Registered User
Joined: 11/28/2005
Posts: 15
|
I'm trying to create my own new PCF that uses the formula logic of the Moneystream Surge - 1 Wk. system PCF. You can display the formula criteria of this system PCF as:
100 * (MS - MS1.5) / (MAX(MS1,99) - MIN(MS1,99))
The PCF I'm creating produces the difference between a 1 week MS surge and a 2-day MS surge. What I'm finding is that the above system formula produces erroneous results. In fact, I created an exact duplicate PCF using the formula above, and when I display the results of both the system PCF and the clone PCF in side-by-side columns, the values are different! I need to confirm if the above formula is in fact the one being used in the system PCF, or if the criteria info. display for this PCF is bogus. I've recalculated everything and can't see any reason why the results are not the same.
I called Telechart support and they directed me here. Any help would be appreciated. Thanks.
|
|
Worden Trainer
Joined: 10/1/2004
Posts: 18,819
|
We saw this, bluto (deleted your reposting).
I duplicated what you did and am looking into it. Stay tuned.
- Craig
Here to Help!
|
|
Registered User
Joined: 11/28/2005
Posts: 15
|
QUOTE (Craig_S) We saw this, bluto (deleted your reposting).
I duplicated what you did and am looking into it. Stay tuned.
Ooops...sorry. I wasn't sure if I had posted in the right forum. Thanks for confirming, Craig. It was driving me nuts! 
|
|
Worden Trainer
Joined: 10/1/2004
Posts: 18,819
|
I regret not letting you know I was looking into it.
- Craig
Here to Help!
|
|
Registered User
Joined: 11/28/2005
Posts: 15
|
Hey....have you guys forgotten about this? I really need an answer so I can finalize my PCF. I'd imagine this should be a simple thing to verify. 
|
|
Worden Trainer
Joined: 10/1/2004
Posts: 18,819
|
I've not heard back from the proper people yet. I will poke with a stick again.
- Craig
Here to Help!
|
|
Worden Trainer
Joined: 10/1/2004
Posts: 18,819
|
Just conferenced with the Data department on this. It seems to be problem that may involve the develepment team. They are working on getting it fixed ASAP.
- Craig
Here to Help!
|
|
Worden Trainer
Joined: 10/1/2004
Posts: 18,819
|
In the meantime, create a PCF with this formula:
100 * (MS - MS1.5) / (MAX(MS1,99) - MIN(MS1,99))
it will give the correct values.
- Craig
Here to Help!
|
|
Worden Trainer
Joined: 10/1/2004
Posts: 18,819
|
The issue was finally tracked down and the system criterion will match any created PCF by tomorrow's closing download. Thank you for your patience, it was not an easy problem to diagnose.
- Craig
Here to Help!
|
|
Registered User
Joined: 11/28/2005
Posts: 15
|
QUOTE (Craig_S) The issue was finally tracked down and the system criterion will match any created PCF by tomorrow's closing download. Thank you for your patience, it was not an easy problem to diagnose.
Terrific, Craig! Thanks for your persistence on this.
|
|
Worden Trainer
Joined: 10/1/2004
Posts: 18,819
|
Thank you for finding something we did not realize was an issue.
- Craig
Here to Help!
|
|
Registered User
Joined: 11/28/2005
Posts: 15
|
QUOTE (Craig_S) The issue was finally tracked down and the system criterion will match any created PCF by tomorrow's closing download. Thank you for your patience, it was not an easy problem to diagnose.
Craig - Finally had a chance to check and the issue has in fact not been resolved. The criteria formula of
100 * (MS - MS1.5) / (MAX(MS1,99) - MIN(MS1,99)) continues to display for the system PCF of "MoneyStream Surge - 1 Week" when you click the "view formula" icon. I confirmed this by creating my own PCF again using the above formula and comparing the results to the values produced by the system PCF. Either the above formula is wrong, or the computation inside the system PCF is incorrect and different than what the above formula indicates. I was assuming as you mentioned that all of this would be remedied and in place by the Tuesday end of day update.
 
|
|
Worden Trainer
Joined: 10/1/2004
Posts: 18,819
|
That was ma hope as well. It fell on the priority list and is scheduled to be corrected in tonight's end-of-day download.
To be clear, the problem is with the built-in condition. Your PCF is showing the correct values.
- Craig
Here to Help!
|
|
Registered User
Joined: 11/28/2005
Posts: 15
|
QUOTE (Craig_S) That was ma hope as well. It fell on the priority list and is scheduled to be corrected in tonight's end-of-day download.
To be clear, the problem is with the built-in condition. Your PCF is showing the correct values.
Whew....in the meantime, can you tell me how I could devise a shell formula for a "MoneyStream Surge - x Days" PCF and which components I'd need to change depending on how many days? I'd like to build a series of these for different time periods. Thx.
|
|
Worden Trainer
Joined: 10/7/2004
Posts: 65,138
|
bluto,
Let's break down the formula:
100 * (MS1 - MS1.5) / (MAX(MS1,99) - MIN(MS1,99))
The (MS1 - MS1.5) portion is the difference between the current MoneyStream value and the value 5 Periods Ago. If you wanted a "MoneyStream Surge - 10 Days" PCF, you would need to change this portion to (MS1 - MS1.10).
The second portion, (MAX(MS1,99) - MIN(MS1,99)), is the difference between the Maximum and Minimum values of MoneyStream during the last 99 Periods. It is necessary for two reasons, a surge is by definition relative, and because MoneyStream is a Cumulative Indicator (meaning its actual values are meaningless and can only be compared to other values of MoneyStream for the same symbol calculated at the same time). Using 99 days is fairly arbitrary, but you might wish to use it for your other MoneyStream Surge formulas to make direct comparisons easier.
It should probably be noted that once the MoneyStream Surge - 1 Week System Criteria is fixed on the servers, it will still only match its associated Personal Criteria Formula when that PCF is updated using Market Close data because the System Criteria is only updated at the end of each day. Once you have downloaded the following day, it will no longer match because the data on your system will be newer. At that point, it should match the following PCF instead:
100 * (MS1.1 - MS1.6) / (MAX(MS1.1,99) - MIN(MS1.1,99))
You may wish to review the following:
Dealing with OBV & MS in PCF's - how to interp their "values"
-Bruce
Personal Criteria Formulas
TC2000 Support Articles
|
|
Registered User
Joined: 11/28/2005
Posts: 15
|
Awesome, Bruce You and Craig are "Da Men". 
|
|
|
Guest-1 |