Specify Strikes, Offset, Min/Max Premium as Percentage of Underlying
under consideration
Brian Chia
Problem
For long-duration backtests and persistent live strategies, a fixed values for premium targets, spread premium targets, spread offsets, and entry premium filters can represent materially different trade setup and risks as the value of the underlying changes.
Suggested Solution
Add percentage-of-underlying options for strike selection and entry premium filters.
- Premium target based on Percentage of Underlying Price
- Spread Premium target based on Percentage of Underlying Price
- Child leg offset based on Percentage of Underlying Price
- Child leg Premium target based on Percentage of Underlying Price
- Child leg max offset based on Percentage of Underlying Price
- Min/max entry premium based on Percentage of Underlying Price
Brian Chia
updated the status to
under consideration
Brian Chia
Merged in a post:
Add a new "Use Min/Max Entry Premium as a % of Underlying" filter
D
Dario Luparello
The current filter uses absolute values, but that's problematic when backtesting across many years, as the underlying can change much. This is similar to using points vs delta, to choose strikes. The UI could simply have a dropdown where you can choose $ or % (similar to other UI elements)
Brian Chia
Merged in a post:
Add % Offset to the Parent Order Strike
I
Izzy
Add the option to offset the parent order by %.
For example if the 25 delta lands at 6500 for a call and I want the link/child to be 1% of the underlying away the BT will select 6565 as the strike for the linked order.
M
MrT
Agree, % of the underlying for min/max should give more accurate results for most trades. Otherwise the backtest/auto trades don't account for options pricing changes as the underlying moves up/down, especially over longer date windows.
Brian Chia
Merged in a post:
Strike (Width) selection based on % of Notional Size of Underlying
B
Ben
This would allow the trade to scale its risk profile to match the size of the underlying, which in the case of SPY has doubled since 2017. This would allow testing strategies on longer timeframes.
K
Kevin Smith
I think this is a great idea. Currently the available parent off sets are delta, parent offset and fixed premium. The SPX has gone from 4,000 to 7,000 since 0 DTEs launched. This means parent off set of 60 or 70 is VERY different for the start of the testing time frame (SPX=4000) versus more recent times (SPX=7000).
This problem makes back testing butterflies more difficult. I currently break the back test into 4,000 - 5,000, 5,000 - 6,000 and 6,000 - 7,000 ranges of time.
Percentage offset would help a lot here.