BlackBars
54 posts
msg #138449
- Ignore BlackBars |
9/29/2017 1:06:37 AM
So over the past months I have taken Stratasearch very seriously with my trading. I have read the entire manual and still have questions over some of the features that I would assume long time users of SS would know. So if any experienced SS user wouldn't mind answering a couple questions for me I'd be willing to pay them for their time.
|
mahkoh
1,065 posts
msg #138458
- Ignore mahkoh
modified |
9/29/2017 1:11:31 PM
Just ask ahead. If there's no response to your question it's very possible that nobody has a conclusive answer. Also try the forum and look if your question has been asked there. And about that forum: You may want to download an offline version
while it's still up.
|
Kevin_in_GA
4,599 posts
msg #138459
- Ignore Kevin_in_GA |
9/29/2017 1:42:59 PM
Ask away at no cost.
|
BlackBars
54 posts
msg #138466
- Ignore BlackBars |
9/29/2017 11:38:12 PM
I have been using the "percent of perfect" oneclick on many different ETF's. If I was to turn all these into multi system signals based upon the percent of perfect's criteria, will it STAY consistent as it has been through the results? Or will the percent of perfect's criteria eventually stop being profitable after a while?
|
mahkoh
1,065 posts
msg #138467
- Ignore mahkoh |
9/30/2017 8:43:49 AM
Why don't you just merge the signals? If you are happy with the one click result right click on it to create a signal, then under signal setup merge it with the signal for another etf.
Whether your results will be as profitable going forward as they were in the test is anyone's guess.
|
Kevin_in_GA
4,599 posts
msg #138471
- Ignore Kevin_in_GA |
9/30/2017 12:48:29 PM
My thought is that you should periodically reoptimize/rerun your OneClick setup - perhaps every six months or when you start to see results below prior backtest performance. It can be done in a day or so during a period where you have no trades ongoing.
|
BlackBars
54 posts
msg #138472
- Ignore BlackBars
modified |
9/30/2017 12:48:41 PM
Reoptimize?
|
mahkoh
1,065 posts
msg #138475
- Ignore mahkoh
modified |
9/30/2017 2:10:47 PM
Let's assume you ran your test from 2015/1/1 until 2017/1/1 and you notice that the system performance is deteriorating. You then clear the results, change the test period from 2015/7/1 until 2017/7/1 and rerun the test.
An additional thing you could do is go to the alternate data tab in the one click strategy and fill in 2017/7/1 until 2017/9/30 as alternate data#1. Then you go to strategy performance and create a new evaluation field "out of sample/avg return" and put ($F_A1MCAvgAnnReturn/ $F_MCAvgAnnReturn)*100 in the definition field. Set best rated value as 100 and worst as 10. Now a strategy that has a better result in the out of sample data period will rank higher in the score, depending on how important you rate this new evaluation criterium.
|