Our current three positions are TMF, SSO, and IFN. There were no position changes in the country model this week. The stops and targets model is in 2/3rds a position of IFN, full position in SSO, and stopped out of TMF.
This Week’s Strategy Lesson: How to Trade the ETF Models (Part 3)
In the first part of this series we covered the overall objectives and basic rules that shape the contours of the ETF strategies. Last week we discussed the differences between the basic rotation and stops and targets versions of the models. This week we will cover how to approach rebalancing.
Portfolio Positions
We use a fairly basic and consistent set of rules for money management in the ETF Strategies. We initially divide our allocated capital evenly among the positions (three positions for each individual model or nine for the complete model). Once this capital division occurs, we treat each of these “pools” of money as separate portfolio slots. When we rotate out of a position into a new one, the new position takes on the same position size as the old one.
This is in-part necessitated because we don’t know how long we will be in a trade or when they might occur. But as time goes by and different positions earn potentially very different returns, these portfolio slots can quickly get unbalanced.
For example, just looking at the Sector Stops & Targets portfolio from January 2014 through July 2014 the entire portfolio was up around 28% but the performance variance between the portfolio slots was quite varied. The first slot was in semiconductors the entire period and was up 22%. The second slot had several rotations between technology and real estate and was only up around 11%. The final portfolio slot had great trades in healthcare and energy and was up over 53%.
The average six month period might not typically have this much divergence but it’s not hard to imagine if one position is reaching targets while another position is negative or get stopped out, that there can be a big divergence between these portfolio positions.
Diversification
In many ways, it makes sense to let the winning positions grow larger as they are outperforming, but it makes less sense that when you rotate out of that trade into a new one, the new position will be artificially larger than all your other positions. And if this continues for a while, you could have a situation where a majority of your capital is one position while the other two are very small.
When this happens we lose some of the benefits of diversification. We don’t know which one (or two or three) of our positions will do well or which might move against us. By maintaining roughly equal position sizing we minimize the risks of an outsized loss when one position moves against us.
Of course, rebalancing every day would ensure that the positions are always equal, but that would require a lot of work to execute and depending on how much capital you are trading with, the transactions costs of buying and selling partial positions every day would likely dwarf any daily profits. So rebalancing is about balancing the costs versus the potential benefits.
Rebalancing
There isn’t necessarily any one optimal way to rebalance. During any given period it will make difference, but in our testing using over seven years of data from the ETF models, different methods of rebalancing tended to average out to small differences over very long periods of time. In addition, since not everyone started following the models at the same time or use the same position sizing methodology, it wouldn’t be possible to tell you when you should rebalance or the exact steps to take for your account.
For these reasons, we have tended to leave it up to the users to rebalance at their own discretion, however, we will cover here two different methods that you can use to keep your own portfolio balanced.
Perhaps the easiest one to use is a simple calendar-based rebalancing. This means that perhaps twice a year (January and July for instance), you calculate the total value of your portfolio including cash positions or profits you have taken out of the market and divide that by three (or nine). From this point you would buy and sell shares of your different positions until they are roughly equal to that 1/3rd value.
So if $5000 is the average position size and you have $4000 in ABC and $6000 in XYZ, you would sell $1000 worth of XYZ and buy $1000 worth of ABC so that both have around a $5000 value. With share prices being different, you might not be able to get them exactly even but that will not make much difference.
You can decide if the variance between the position sizes is large enough to warrant incurring the trading costs and commissions. For smaller accounts or for those who pay high commissions, it might not make sense to attempt to balance out small differences.
The other method you could use would be a percent difference method. For this, you would only rebalance if the largest position was “X” percent larger than the smallest position. Depending on your account size and costs, you could choose a large number like 25 or 30 percent or a smaller number like 10 percent.
We have tested both of these methods and the resulting differences between them are small enough that we can’t recommend one method over another from a performance standpoint. So it really comes down to your own personal preferences and which style might be easier to execute.
We hope you have found this short series on how some features of the ETF models work and some guidelines for trading the system. We hope you have a wonderful and refreshing Thanksgiving holiday weekend.
The Current Condition of the Model
For the country model, we are in TMF, SSO, and IFN. All three positions are up on the week. Treasuries made a nice move this week and IFN is only about 1 ATR (30 cents) from its second target.
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
Trader & Analyst
MarketGauge