weighted.  And when the S&P 500 was segmented intoits sector constituents, like energy and financials, those too were cap weighted.

Just reflecting this weighting process, stocks that have performed the best most recently will get relatively higher weight in the basket and stocks that have performed the poorest will have relatively less representation.  But in standard finance theory, the market is assumed to be “efficient,” and stock prices ALREADY reflect recent good and bad fundamentals.  So market weighting in this theory results in buying high and selling low - adding weight to stocks that are already up and  taking weight away from stocks that are already down.

Strictly speaking, then, “smart” weighting is something which tries to avoid “buying high.”  The first such ETFs substituted, for example, earnings rather than market capitalization as a way to weight the securities.  So take two stocks with the same amount of total earnings.  In the “dumb” weighting, those earnings do not matter and the company with the higher market capitalization of the two will get weighted relatively more heavily.  Thus the second company will have a lower price to earnings ratio than the “overweighted” company and it is thus presumably “cheaper” and should perform better than the more “expensive” stock.  Soon after the earnings weighting approach there followed an approach that weighted stocks in the index by their revenues.  More recently there has been a raft of ETF product that weights the constituents by a factor believed to promise performance that is above “average” (the popular benchmarks like the S&P 500 are considered “the average” and they are usually are organized by market capitalization).  These factors include, for example, overweighting companies with growing earnings, or growing dividends or relatively high dividend yields.  There are factors that overweight “high quality” (and “low quality” too!), low volatility, high momentum and on almost endlessly.

While there is general research that is used to show a factor in good light, the facts are that there are better and worse times to be exposed to the factor.  In other words, for example, cheap stocks do not always outperform the averages over time horizons other than assumed in the research.

Our ETF Strategist Tool uses a quantitative process to assess the attractiveness at any one moment for over 800 ETFs.  Among the ETFs we rate, a good portion of them can be classified as in the “smart ETF” categories (http://www.qas-service.com/etf.html).

Again, while the relative attractiveness of a factor changes over time, it is very interesting that currently we see very little difference in the quantitative rankings between a whole group of smart ETFs and capitalization weighted ETFs.  The factors which just show average attractiveness are dividend yield, dividend growth, value, earnings growth and low volatility.  (The exception currently is that some of the US ETFs that are promising above average performance are capitalization weighted, but in their case, the category is “small cap.”)  We think it is important to keep monitoring these distinctions as they can often help judge the overall market character.  One way to think of the environment currently is that investors have very low conviction.  That is another way of saying “the market could go either way.”  While that is literally always true, this time there is a substantial amount of quantitative backing for the notion. 


Copyright ©2017 Quantitative Analysis Service, Inc. For confidential use only. Not for circulation, distribution, or publication, except pursuant to license from Quantitative Analysis Service, Inc., the proprietor of all rights herein. All rights reserved. No part hereof may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the express written consent of Quantitative Analysis Service, Inc.

ETF Strategist - Smart and Smarter
Copyright 2015 Quantitative Analysis Service. All Rights Reserved