The Stops & Targets model had a very strong week outpacing the benchmark by almost a percent. The basic model lagged behind the benchmark this week as TMF dropped off a little. We rotated out of that position into SSO (2x SPY) on Thursday.
Global equity markets had a strong week with the ACWI putting in a new recent high and the U.S. markets hovering at or near their highs. FXI is near its first target of $52.67.
This Week’s Strategy Lesson: Relative Performance (Part 3)
We are continuing our series on alternative ways of measuring relative performance. In the first part, we focused on ratio charts and how they can be used to easily show the differences in relative performance between two stocks or ETFs. Last week we looked at some rolling period return charts where you can see how the average return for the different instruments can change over time.
This week, we are going to look specifically at drawdown charts. This is one area where, depending on how you choose to evaluate it, the ETF models can have lower, higher, or similar drawdown characteristics as the broader market. Let’s look into why that is the case.
Draw Down Chart
A “Draw Down” chart (or sometimes it’s called an “Underwater” chart) basically tracks how many percentage points you are off the all-time (or period) highs for each day or week. So if a stock was putting in a new high every day, the chart would basically plot a straight line at 0%. If the stock moved down 10% off the highs, the chart would plot the line or area as negative 10%.
In this chart above, I have plotted both the equity and draw down of the SPY so that you can see how they relate to each other. The equity is in blue and is based on a $1000 investment in the SPY at the start of 2007. The dollar values are plotted on the left axis.
The drawdown is plotted on the right axis (0% to -100%) and shows up in orange. During the depth of the financial crisis, the SPY lost over 50% of its value from its all-time high in October 2007. You can see the orange area dropped below -50% in early 2009.
In the last two years, the SPY has been putting in frequent new all-time highs, so the drawdowns have tended to be smaller and you can see that the orange shaded area in that portion of the chart is, likewise, very small.
In this chart above, I have plotted the same type of drawdown, but now comparing the ETF Complete model to the SPY (and ignoring the equity plot). There are a few things that come immediately to mind. The huge drawdown in the SPY in 2007-2008 dominates the chart and it took the SPY almost five years to recover. The ETF Complete model had relatively tame drawdowns with the largest being just around 13%.
This difference should not be glossed over. All of our ETF models (Sector, Country, and Complete) had significantly smaller drawdowns than the overall market and even managed to emerge from this period in the green. You want to know that your trading model is robust enough to work well in a variety of market conditions—especially poor ones.
However, when looking at the data, it’s important to both consider the “outliers” like the financial crisis, but also see how the model compares in more “normal” situations.
The above chart looks at relative drawdowns if you had arbitrarily started tracking the data at the bottom of the crisis (so it ignores the fact that the SPY was down over 50%). Here we can see that if you ignore the “outlier” of the giant SPY drawdown, the average drawdowns of the SPY and Complete are much more similar.
The Complete still outperforms the SPY on this metric, but the average difference is a lot closer. We can also see that some drawdowns are correlated (meaning they happen around the same time and to a similar degree) and some are uncorrelated. Variance in this just depends on the type of holdings, how much cash is in the portfolio, or what stages the component trades happened to be in at the time.
When we compare the post-financial crisis drawdowns between the SPY and the Sector Stops & Targets we can see that the Sector Stops & Targets has larger draw downs. We should probably expect as much since the SPY is the average of over 500 stocks whereas the Sector model has at most three holdings and they tend to be more highly leveraged.
Fortunately, the Sector model compensates for this deficiency with higher average performance. But it is important to realize that it will have higher overall drawdowns and account for this increased volatility by using less aggressive position sizing and capital allocation.
Understanding the types and characteristics of both “typical” and “unusual” drawdowns is an important part of decision process when it comes to trading. It may not be possible to know everything and always be completely prepared for every situation, however, the more information you have and can keep “in front” of you, the less likely you are to be blown out by market action.
The Current Condition of the Model
We made one position change this week. The model rotated out of TMF and into SSO. TMF was a great trade, lasting just over a year and netting around a 60% gain for the Basic model. The Stops & Targets model reached two targets and had been stopped out of the trade since last fall.
Our two other positions are FXI and EWJ. FXI is near its $52.67 target (came within 11 cents a week ago).
Stay tuned to daily updates for any position changes.
Here is a summary of the weekly performance of all the ETFs that the strategy monitors.
Best wishes for your trading,
James Kimball
