Quantitative Analysis Service, Inc.

ETF Strategist - Dig We Must
By: Ted Theodore | 16 Mar 2015

Since the stock market low point six years ago, about $1 trillion more has been invested in US bond mutual funds than in US stock mutual funds.  Yet stocks have about tripled in value during this period.


Active or passive?

This caution among individual investors comes despite all time record levels of GDP, retail sales, employment and profits.  As if to buttress investor caution there have been a growing number of articles and discussions about passive investing.  Those who are looking at the subject are likely aware of the statistics showing that active investment managers have recently substantially underperformed passive investing and index funds.  And more are finding out every day as shown below in the index of the number of searches in Google for “passive and active investing.”

Source: Google® Trends (Data: Total number of searches, normalized)

The extraordinary growth of Exchange Traded Funds has paralleled the growth in passive investing.  Yet while seeming to be a kind of binary decision (active or passive), in reality there is a fair amount of nuance between what seems to be mutually exclusive strategies.

A stock market or a market of stocks?

As it turns out, there are better and worse times to be active or passive.  It can be thought as the difference between “a stock market” and “a market of stocks.”  In the former, big themes and issues are so important to investors that they relegate individual company developments to the back burner.  As these concerns rise, the differences between companies seem to fade.  It is a time when “macro” economic and/or investment themes carry most of the weight.   The result is that individual stocks move more and more in lock step with each other as correlations between stock movements rise in “a stock market.”

The opposite series of developments take place when the response to macro issues fade.  The specifics of a company and its stock gradually take on more importance.  The “systemic” influences become less important and, as a result, the “specific” differences between stocks come to the fore.  The result is that correlations between stock movements decline in “a market of stocks.”

The simple arithmetic of the exercise suggests there is typically more opportunity to select above-average performing stocks when macro risks are low.  By the same token, there is less likelihood to be able to pick out the winners when correlations are high.

Knowing when

There is a fairly wide range of ways to characterize these market environments.  The most popular way to measure the environment is to correlate the percentage price change of each stock in an index to every other member of the index - separately.  For the S&P 500, that is almost 125,000 “pairs.”  For the market as a whole that is more than 6 million pairs.The degree of correlation in the market at any one time is the average correlation for all the pairs.  So the barriers to observing the phenomenon are pretty substantial.  The literature on the subject suggests that most of the differences in the particular methods are not crucial to characterizing the overall environment between a stock market and a market of stocks.  More or less, the cycle impacts most stocks in a general way.  And, importantly, there are short cuts to measuring this cycle.

The factor behind the short cut lies in understanding the “demographics” of the stock market averages.  Only a small number of the S&P 500 are mega cap stocks.  About 10% of the stocks in the index account for almost 50% of the total value.  And their typical value is almost 10 times the market value of the other 90%.  So, really, the S&P 500 index, which is a capitalization weighted index, reflects the movement of extremely large companies.  But if the stocks were equally weighted, rather than by capitalization, we would get a sense of what the “typical” stock is doing.

Just as there are S&P 500 “market cap”indexes and ETFs(most popularly either SPY or IVV), there are also indexes and ETFs for versions of an equal weighted S&P 500, particularly RSP (Guggenheim S&P 500 Equal Weight ETF®).

In the first step of the short cut, we recognize that changes in RSP (all stocks are weighted equally) more resemble the changes in the “typical” stock than changes in the mega cap stocks.  But we are not interested here whether small caps are outperforming large caps (that is a separate cycle itself), but whether there is a high degree of correlation between stock movements.

In order to do that we need a second step.  Similar to the index which equal weights stocks, there is an index which equal weights sectors.  Sectors are, of course, groupings of stocks.  There is an energy sector, comprised of all the individual energy stocks.  There is a financial sector, comprised of all the individual finance stocks.  And so on.  As it turns out there is an index that underlies an ETF, EQL (ALPS Equal Sector Weight ETF®), which equal weights the sectors.  Whereas sector capitalization weights have varied from about 3% to over 25%, EQL weights the nine major market sectors equally. 

