Last week we ended our weekly blog with a brief discussion on using Trailing Stop Orders (TSOs) to help protect a portfolio. This week, as promised, we'll take a deeper dive into this, plus examine the peaks and troughs from Jan. 1, 2004 through Oct. 8, 2013, for reasons which will become apparent soon.
How TSOs Protect a Portfolio
The point of using TSOs is to provide a measure of loss protection in a portfolio. For context’s sake, let's assume you have a portfolio consisting of 10 holdings. Some of the holdings are riskier than others. Overall, the portfolio has a certain degree of risk, depending on the percentage allocated to the riskier assets. In 2008, we learned that diversification alone is inadequate in the face of a market meltdown. In fact, during 2008, diversification disappeared and correlations rose! How can you protect the portfolio? With TSOs.
Where to Set Your Stops
In the event of a market collapse, which holdings would you be most concerned about? The riskier assets, of course. Think of using TSOs as a safety net underneath these assets. Hence, using mutual funds for short-term bonds is prudent as the worst case scenario is not that bad. Perhaps individual bonds and a few mutual funds could complete the remaining bond allocation. However, for riskier assets, mutual funds provide little protection as they can only be sold at the end of the trading day. Therefore, ETFs with TSOs are prudent for riskier assets since you can place stops under each position.
Determining where to set your stops for the TSOs is not an easy task. Since a TSO sets a price at which the ETF will sell, it's important not to set your stop so low that routine volatility would trigger a sale. I could think of a few ways to test this, but here's where I began. I used data from the Dow from Jan. 1, 2004 to Oct. 17, 2013. I should note that by Jan. 1, 2004, the Dow had already risen from its bottom on Oct. 9, 2002. Hence, the period I used includes a sideways trend of roughly 16 months before the second and final leg in the post-tech bubble rebound.
I plan to perform this same analysis segregating bull and bear markets in the near future. In any event, I then calculated all "peaks to troughs" in which the decline was at, or greater than, 5.0%. Finally, I divided the entire period into two halves, using the March 9, 2009 market bottom as the dividing point. The results are found in Tables A and B.
Table A covers the period from Jan. 1, 2004 to March 9, 2009 in which there were 1,894 days, 25 dips of 5.0% or more, and an average decline of negative 9.4%. Moreover, these periods of decline lasted an average of 19.6 days and the worst loss was a negative 22.0%.
In Table B, from March 9, 2009 to Oct. 8, 2013, there were 1,674 days, 16 dips greater than 5.0% and an average decline of negative 7.8%. The average decline lasted 25 days and the worst was negative 15.8%.
During the period included in Table A, there were seven double-digit declines, all occurring between May 2008 and the bottom on March 9, 2009. In the rebound, there was only one loss exceeding 10%. Realizing risk is an important factor, I also examined the VIX during these periods.
During the first period, the VIX averaged 19.2 with a standard deviation of 11.78. During the rebound, the VIX averaged 21.5 with a standard deviation of only 7.21. However, if you consider only the period from the market top on Oct. 9, 2007 to the bottom (in the first period), it's a different story. The VIX averaged 24.2 during this period with a standard deviation of 15.25!
At the present time, I plan to continue using a 5.0% TSO, pending further study. My next step is to segregate bull and bear periods and examine the declines in each. I'll write more on this as I complete my research.
Thanks for reading and have a great week!