I'm interested in playing around with the Self-Adaptive trading feature. If it generates improved inputs but there's a pre-existing position, how does it handle that position? Will it close it if the new inputs deem we should be flat? Will it reverse it if it deems we should be on the other side of the trade? And will it open a position if we are flat but it deems we should have a position (or does autotrading turn off in this case)?
Also, is there any way to simulate Self-Adaptive trading in backtesting?
Thank you
Self-Adaptive trading and positions
- Tammy MultiCharts
- Posts: 200
- Joined: 06 Aug 2020
- Has thanked: 6 times
- Been thanked: 65 times
Re: Self-Adaptive trading and positions
Hello Mr. Beeblebrox,
The position may only be closed or, if you are flat, opened by new orders generated by your strategy after the new inputs are applied.
Please find more information on how Self-Adaptive Trading works here.
If new inputs are generated and applied to the strategy, the current position will be saved, but all the active orders generated by the strategy before that (with the old inputs) will be cancelled.If it generates improved inputs but there's a pre-existing position, how does it handle that position?
The position may only be closed or, if you are flat, opened by new orders generated by your strategy after the new inputs are applied.
Please find more information on how Self-Adaptive Trading works here.
Self-Adaptive Trading was designed specifically for optimizing the strategy during Auto Trading. It's not possible to imitate it in Backtesting. You can, however, optimize your strategy on the set of data plotted on the chart before going live.Also, is there any way to simulate Self-Adaptive trading in backtesting?
Re: Self-Adaptive trading and positions
Hello,
Since you already implemented the ability for MC to change inputs while in auto-trading (with self-adative), it would be awesome if one could do the same... manually !
Thanks for listening.
Since you already implemented the ability for MC to change inputs while in auto-trading (with self-adative), it would be awesome if one could do the same... manually !
Thanks for listening.
Re: Self-Adaptive trading and positions
Ok, thank you for the information (and sorry for the delayed response). It sounds like I could essentially produce relevant backtesting results by continually changing the amount plotted on the chart and keeping track as it moves forward. The optimization seems like it could be a very powerful feature in some circumstances. It would be great if it could be backtested in a future version of MC. It would probably be very resource intensive but maybe if you already had a base strategy you were happy with and set a relatively small range for the inputs in optimization. Probably a good job for a powerful virtual server or DaaS. Thank youHello Mr. Beeblebrox,
If new inputs are generated and applied to the strategy, the current position will be saved, but all the active orders generated by the strategy before that (with the old inputs) will be cancelled.If it generates improved inputs but there's a pre-existing position, how does it handle that position?
The position may only be closed or, if you are flat, opened by new orders generated by your strategy after the new inputs are applied.
Please find more information on how Self-Adaptive Trading works here.
Self-Adaptive Trading was designed specifically for optimizing the strategy during Auto Trading. It's not possible to imitate it in Backtesting. You can, however, optimize your strategy on the set of data plotted on the chart before going live.Also, is there any way to simulate Self-Adaptive trading in backtesting?
Last edited by ZaphodB on 20 Nov 2021, edited 1 time in total.
Re: Self-Adaptive trading and positions
You can somehow simulate that with the Walk-Forward optimization, but it lacks OOS aggregated reports.It would be great if it could be backtested in a future version of MC. It would probably be very resource intensive but maybe if you already had a base strategy you were happy with and set a relatively small range for the variables in optimization.