The individual or specific stock risk within a sector is reduced when you use a “sector-weighted” index.  We can assume that when a “sector-weighted” index that underlies the EQL ETF is outperforming a “stock-weighted” index that underlies the RSP ETF, this would indicate that macro concerns have increased.  In reverse, when the RSP ETF (tracking a “stock-weighted index”) is outperforming the EQL ETF (tracking a “sector-weighted index”), macro concerns are receding.

At this point of the analysis, it becomes important to take into account the market’s well known predilection, as we mentioned above, to favor either large caps or small caps, a phenomenon which can be tracked with “relative strength” analysis.  So, in the third step, to focus as much as possible on just the change in correlations, we should “normalize” the EQL/RSP ratio (represented by the underlying indexes values data) by the capitalization cycle.  We do this by dividing the first ratio by a second ratio which divides a large caps index by a small cap index.  We use the Russell® index family. The result is shown below.

Data Source: Bloomberg® (Indexes)

Going deeper - or not

ETF/indexing investing has become more and more popular among investment professionals compared to individual stock investing. There are many reasons for this phenomenon, including individual stock risks reduction, “quick” macro/theme investing and others. 

The number of ETF products increased dramatically over the past 5-6 years (over 1600 in 2015). Today’s ETF product offerings cover every industrial sector, sub-sector, market segment andasset class across all countries and regions, and therefore the selection process is becoming more complicated.

Reliable ETF selection tools and methods are becoming more and more popular among ETF strategists, portfolio managers and researchers.

Quantitative Analysis Service, Inc. (QAS) ETF Strategist search engine provides comprehensive method to effectively analyze ETFs (For more information visit:


QAS’s score-based “7-Zones” quant-ratings system clearly helps in the selection of ETFs by sorting all ETF in its database into seven risk zones, where Zone 1 represents the lowest risk/high reward and Zone 7 represents the highest risk/low reward opportunity.

In addition, having some sense of whether the stock market is delving into more specific factors analysis, as when correlations are low, should raise the odds on the selection process even further.

As an example (as of early March of 2015), the ETF for the major energy sector, XLE, had our most negative rating (Zone 7).   However the PowerShares WilderHill Progressive Energy ETF®(PUW) had a positive bias (Zone 6+).

Also, the broad sector ETF for healthcare, XLV, had a neutral rating(Zone 6+).  In contrast, IBB, the healthcare sub-sector biotech ETF within healthcare had a more positive rating (Zone 4).  In financial segment, XLF (broad sector ETF) also had a neutral, positive bias rating, while KBWR (PowerShares KBW Regional Banking®) had a much better rating(Zone 3, Positive, Early Stabilization).

Interestingly, it has been announced that fairly soon, MSCI and S&P will be separating Real Estate stocks from the financial sector.  As of early March of 2015, ICF, iShares Cohen & Steers REIT® ETF has a stronger current rating (Zone 2, Positive Maturing) in comparison to XLF (Zone 6+, Neutral, Positive Bias). 

In sum

ETF strategists and investors can take advantage of the current low correlation environment by digging deeper within the ETF offerings.  But if correlations rise, they should also “back out” of the more specific company and industry related factors because those factors will mean less when we are, say, in a recession or a financial crisis.

In other words, we can seek selection picking “alpha” in low correlation markets, but hide in index “beta” when correlations rise.

Finally, if all of today’s sideline cautionary money – which extends to institutions and corporations, as well as individual investors - does eventually find its way into the stock market (as history strongly suggests), it might seem as if all that money will force a period of higher correlations.  But that is unlikely as higher correlations are typically associated with market stress.  Rather, the flows back into stocks will very likely come as a result of a realization that the macro fears are no longer warranted and that getting down into the selectionweeds will be the more appropriate investment strategy.  The chart shows a fair amount of potential for correlations to fall even further.


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 - Dig We Must
Copyright 2015 Quantitative Analysis Service. All Rights Reserved