I’ve discovered a way to make lots more money!

		Init. Price $		Gain ‱
Symb	Buy	Purch.	Stop	Old	Craig	Mine	Actual
SCHV	JA04	28.56	28.21	-12	-12	-2	+5
DRR	JA06	48.04	45.15	-40	-40	-40	-66
OUTD	JA10	7.57	7.24	-14	+37	+56	-3
ATML	JA10	13.49	12.49	+6	+6	+6	+12
TNA	JA13	77.20	70.60	-61	-61	-61	-134
GGG	JA20	40.87	39.55	+10	+20	+19	+17
OKE	JA21	58.47	56.36	+1	+1	+1	+0
GBX	JA25	22.71	20.60	+38	+65	+65	+43
BAL	JA25	82.30	75.00	+33	+69	+48	+27
BAGL	JA26	15.95	14.65	-53	-53	-53	-47
SE	JA26	25.82	24.86	-3	-3	-3	+5
FLIR	JA27	30.35	29.26	+36	+36	+25	+6
ETN*	FE03	108.38	104.50	+29	+28	+23	+29
GIII	FE03	37.32	33.95	-19	-19	-19	-12
ATML*	FE04	14.94	13.44	+100	+100	+91	+100
MKL*	FE04	411.50	400.70	+10	+10	+11	+10
(Average)				+4	+12	+10	-1

Yesterday I reread Craig Ferguson’s website about how to do swing trading.  It’s the sort of thing I have to reread every month or two because the words don’t mean anything until you have some relevant experience.  I happened upon his 2-for-1 strategy page and thought, “IT COULD WORK!” (see Gene Wilder @ YouTube).  The 2-for-1 strategy (which is recommended for newbies) is to sell half your shares if the stock rises above your initial purchase price by the same amount that your initial stop was below that.  But Craig’s stops are much higher than mine (he prefers 2-5 day holds), so maybe I should sell half the shares if the stock rises by half the initial stop depth?  And if it rises by that amount again, sell another quarter?  And then an eighth, and so on until the remaining shares are stopped out or MACD turns negative.  Would that be better?

The table at left shows how profitable this new rule is.  “Old” was the New Rule for last week (use pSAR for stop prices, then wait to get stopped out).  “Craig” is the 2-for-1 rule, with one sale at the next opening price after the closing price exceeds purch+(purch−stop).  “Mine” is my revised rule, with sales at purch+N×(purch−stop)/2 where N is a positive integer.  “Actual” is what actually happened.  For stocks I still owned as of last night (shown with an asterisk), the “actual” value was yesterday’s closing price.  For this set of stocks, Craig’s rule is slightly better, but I like my rule because it lets me sell earlier and oftener.  Still, it probably needs more tweaking.

For now, my sell-stops are automatic (they can trigger any time of day) but my revised-rule limit sales are manual (they get the next day’s opening price).  I can’t automate both of them because that would need an order like "sell-stop all shares at $X−Y, or sell-limit half of shares at $X+Y/2 and then remove those shares from the sell-stop”.  That’s too complicated for retail trading.

I put in trades this afternoon for any stock that my revised rule said should already have been ½ sold—which was most of them.  Then the market went down later in the day so I locked in my profit!  Now the pie chart looks very busy, but that should clear up by Friday when today’s sales become cash.  I wasn’t able to sell half of ETN because all of it got stopped out earlier today at +12‱ (would have been +16‱ if I had put in my trade last night and gotten the opening price).

I called Schwab to ask WTF happened with GIII, which got stopped out yesterday even though my stop price was lower than the day’s official low.  They said there was a extra-low bid at the open, which no one accepted but still it triggered my stop and caused me to get the lowest price of the day.  This doesn’t matter much (I would have gotten stopped out of GIII today in any event), but still I dislike the idea that a stop can be triggered by a hidden bid that doesn’t show up in the exchange’s history of what happened that day.  This makes hypothesis-testing difficult because I can’t really tell which day a particular stop price would have triggered on.  Craig has stopped using stops for this reason, but I think I'll stick with them a little longer, just to make sure that I sell when I said I would.