|
PureBytes Links
Trading Reference Links
|
Thanks for your reply. I was trying to determine why a 14 day RSI in
Metastock differs (by approximately 10 points) from Reuters' RSI. Equis
stated they were using the formula in New Concepts in Technical Trading
(which apparently is much more in depth than the one indicated in their
manual).
At 10:35 AM 11/2/1998 -0800, you wrote:
>Heidi,
>
>Are you using the same number of periods (i.e., 9 or 14) as the default?
>Each service "tweaks" their own set of numbers. Wilder suggested 14 in his
>book, but he personally suggested 9 to me two years after he wrote it. You
>might want to "run" a number of RSI's for a while, until you find the
>periods you are most comfortable with.
>
>Steve Karnish
>CCT
>
>----------
>> From: Heidi Stubner <stubner@xxxxxxxxxxxxx>
>> To: metastock@xxxxxxxxxxxxx
>> Subject: Re: Relative Strength Index
>> Date: Monday, November 02, 1998 9:10 AM
>>
>> I'm trying to calculate how Equis comes up with the RSI figures. I've
>> tried using the formula in the manual (100-(100/1+U/D)) where U = change
>of
>> up days/time period and D= change in down days/time period.
>>
>> I'm coming up with at least a 5 point difference. I've contacted Equis
>and
>> they indicated that the real formula is in Wilders book: New Concepts in
>> Technical Trading Systems. Unfortunately she wasn't able to send me a
>copy
>> of that formula. My Reuters RSI is also different from the Metastock #
>> (and the # I'm coming up with).
>>
>> Is there someone who has a copy of his book and could either post the
>> formula or could fax it to me? I'd appreciate it very much.
>>
>> One question is how they would adjust for a period where there were no
>down
>> days during that time frame? Obviously you couldn't divide the # of up
>> days by zero.
>>
>> Any assistance is greatly appreciated.
>>
>> Heidi
>>
>
>
|