News Focus
News Focus
icon url

capt_jmj

04/25/06 3:53 PM

#29734 RE: stock-trace #29732

PT, re: 1000x, please see my post from this morning on SC. We are there, Master! (Got one bullet left for today; trying to save it)

Great, another partner in crime! Are you going to try the programming changes yourself? If so, I would offer my humble assistance and/or consultation if needed, and I am very interested, as I am sure PT and others will be, in any results you get which improve the bottom line. Just be careful with those logic statements; they can be a real bear! ;-) As a starting point, I would suggest getting with PT and figuring out how your ideas might fit in with the current rule hierarchy, because in trying to improve the results, you often end up initially with the opposite, and then you sit there scratching your head (or the other end), wondering if the theory is faulty or you just don't have it coded right, and most importantly, in the proper order of dominance. The first condition satisfied in a logic statement that may go on for two lines will end the evaluation.





icon url

stock-trace

04/25/06 8:51 PM

#29759 RE: stock-trace #29732

cj:pt:tc:

well to keep it simple, i suggest that

1. in bull mode: no changes.

2. in bear mode- if rsi<28, override variation:

a. in bear mode, if nasi > (- 250) no changes.

b. in bear mode, if nasi < (- 250) wait for bull mode prior to long override.

i think it may reduce some of the drawdowns. may also make it simpler by substituting nasi 0 value instead of -250.

thx much

st