|
PureBytes Links
Trading Reference Links
|
I saved this post because I like systems which work from patterns rather than
indicators - got a chance to check it out this weekend using S&P continuous
contract data 1990-present. Used entry at next day's open with 2 day hold and it
tested 67% profitable, however I didn't care for results as the ratio of avg
win/loss was barely over 1 and the drawdowns were very large. Based upon
observation that many losers occurred because market declined rather than
rallied following the inside day, I changed entry to high of day preceding
inside day + 1 tick and that eliminates most of the losers as well as much of
the drawdown. Works well with day prior to inside day being lowest low in 4-16
days and fixed holding period of 2-3 days so the results do not appear to fall
into the over optimized category. It does not trade very frequently, however
most trades are in non-trending markets. Here is the complete system for SC or
TS. The use of inputs allows testing over a range of values:
Inputs: LowDays(8) and BarsHold(3)
Long entry: (High <= High[1]) and (Low >= Low[1]) and (Close < Close[1]) and
(Low[1] = Lowest(Low,LowDays)[1])
Buy Next Bar: High[1] + .10
Long exit: BarsSinceEntry = BarsHold
Exit Long This Bar On Close
Some additional testing indicates that the system trades almost twice as
frequently and profitably without any requirement relating the close of the
inside day to the close of the previous day.
Any questions regarding this post should be addressed to the list.
Earl
----Original Message-----
From: POMPATIS@xxxxxxx <POMPATIS@xxxxxxx>
To: RealTraders Discussion Group <realtraders@xxxxxxxxxxxxxx>
Date: Wednesday, July 29, 1998 11:46 PM
Subject: Re: S+P inside day w/ DOWN close
>Check this out:
>
>According to a table I found in an old L. Williams book, the percentage of
>time SnP closes up after an inside day where the inside day closes down and
>the day prior to the inside day is a 10 day low:
>
>71% the first day
>71%the second day
>
>I have no idea how much data this is based on. But there you have it folks.
|