If I'm calculating right, the first 77 trades of the BB system have a negaive expectancy. Just so we don't duplicate each other's work. I've tried it without taking Fri triggers and not taking stocks below $15 and neither improves the system. My other ideas are 20d SMA is higher than it was 20 days ago, has a open red candle, only on second signal, only if stochastic rsi (default) is 0, only if rsi is < 30.
If you've tried any of these please let us know. That way I won't have to do each one.
Thanks,
h

Howard,
As is the Bollinger Band system has produced -5.8%, if my calculations are correct. If you were to incorporate the two filters I suggested into the 77 trades Larry has listed you would reduce the number of trades taken to 32 but raised the profitablity to +12.6%. I'm not sure if this is just chance or not but an 18.4% difference does seem significant. I supect that 32 trades in a month is more than enough for most of us (there were more that Larry didn't post) It seems like it would be beneficial to wait for the trades that have the best chance of winning. My theory is that the Bollinger Band strategies and VTO worked very well in the first leg down of the bear market because everyone assumed that it was just a correction in the great bull market. As the reality sunk in that we're not in a bull market anymore the urge to aggressively buy the dip has disappeared. I suspect that most of the strategies that worked really good from 2000-2003 are probably not going to produce great results anymore. We're just going to have to come up with strategies that perform well in a stagnant market.

Jim

I thought I might list some of my revisions and deletions to the various ideas I have posted and possibly get feedback on improvements.

The B/B I already listed in the previous post.

The short system is exactly the same except I'm going to add the under 200 dma rule to it. It doesn't seem to make any significant difference if the 200 dma is pointing up or down.

The filled Black candlestick/Gap up strategy I haven't modified except I don't limit myself to just S&P and Nasdaq 100 stocks. It doesn't occur enough to just use only 550 or so stocks.

The Gap up hollow white candlestick strategy. I'm going to add the above the 200 dma rule.

The hollow red candlestick I made a lot of changes to. First the stock must be above the 200 dma. Second the price action for the day must of at least moved below the 10 dma but not to or below the lower Bollinger band. I also don't take signals that close above the upper bollinger band.

I'm also going to start using the crossover system that Gary posted the other day. So far it has tested better than any of the other systems. I am going to add one rule to the system though. If the trade goes against you enough to become either overbought (+70) or oversold (-30) double the trade. This rule improved results to about 99%.

Jim

Remind me Jim what your two filters are?
h

Jim
forgot to add, could you send me the trades so I can plug them in without having to find them all over again. The ones that meet your filters I mean,
h

First the stock has to be above the 200 dma and second if the day that the stock closes below the lower bollinger band has more than 1.5 times normal volume pas on the trade. Here are the trades that would have passed the filters in order:
1. AYE
5. XOM
8. COP
13. CMCSA
15. SGP
17. ABI
18. BDX
19. ABT
20. BCR
22. HCA
24. THC
30. ESRX
31. PBG
34. DGX
38. MHS
42. T
44. WWY
49. GNTX
50. LNCR
52. GDW
53. IR
54. JNJ
55. KO
57. AMGN
58. ANDW
59. DIS
63. TEVA
64. AAPL
68. APPC
69. BNI
70. GENZ
71. UNP

Jim

I tested the above 32 trades using the 2 filters and the hold till the stock closes above the 10 dma. I found a sizable mistake with GNTX as Larry has it listed as a 3.2% profit but I think it was actually a .8% profit.
Using standard rules I'm showing an 8.2% loss.
Using standard rules and the 2 filters I'm getting 7.7%.
Adding the hold till a close above the 10 dma I'm getting 24.2%.
One loss could potentially wipeout the 7.7% using the standard rules plus the filters. I'm inclined to say that holding is the way to go with this system.

Jim

I've come to the conclusion that the way to play the short system is to incorporate the crossover rules into the system. Only take signals that are under the 200 dma, 200 dma in an obvious down trend and cover on the first close of an RSI(10 day) of -30 or under or 2 months which ever comes first.

Jim

Larry,
Have you tested the B/B strategy with a years worth of data? I started last night and I'm not sure the system has a positive expectancy no matter what filters I use. I've only done the first 10 stocks in the Nasdaq 100 though. If you already ran the test I'll save myself the trouble.

Jim

Jim,

No, I haven’t except to do some spot checks. I’m going to try it for awhile looking for divergences -- stocks making lows after trading below the lower BB with an oscillator like MACD histogram or whatever ticking up, with a little bit of intuition thrown in. I think I can get much better entry points and most of the losses can be avoided. I just haven’t decided when to exit yet. If I can’t make that work for significant profits, I’ll just scrap it. It’s too much hassle for small profits or, worse, small losses.

Larry

Larry,
I have an idea. A few of us have been experiencing difficult fills. My thought is to observe the opening price as its listed then place a limit order 2% below that price. At first glance it looks like a lot of the trades will get filled. If it ends up being a losing trade well we just saved 2%.

Jim

Larry,
I started testing my 2% idea, the results were not good. I'm not sure any combination of filters can make this system work. When the market reverses and heads down this system just falls apart. I think I'll leave this one to you, I'm fresh out of ideas

Jim

Larry,
I didn't test this but I wonder if taking signals at B/B level (10,1.5) like Altucher did would produce more winners and offset the large losses.

Jim

Jim,

What do you mean? How is that different from what we were already doing?

Larry

Larry,
Instead of waiting for the signal to drop 1.9 standard deviations take it at 1.5. I believe that is the level Altucher used. You should get a lot more signals. Maybe taking the trades at that level will result in enough wins to make the system profitable. I'm just guessing since I haven't tested it. I have been assuming that the more oversold a stock is the greater the chance of a bounce. Perhaps the opposite is true, when a stock gets extremely oversold maybe that increases the odds that its not going to bounce.

Jim

Oh. The reason I use 1.9 is when I read Altucher's book I couldn't get my BB readings to match the examples he showed using 1.5. I was using Stock Charts and he was using a back tracking software program. By adjusting Stock Charts to 1.9, it more closely duplicated his examples than using 1.5.

Larry