I have been working these last few weeks on developing a method to limit losses on those months which inevitably occur when the market moves against your initial entry buy for the cycle.
These months do not happen often, but when they do we want to suffer as little damage as possible. To mitigate this loss on these months I have added a few Rules to GLAD/3%. I am more than happy to entertain critiques and help in improving these Rules, so sound off if you have constructive additions to these Rules.
1) Do not enter positions (GLAD or not) when the entry PPS is 4% or higher than the last (X-1) closing PPS.
2) The period or cycle can be dissected into three sub-periods of approximately 7 days each. The first two periods define the U shape bottom of the period, with the last seven days the growth side of the U. We do not expect this last seven days to go down, but sometimes stuff happens...;-)
Therefore, do not enter GLAD/3% if you have not yet had entry before (X-7). So no entry even if GLAD met during the last 7 days of the cycle.
3) The stop loss I have found best is 2.5%. This 2.5% is only to be employed on the last 7 days of the cycle. So the 3% Rule can be used up until the last 7 days only. This is a change. Example:
Two 3% entries have been made, so with the original purchase you have three tranches, before (X-7). Suppose the entries were on PSEC @ 10.80, 10.48, 10.17. So your average is 10.48. Now it is X-7 date. You take 2.5% stop on the average position BE of 10.48, which is 10.22. That is your mental stop on the closing PPS's from (X-7) all the way to the exit on (X-1).
So if the PPS never sees a close of 10.22 or less on those 7 days pre- EX, you close out on the normal (X-1). If the close hits 10.22 or less on the close of any of those days, you are stopped out and immediately close out your entire position.
That's it for now. I welcome discussion.
Doc, a question on 3) the use of 2.5% stop loss.Using your example above. You use the BE on the 3 purchases on (X-7) to set your stop loss, why would you not use the pps at that time if it is above your BE. Lets say the pps is 11.00 on (X-7) why not use that price to set your stop loss? Bow
You can do that if you wish Bow. Use the higher of the closing PPS on (X-7), or the BE figure, and deduct 2.5% for the stop. You can keep moving your stop up with each day's gain also. You don't want to take the PPS on any close to calculate the 2.5% stop if less than your BE, as this would stop you out at a greater loss than 2.5% on your total position.
The problem is with a higher stop PPS and the likelihood(most months) of an escalating PPS is that you might get stopped out B4 the X-1 date, whereas with a lower stop PPS you allow for more volatility as the PPS climbs to X-1. It is a balancing act for sure.
Some months are terrible like Nov 2012. GLAD got you in at 29.94 on Nov 2, 2012. Remember Nov 14th when the market had the flash crash? That day, according to the Rules would have stopped you out on the close @ 27.50, just to see a recovery two days later at 29.52 for over a 2.00 loss. I guess some consolation is that the prior August returned over 3.00. It "way" works out in the end although these losing whipsaws are painful.
I have wrestled with trading with no stops. Trading 3% Rule up until X-1, and other iterations. I am all ears for help on capturing as much gain while suffering as little loss as possible.
Hi Doc -
I haven't had time to do much lately, but your approach seems like a good one to me. The average curve on most of these low PPS monthly's we have been discussing is a shallow curve as you describe. With the shallow curve and short time period, absolute gain is limited on most to 0.20 - 0.40 on average. (not a bad ROI annuallized) Limiting the losses to increase the gain would seem worth study. I haven't had much luck so far, with short term trailing averages in defining entry points. I am expanding it now to look to the prior period. Another area worth looking at might be confidence factor, although VIX does not seem to go well with these stocks. Interesting stuff!
I downloaded CRF data today. Looks like another good one at first glance. Thanks for sharing.