October 27: How did the Selects Switching System perform on "Gray Monday"? Overall, pretty well all things considered. The primary holding at that time was Energy Service. It fell about 8% that day, a bit more than the market averages, but it had been essentially unchanged on Thursday and Friday of the previous week when the Dow fell over 300 points. In the following two weeks, that fund regained the drop and exceeded its earlier high by a small amount. Later in November, the fund fell very rapidly, which is discussed in a later question.
There were stop-loss sales in some accounts that employ that tactic, and, unfortunately, these worked very badly. The closing prices on October 27 triggered the stops, which were placed for execution at the 10:00 AM prices on Tuesday, October 28. In this case, that was exactly the wrong move because the market fell sharply in the first hour or so and then rebounded strongly with Dow recording its largest one-day point gain ever. That meant the stop sales were at the lowest point possible point and resulted in losses in the 12-15% range. This is far below the 7% trigger level and the 8-9% that I would expect due to the nature of mutual fund transactions.
Stop-Losses: Given how poorly the stops worked in late October, would it be better not to use this tactic? Even though the use of a stop-loss backfired this time, I do not think anyone should abandon them at this time. The main reason for stop-losses with the Selects Switching System is to prevent enormous losses such as those possible during the 1987 crash. As is true for any strategy or tactic, sometimes it will help and other times it will hurt. This time it definitely hurt. However, if your reasons for electing the strategy still apply—usually a desire to avoid huge losses on a single trade, then you should still use stop-losses. I expect the effect on investment returns from the use of stops over a period of several years to be small, but as we have seen, this won't be true for each instance when they apply.
It is important not to judge a strategy or tactic by short-term performance. Had the market meltdown continued resulting in much lower stock prices, then those who used the stops would be pleased and those who did not might well be asking if they should start using them. (My answer would be no if the reasons for not using stops were still applicable.) While it would be nice if we knew when the stops would help and when they would hurt, I will readily admit to having a very cloudy crystal ball that is of little use in any specific instance.
Energy Service in late November: We gave up over half of the profit in Energy Service in the last two weeks of the trade. Had something like this ever happened before, and should we have gotten out sooner? In this case, it obviously would have been better to get out a week or so earlier, but the system did not call for that. The use of trailing or retracement stops would, which some strategies employ, also would have preserved much more of the potential profits. I have studied such tactics and concluded that they would probably decrease overall returns, but they would also decrease volatility and possibly drawdowns. In late 1996, the system stayed with Electronics during similar highly volatile ups and downs, which was correct then because the fund resumed its upward movement. As is true for stop-losses, no strategy or tactic works best all the time. What I try to do is find those that work well over the long haul and stick to them. I emphasize total returns, which may mean living through greater volatility.
Depending on how one evaluates the drop, with the exception of the 1987 crash period, what happened to Energy Service from November 10 to 24 was the worst or second worst drop by a fund that might have been bought by the system since the start of 1987. I tried to put price changes in context by comparing the weekly changes in fund prices with those of the Vanguard Index 500 fund. From November 10 to 17, Energy Service fell 5.1% while the market was up 2.8%. The 7.9% difference was large, but not nearly a record. The following week, the fund fell 10.4% while the market was virtually unchanged. That difference was the largest except for the 1987 crash and once in 1992. The two weeks combined were unprecedented. The few instances with behavior similar to the first week did not reveal any meaningful patterns. For now, I consider what happened to be a rare (I hope) aberration. After all, the system and trade produced profits that were typically above 20% in four and half months. I think we should be happy with the end result and not focus too much on what might have been.