Strangle tested on one side: roll or close?
Defending a short strangle
The stock has run to one of your short strikes. Rolling can buy time and more credit, but it keeps the risk on; closing ends it at a known loss. Decide the stop before you need it.
Set the stop before you sell: close at a loss of about 2× the credit. Before that, roll the untested side closer for credit, and roll out in time only for a net credit. Never roll just to avoid taking a loss.
Roll
Best for
A move that stalls or reverses, with enough time left to collect the extra credit.
What you get
More credit and wider breakevens on the tested side, but the risk stays undefined, often larger, for longer.
Close
Best for
A move that keeps going, or an event ahead you would not hold through.
What you get
A known, limited loss now, and the capital free for the next trade.
Side by side
| Roll | Close | |
|---|---|---|
| What you do | Roll the untested side closer, or the tested side out in time | Buy both sides back |
| Credit | Adds to it | None: you pay to close |
| Risk afterwards | Still undefined, often larger | None |
| Works when | The move stalls or reverses | The move continues |
| Cost | More time in the trade; more gamma near expiry | Realises the loss |
| Our backtest | Rolls when the price is within 1% of a strike, a few times at most | Stops at 2× the credit; exits at 21 days left |
Rolling, done right
- Roll the untested side toward the price only for a net credit, and never past the tested strike (that inverts the strangle).
- Roll the tested side out in time only if the new position pays a net credit; otherwise you are paying to keep the risk.
Closing, done right
- Write the stop down when you open the trade: about 2× the credit is a common one.
- Close before earnings or other events you would not sell a new strangle into.
Payoff sketches are at expiry on a $100 stock, for shape only. Education, not advice: disclaimer.