Show Latest Posts:

The 10 Most Recent Messages By gritton

Take me back to where I was.
  • Date: 9/1/14 4:27 PM
  • Number: 252778
  • Recommendations: 1
IIRC the Gritton backtester (not the same as GTR1) has only the standard 1700 stock edition as its universe. So you don't have to do anything explicitly to limit yourself to that universe.

Quite so. If you limit yourself to
(Continued...)
  • Date: 8/4/14 11:42 PM
  • Number: 252271
  • Recommendations: 7
For calculating the Sharpe ratio myself over the desired period, what is the source of the 1-year bond?

It's actually the 90-day T-Bills, data from Yahoo's ^IRX. OF course if you're calculating it yourself, you're free to use 1-year
(Continued...)
  • Date: 7/28/14 1:08 PM
  • Number: 252127
  • Recommendations: 3
Wow. Bunch of old people

When I started here, I used to belong to the "under 40" group. Of course, I used to belong to a long of younger groups back in the day...

- Jamie
  • Date: 7/25/14 12:47 PM
  • Number: 252041
  • Recommendations: 2
If I could, I would have answered both "it's all me" and "100% lucky." C'est la vie.

- Jamie
  • Date: 7/4/14 5:17 PM
  • Number: 251738
  • Recommendations: 11
Relevant: http://xkcd.com/882/

- Jamie
  • Date: 6/30/14 1:02 PM
  • Number: 251656
  • Recommendations: 5
At least with the newsletters the performance is all out of sample.

Yeah, that's been a problem. There was an effort a while back to put discovery dates on the screens, but it was pretty spotty. It would be nice to have an (easily
(Continued...)
  • Date: 6/29/14 1:33 PM
  • Number: 251627
  • Recommendations: 9
Q: Why is *this* one a good one?
A: It has survived so far.


Ummm how is this any different from stock screen selection?

Very different. The bad stock screens we have here have mostly "survived". (Continued...)
  • Date: 6/12/14 12:55 PM
  • Number: 251311
  • Recommendations: 21
Backtest TA methods and half of it is too vaguely defined to be testable at all, 40% fails the tests, and 10% works.
Let's be generous and say 20%, giving it a 4 our of 5 stars on the mumbo jumbo scale.

It's true that something that

(Continued...)
  • Date: 6/10/14 11:31 AM
  • Number: 251253
  • Recommendations: 4
backtest.org's behavior isn't related to Value Line's strange 1-year TR behavior, but is instead tied to a calendar weirdness of its very own. Well, not quite its own but an attempt to recreate the schedule that
(Continued...)
  • Date: 4/22/14 11:15 AM
  • Number: 250415
  • Recommendations: 8
I'm curious, are there other data sources available for a nominal fee that has cleaner data?
IOW, what's the next step up from Yahoo cost?


Yahoo's free data is good if you know how to use it. You can't use their easy-to-download
(Continued...)
Show Latest Posts:
Total = 10

Take me back to where I was.
Stock Folders: A B C D E F G H I J K L M N O P Q R S T U V W X Y Z