Support Board
Date/Time: Thu, 07 May 2026 19:26:32 +0000
Post From: Multi-Year Backtest Performance and Simulation Account Issues
| [2026-05-07 09:51:11] |
| User499597 - Posts: 4 |
|
Hi, I’m encountering issues with the auto-trade bar-based backtest when running longer periods (10+ years). Context: • A separate simulation account is required for each year because processing all years together takes too long. • When loading 10+ years at once, after 2 years the backtest speed drops by roughly 50%, and it continues slowing down as the test progresses. However, this may be related to ACSIL complexity and bar counting, but I’m not sure. • A simulation account resets automatically when starting a new backtest, so a single account cannot be used across multiple years. • Drawdown statistics are incorrect because they reflect only the individual simulation accounts instead of the combined multi-year performance. Questions: 1. Is it possible to run a multi-year backtest using a single simulation account while the years are split/loaded after each other? 2. Is it possible to disable chart visualization and everything unrelated to the backtest in order to improve speed and avoid deceleration? 3. After more than 10 simulation accounts for each year, the trade log order becomes inconsistent (e.g. the 10th year appears after the 2nd). How can this be solved? Thank you. |
Private File |