The Basic model had a solid week, beating out the benchmark by a small margin. Meanwhile, the Stops & Targets model, with only one position and that one selling off considerably had a very poor week down just over 4%.
The SPY closed positive thanks to strong gains on Monday and Friday that bookended a relatively quiet week. Earnings season has officially kicked off. Expectations are very low, which doesn’t speak well to the state of the economy, but low expectations can often lead to positive upside surprises.
DRN had a really rough week, tumbling over 8% and dropping into 4th place. It is near the stop in the Stops & Targets model and we could see some position changes next week regardless of if it hits that stop.
Stay tuned to the daily updates should they occur.
This Week’s Strategy Lesson: Relative Performance (Part 1)
We decided to change gears a little and focus on relative performance for a few weeks. The Sector models (particularly the Stops & Targets) are both negative on the year and continue to lag the SPY benchmark. This is a good time review the performance and how it might be similar or different to what we can expect based on the historical performance.
There are a lot of different components and ways to analyze relative performance: Annual returns, rolling period returns, drawdowns, etc... We would like to cover a number of these metrics, but we will start with one method for comparing relative performance: Ratio charts.
A ratio chart simply takes two financial instruments and divides one against the other for each period and then plots this value on a chart. Whenever the line is increasing, that means that the first instrument is temporarily outperforming the second and whenever that line is decreasing, the first instrument is temporarily underperforming the second instrument.
Furthermore, if the line is above the starting point, that means the first instrument has outperformed the second over the whole time period being evaluated.
A ratio chart doesn’t tell us anything about the absolute performance of each instrument, only about the relative performance. Meaning, over the period above, both might be going down, however, the first one is going down slower than the second (outperforming on a relative basis).
Sector Stops & Targets
For this analysis, we wanted to narrow down to the Sector S&T model. We could have just as easily used Sector Basic, or one of the Country models or the Complete. The analysis would be the same, but the relative results would all fall along a spectrum.
This is a broad ratio chart of the Sector S&T model divided by the SPY over the last 8+ years. There are a few initial observations we can make. The chart moves from 1 on the left to around 5 on the right. This means that the Sector S&T model outperformed the SPY by 5 times over this period. On an absolute basis, it was up closer to 600%, but once you factor that the SPY was up around 45% over this period, that accounts for the difference.
Also, ratio charts can be deceptive sometimes. While if you look at this chart from the beginning of 2014 to now, it might look like the model might be right back to where it was when it started that period (it is on a relative performance basis to the SPY), it is in fact up more than 10% over that period.
Looking back through the whole history of the chart, we can see a number of periods where outperformance increased, decreased and chopped sideways (slope of blue line). Let’s zoom in on one of these periods and see what there is to discover about how the model performance can vary compared to the benchmark. Looking at the 18-month window from September 2011 to the end of April 2013:
When we zoom in on this period we can really see the volatility and relative trading characteristics of the Sector S&T and SPY. Over the entire period, the ending performance of both models were the same (chosen on purpose), however there was a lot of relative changes throughout this period.
The SPY (orange line) was up 35%, but this included a number of large drawdowns (from peaks) and a 6-8 month period where it was basically flat. The Sector S&T (blue line) also ended up 35%, but it had significantly higher volatility and even spent some considerable time negative in both absolute and relative performance. At its peak, the Sector S&T model lagged by as much as 23% (Green line on 11/23/12) although the maximum drawdown in this period in absolute terms was only 10%.
The swiftness with which the model can claw back gains is perhaps equally as striking as the speed of its drawdowns. This high volatility but strong performance is why we emphasize position sizing to comfortably handle the drawdowns so that you are “in” when the model goes on a good run.
The recent volatility and swings in the model in the last couple months fits this mold well. Both Sector models went from decent gains for the year to a loss and lag behind the SPY (which is currently just barely up on the year).
However, this type of behavior is not unusual. There will be periods where the model will either lag in relative performance to the SPY or in absolute performance for the year. So far the model has always been able to recover the lead (eventually), and based on historical evidence, there is no reason to believe that this time will be different. No one can foresee the future, but we still strongly believe the models provide a valuable “edge” in trading.
Next week we will continue this series looking at ratio charts, rolling-period returns and drawdowns as we examine the different elements of monitoring relative performance.
The Current Condition of the Model
Our three positions are CURE, DRN, and SOXL. The Stops & Targets has only one position DRN. DRN had a really rough week, tumbling over 8% and dropping into 4th place. It is near the stop in the Stops & Targets model and we could see some position changes next week regardless of if it hits that stop..
Stay tuned to the daily emails for any position changes and updates.
Here is a summary of the weekly performance of all the ETFs that the strategy monitors:
Best wishes for your trading,
James Kimball
