Support Board
Date/Time: Wed, 26 Feb 2025 01:58:55 +0000
[User Discussion] - Inconsistent back / live tests on two installs - despite using "Accurate back test mode&quo
View Count: 420
[2021-05-05 19:01:07] |
User963483 - Posts: 22 |
So I have two installs one on a laptop the other desktop. I'm testing an automated trading system LIVE. 1. On the desktop it took no trades yesterday. 2. At market close, I then ran a back test and no trades executed - which is fine or so I thought. 3. I opened up my laptop version and ran a back test. It took 3 trades during the RTH. 4. I deleted the desktop version and copied over the WHOLE 16GiB from laptop to desktop. 5. I ran a back test and it worked with 3 trades like #3. 6. Today, I left the desktop running and it took 4 trades, 3 longs and 1 short during RTH. 7. After markets closed, I closed the desktop version and opened the laptop one - they SHOULD be identical as per #4 except for update SCID. 8. After it downloaded missing data ie bring it back up to speed. I ran a replay same settings - technically I would think they should be the same. As the laptop just has to download the data since last night. 9. It took 2 trades in total, 2 longs vs #6. 10. I copied the SCID file from the desktop to the laptop and suddenly it takes 4 trades the same as #6. What is a major concern for me is the inconsistencies in the data file. I'm using CLEM21 which starts around 16 April with a VWAP Rolling study. 1. When I start a FRESH installation, I copy over the cht, defaultspread...scss and my twconfig file. 2. I setup the datafeed (cqg), open my chart which loads up CLEM21. Visually it "looks" fine starting with data around the 16th. 3. When I run a back test it shows roughly -60 ticks for the month. 4. When I copy the SCID from desktop or laptop the same back test shows around +500 ticks. Chart settings are 30 days and market depth 10. Why would the data be out by so much just a few hours later on two identical installs? Please help, what am I doing wrong? Date Time Of Last Edit: 2021-05-05 19:02:49
|
To post a message in this thread, you need to log in with your Sierra Chart account: