ETF Movement as Research For Understanding Bigger Themes

Feb 03, 2011

"I measure what’s going on, and I adapt to it.” - Martin Zweig

ETFs represent well-defined baskets of securities --- indexes --- and we can use these index movements to better understand money flows and trends. The following list is a blend of ETFs that was based on the most popular securities at ETFreplay. This list of 15 ETFs represent a good sampling of bonds, stocks, commodities etc... Some editing to avoid duplicate indexes from different providers was used.

First lets take a look at what happened if we just held all 15 in equal-weight for the entire year 2010 using the Backtest Portfolio Allocations App (you can hand-input up to 10 symbols in the free version of the page).

Auto-Load Symbols From An Existing Portfolio:


Press 'Test Strategy' Button on Right To Run Chart Of Portfolio:

Scan Supporting Statistics:

Note that the correlation of this equal-weight portfolio is 0.97 to the S&P 500 (see SPY correlation figure in the top-middle of the 3rd chart image above) . This reading actually isn't that big of a surprise. Most buy & hold ETF allocations that contain a large allocation to equities will result in high correlation to an equity index like the S&P 500.   

Now let's take a look a little deeper at how one particular tactical asset allocation (TAA) strategy found leadership in 2010. We will study 2010 by creating a model and then observing which ETFs were held the most during the year within that model. Whatever was held most might be considered a 'theme' for the time period under discussion.

What relative strength does is tries to understand what is strongest and hopefully thereby find ETFs that are beginning to trend. Every leading ETF starts out with relative strength and then either trends or mean-reverts. Judgment based on experience will be necessary ultimately -- but having a quantitative process locate the strongest ETFs will give you thematic ideas as they develop to consider.

We used a monthly rebalance schedule and a basic 3-factor model to choose among the 15 ETFs. We set the backtest to include the top 3 at the end of each month and hold until they fall out of the top 3, at which point they are replaced with the new top 3.

The next table summarizes which ETFs were held using this strategy for 2010:

As this model correctly identified at the time, U.S. Tech stocks (QQQQ) U.S. Midcaps (MDY) were themes. Bonds had strong relative strength bull move into the summer and Treasuries (IEF) and U.S Corporate Bonds (LQD) both show significant days of being held. Using this strategy, Europe (VGK) was NOT held for even a single day during 2010. Below is a chart of a comparison of these 3 ETFs (QQQQ, MDY, IEF). You can easily visualize what the model was picking up.

Forwarding to more recent action, emerging markets have moved down the relative strength rankings while agricultural commodities remain highly ranked. Underweighting emerging markets equities and overweighting agriculture have been recent themes. Will this continue? It is to be determined --- but this is where a quantitative backtest would say to be --- but there will be future rotations in 2011 so we will just have to continue to...measure what is going on and adapt to it.




Follow us on Twitter:




Comments are closed