|
PureBytes Links
Trading Reference Links
|
Well, I'll be darned. Look at that. Thanks Alan for a really cool post.
Walt Downs
CIS Trading
-----Original Message-----
From: Rob <sp500@xxxxxxx>
To: RealTraders Discussion Group <realtraders@xxxxxxxxxxxxxx>
Date: Wednesday, July 15, 1998 11:10 PM
Subject: RE:fdates
>Hey I want to take a chance to thank Mr. Sears for his W-dates. I really
>mean it!!!!
>
>This is the kind of stuff that I love and have already come up with a nice
>little system for W-dates.
>
>We all need to share more of this stuff. I will be posting my little system
>with W-dates in a couple of days.
>
>Thanks you Alan
>
>Rob
>
>At 02:48 PM 7/15/98 -0400, you wrote:
>>I am re-sending this as I never saw it posted and a lot of people
>>are asking the same question
>>
>>>
>>>What in the heck are WDates???
>>>
>>
>>
>>Wow, I have opened a can of worms here !
>>
>>My post was simply meant as a lighthearted parody of fdates :)
>>
>>My "extremely accurate" W-dates are simply Wednesday-dates
>>as noted in my initial message, the blue bars on the chart are
>>simply Wednsdays ! No secret, no membership required to get
>>these highly accurate turning points. Although for those interested
>>I could setup a subscription based service where I would email
>>everyone Tuesday to advise them in advance that the following
>>day was a W-Date <G> ie: Wednesday :)
>>
>>Using the exact rules of +/- 1 bar and the high / low not being violated
>>for 3 days following (Same as Fdate rules). Wednesday has a hit
>>rate of 80% or better !
>>Did I pick Wednesday at random ? NO
>>I ran tests on 20 years of historical data for many markets to arrive
>>at this conclusion. Monday & Friday had a hit rate of 50-60%
>>and Wednesday was consistently above 80% for all markets
>>tested.
>>Why did I do this ? To prove to myself and others that methods
>>such as fdates are no better than random.
>>Did I prove this ? I think so, who would have guessed that
>>using Wednesday as a pivot date using fdate rules would
>>give equal or better results ? :-)
>>
>>What I did was setup a spreadsheet to identify ALL pivots
>>that met fdates rules. Example: there may have been 3000
>>pivots in a data file which met this criteria, 80% of these occured
>>on Wednesday +/- 1 bar. I found the results to be statistically
>>significant.
>>Previous examples of Bean Oil & Sp500 proved my findings
>>Attached is a chart of cocoa showing the same amazing
>>accuracy !
>>
>>
>>
>>
>>Attachment Converted: "c:\internet\eudora\attach\Wdates3.gif"
>>
>
|