Support Board
Date/Time: Wed, 07 May 2025 14:47:05 +0000
Time unit significance for backtesting
View Count: 201
| [2024-11-22 16:08:30] |
| gtaranti - Posts: 94 |
|
Hi, I have an intraday 15-min chart with an ACSIL trading study. (Single chart chartbook) I have set the "Intraday Data Storage Tme Unit" in "Data/Trade Data Service Settings" to 1 min Do you think will have any difference changing this value to 30 sec, or lower (or even 1 Tick) in my backtest results? The idea was that due to the signals in the study being set only at the bar close (that is, every 15 min) I should NOT have different results. Is this correct? Or the smaller the time unit, the better the accuracy in any timeframe? Date Time Of Last Edit: 2024-11-22 16:09:03
|
| [2024-11-24 23:43:27] |
| Tony - Posts: 569 |
|
When you set the "Intraday Data Storage Tme Unit" in "Data/Trade Data Service Settings" to 1 min The replay will give you one snapshot every minute, which has a very unrealistic looking. I set it to 1 sec, which is close enough to real streaming data, set it to 1 tick will make the .scid file significantly larger, and will take much more computer resources during replay No comment on your study, since I don't know what it does. Date Time Of Last Edit: 2024-11-24 23:44:42
|
| [2024-11-25 20:05:23] |
| gtaranti - Posts: 94 |
|
Thanks, can you clarify this : "The replay will give you one snapshot every minute, which has a very unrealistic looking." I don't understand what the snapshot is? But again, on a 15-min bar chart, why should any "Intraday Data Storage Tme Unit" less than 1 min, have significantly different outcome? Date Time Of Last Edit: 2024-11-25 20:31:34
|
To post a message in this thread, you need to log in with your Sierra Chart account: