ETF Complete Strategy Insights: How to Follow the ETF Models (Part 4)

James Kimball | June 16, 2019

The ETF Complete model closed the week up +0.4% compared to the SPY which closed up +0.6%.

Coming off the heels of a large move the prior week and gap higher to start the week, markets largely went sideways and closed lower than it opened the week. The SPY is only a few percent off its all-time high.

Stay tuned for the daily updates and log into the website to see holdings and additional performance data.


This Week's Strategy Lesson: How to Follow the ETF Models (Part 4)


We have spent the last few weeks going through how to start a portfolio and various issues related to following along with the models. Last week we did an extensive walk-through of all the decisions, calculations, and actions you would have needed to make during a 2-month period following the models. In the last two parts of this series, we will look at some considerations regarding timing the models.

Its always easy to look at a stock chart and imagine you would have been able to recognize at the time that that low would actually hold or that peak high was the absolute best time to sell. In real life, its not so easy. There are guides and strategies we can use to help us draw conclusions, but we never know for certain what the future holds.


The chart above shows the SPY from early 2007 through 2010. The red lines show sample negative 12-month return periods while the green lines show some sample positive 12-month return periods. There was an 18-month period starting in March 2007 through late 2008 where it didn’t matter when you got in or at what price, your 12-month holding period return for the SPY would be negative.

At some point late in 2008, that flipped due to the dramatic reversal off the March 2009 lows. Although a position entered in November 2008 would have been positive if held through November 2009, you would still have had to sit through a big decline of almost 30% before it reversed (which shows some of the limitations of this statistic). Of course, most periods in our model samples weren’t quite as extreme as 2008, but the example above easily demonstrates the calculation.


The table above shows the odds of achieving a positive return over any 6 and 12 month time periods if you started an ETF portfolio or bought the SPY during any random week between 2007 and the recent 12-month periods available. The data doesn’t get into the magnitude of the gain or losses over the period or what kind of drawdown you might have had to sit through, which is important, rather it is merely asking the odds of a positive or negative return over the holding period.

Historically speaking, if you picked any week at random during the past 10 years or so, you would have had a 74% chance of a positive return 6 months out and an 87% chance of a positive return 12 months out trading the ETF Complete model. This is higher than the 66% and 74% chances respectively for holding the SPY over 6 and 12 month periods (note that the SPY is also greater than 50% showing that there is also some positive bias in the broader market over the last 10 years, though the ETF models all have a stronger bias than the market).

The high odds of a positive outcome when randomly choosing when to start a portfolio should be reassuring. Profitable trading of the ETF models doesn’t rely on perfect market timing or good luck. But there are still more layers. What if the model just put in a new high or is coming off a period of weakness? Next week we will wrap up this series looking at the performance differences coming off these levels.