All you need are trades with these characteristics:
1) pWin >= 0.5 (the more the better; of course this can theoretically go up to 1.0)
2) reward >= risk
The art is to determine these factors for "the most probable area" on the scale
(ie. the area around the initial underlying spot on the x-axis) --> expectancy.
IMO a good choice is taking the abs pct dist of -1SD (or upto -2SD) also for the other side (this is very unintuitive/unusual).
Of course also important:
3) know that at expiration, IV and volume are irrelevant...
4) risking only 1% to max 5% of AUM per such trade (ie. need to have about 20 to 100 such trades at the same time for being fully invested).
5) use spreads or spread-like constructs (ie. where loss is capped); if early assigned, just let it be, don't fight it with stops etc
...based on research/study/sims, not yet real-world tested.
1) pWin >= 0.5 (the more the better; of course this can theoretically go up to 1.0)
2) reward >= risk
The art is to determine these factors for "the most probable area" on the scale
(ie. the area around the initial underlying spot on the x-axis) --> expectancy.
IMO a good choice is taking the abs pct dist of -1SD (or upto -2SD) also for the other side (this is very unintuitive/unusual).
Of course also important:
3) know that at expiration, IV and volume are irrelevant...

4) risking only 1% to max 5% of AUM per such trade (ie. need to have about 20 to 100 such trades at the same time for being fully invested).
5) use spreads or spread-like constructs (ie. where loss is capped); if early assigned, just let it be, don't fight it with stops etc

...based on research/study/sims, not yet real-world tested.
Last edited:
