Rolling options explained: when to roll out, down and out, or up, why a credit roll is not a free fix, and how to compare rolling with closing in a QQQ example.
Summary: Rolling options means closing an open option and opening a similar one with a later expiration, a different strike, or both, in one decision. In the illustrative QQQ example, a short $680 put that is $1,646 underwater can be rolled down and out to a December $665 put for an $83 credit, but the new position beats simply closing only if QQQ finishes above about $642.68. Because a roll trigger tested on a window where dips recovered looks better than it is, the article ends with a hindsight-free testing routine that runs one roll rule on ten synthetic charts.
Last updated: 2026-10-20.
What Does Rolling an Option Mean?
Rolling an option is closing an existing option position and opening a new one on the same underlying with a later expiration, a different strike, or both. It's two trades entered together, usually as one order. Traders roll to buy time for a trade that moved against them, to move a strike out of the way, or to keep collecting premium on a position that worked.
The words describe the direction of the new option:
- Rolling out — same strike, later expiration.
- Rolling down and out — lower strike, later expiration (common for short puts after a drop).
- Rolling up and out — higher strike, later expiration (common for covered calls after a rally).
- Rolling for a credit — the new option brings in more than the old one costs to close; rolling for a debit is the opposite.
A roll is not a repair. The loss on the old option is real the moment you buy it back; the new option is a new position with its own risk. If you trade short puts, the cash-secured put covers the starting position used below.
The Snooze Button Analogy
Fig. 2 — Rolling buys time; it doesn't undo the move.
The alarm rings and you hit snooze. You get nine more minutes, but you're no closer to being out of bed, and each snooze makes the morning tighter. Rolling for a credit is a snooze button that pays you a little for pressing it — but the move that set off the alarm has already happened.
Key Facts
- Rolling an option closes the current option and opens a new one with a later expiration, a different strike, or both.
- In the illustrative example, a QQQ $680 put sold for $5.03 with 31 days left costs $21.49 to buy back ten days later after a 6% drop — a $1,646 loss if closed.
- Rolling it down and out to the December $665 put (49 days, $22.32) brings in a net credit of $83 and lowers the break-even from $674.97 to $659.14.
- The rolled position beats closing only if QQQ finishes December above about $642.68 — a further drop of about 4.4% from $672 makes the roll the worse choice.
- Below $665 every $1 drop costs the rolled position another $100, so at $600 it loses about $5,914.
- Rolling for a credit doesn't erase a loss; it keeps the original obligation open for longer, with cash-secured collateral of $66,500 still tied up.
- A roll rule backtested on a window where dips recovered measures that window; the trigger it suggests is the parameter to test forward.
When Should You Roll an Option?
Roll only when you would open the new position today on its own merits. Three situations usually qualify:
- Your view still holds, and the new option pays you for the time. A short put tested after a drop, when you'd still be willing to own the stock at the new strike.
- A short option that worked has little premium left. Buying it back cheaply and selling a later one keeps the strategy running with fresh time value.
- You want to avoid assignment or exercise. A covered call that went in the money before you're ready to sell the shares, or a deep in-the-money short put near expiration — in the real market, early assignment risk grows as time value disappears.
Close instead of rolling when your view has changed, when the roll only works for a debit you wouldn't pay for a new trade, or when the position is already larger than you'd open today.
A Worked Example: Rolling a Short Put Down and Out
The Invesco QQQ Trust (QQQ) closed at $718.36 on October 19, 2026, according to the QQQ page on StockAnalysis. The example below is illustrative: QQQ at $715, a 4% interest rate, dividends ignored (QQQ pays small quarterly dividends), Black-Scholes prices with one implied volatility per date. These are assumptions to show the mechanics, not quotes or recommendations.
Day 0 (October 20). You sell one November 20 $680 put with 31 days left at 22% implied volatility for $5.03, or $503. Its delta is −0.19, and the break-even is $674.97.
Day 10 (October 30). QQQ has fallen 6% to $672 and implied volatility has risen to 28%. With 21 days left, the $680 put now costs $21.49. Closing would lock in a loss of $1,646. These are the choices:
| Choice | New option | Net on the roll | Net so far | Break-even |
|---|
| Close | — | — | −$1,646 (loss) | — |
| Roll out | Dec 18 $680 put, $29.81 | +$833 | +$1,336 | $666.64 |
| Roll down and out | Dec 18 $665 put, $22.32 | +$83 | +$586 | $659.14 |
| Roll further out | Jan 15 $660 put, $25.95 | +$446 | +$949 | $650.51 |
Each roll shows a credit, and the cumulative credit is positive — which is exactly why rolling feels like a fix. But look at what you're holding. Rolling out keeps a put that's already $8 in the money. Rolling down to $665 gives up almost all the credit to move the strike $15. Rolling to January adds four more weeks of exposure. None of them changes the fact that QQQ is $8 below where you agreed to buy it.
Take the down-and-out roll. The new $665 put has a delta of −0.42 (versus −0.54 for the old one), so the position is less sensitive to the next dollar, but it now lasts 49 days instead of 21 and still ties up $66,500 of cash as collateral.
Roll or Close? The Number That Decides It
The fair comparison isn't "credit versus loss". It's the result of the rolled position at the new expiration versus the $1,646 you'd lock in by closing today.
Fig. 3 — The rolled position beats closing only above about $642.68 (illustrative QQQ example).
| QQQ on December 18 | Rolled to Dec $665 put | Closed on October 30 |
|---|
| $700 | +$586 | −$1,646 |
| $660 | +$86 | −$1,646 |
| $650 | −$914 | −$1,646 |
| $640 | −$1,914 | −$1,646 |
| $630 | −$2,914 | −$1,646 |
| $600 | −$5,914 | −$1,646 |
The roll comes out ahead above about $642.68 and behind below it. That's a 4.4% further drop from $672 — less than half of the roughly 10% one-standard-deviation move that 28% implied volatility prices over the seven weeks to December. Rolling doesn't remove the dependence on a recovery; it extends it. In my experience, the trades that hurt most weren't the ones closed at a loss, but the ones rolled three times "for a credit" into a trend that kept going.
Fig. 4 — A roll has to pass the same test as a new trade.
Common Mistakes When Rolling Options
- Rolling to avoid admitting a loss. If you wouldn't open the new position today, rolling into it is the same decision with worse framing.
- Rolling further and further out. Each roll adds time and keeps capital tied up; a short put rolled for months can block collateral that would serve a better trade.
- Rolling too late. The deeper in the money a short put gets, the less extra time value a later, lower-strike put carries, so a down-and-out roll may only work for a debit.
What a Backtest Can't Tell You About Rolling Options
Roll rules come with a trigger: "roll when the short strike is touched with at least 14 days left," "roll at 21 days," "roll only for a credit." Backtests of such rules often look excellent, and the reason is the window. In a period where every dip recovered, almost any roll down and out ends above its new break-even — the extra time is exactly what the market rewarded. The trigger that tested best is the one that fit how quickly the dips in that window bounced.
The example shows the dependence. The same October 30 roll ends December at +$586 if QQQ recovers to $700, at +$86 if it drifts to $660, and at −$2,914 if the drop continues to $630. The roll rule is identical in all three; only the path after the roll differs. A window full of V-shaped recoveries reports the first outcome. A window that includes one long decline reports the last one, often repeated through several rolls.
A backtest answers "what would have happened". Hindsight-free testing is testing a trading rule on a market chart that did not exist when you wrote the rule, so the rule cannot be bent to fit the past. For a roll rule, that means fixing the trigger, the new strike, the credit condition and the maximum number of rolls before the chart starts — forward testing on synthetic markets shows how the rule behaves when the dip doesn't come back. The reasoning is laid out in hindsight-free testing.
How to Practice Rolling Options Without Real Money
Pilots practice a go-around until it's automatic, so the decision is made before the runway gets close. Rolling is the same kind of decision: it should be written down before the price reaches your strike.
In the simulator you can roll by hand from the option chain: buy back the short option, then open the new one. Buying back a short always works; for a new put spread, buy the new long put first and then sell the new short put in the same expiration. A short put on its own is cash-secured (strike × 100 in cash), so the protocol below uses a one-contract bull put spread to keep collateral small. One difference from the real market: in the simulator an in-the-money short option is settled for cash at expiration, and there's no early assignment, so the roll is always your decision.
Run It Forward: 10 Charts, One Rule
- Step 1: Write the rule down first. Open a one-contract bull put spread in the expiration closest to 35 days, short put near 0.30 delta, long put two strikes lower. If the price closes below the short strike with at least 14 days left, roll to the next monthly expiration with the same width, choosing the lowest short strike whose new credit is at least the cost to close the old spread; if no strike gives that, close. Allow at most two rolls, and hold the last spread to the candle before it expires.
- Step 2: Generate a chart. In "Model Settings", choose "Essential 1 [ Steady Drift ]" with the "Random" scenario and a length of 6m.
- Step 3: Open the spread in the accepted order. Buy the long put first with the "B" button, then sell the short put with the "S" button in the same expiration.
- Step 4: Step forward candle by candle. When the trigger fires, write down the cost to close the spread, then buy back the short put first, sell the long put, and open the new spread long leg first — no edits to the rule after you see the chart.
- Step 5: Record four things. How many rolls happened, the final result in dollars, the worst drawdown, and the result you would have had by closing at the first trigger.
- Step 6: Repeat on 10 charts. Read the spread of outcomes, not the average, and count how often rolling beat closing at the first trigger.
FAQ
What does it mean to roll an option?
Rolling an option means closing an open option and opening a new one on the same underlying with a later expiration, a different strike, or both. In the example, a QQQ $680 put is bought back for $21.49 and a December $665 put is sold for $22.32 in one decision.
When should you roll an option?
Roll when you'd open the new position today on its own merits: your view still holds, the new option pays you for the extra time, and the size still fits your account. Close instead when your view has changed or the roll only works for a debit you wouldn't pay for a new trade.
Is rolling for a credit always better than closing?
No. A credit roll keeps the original risk open for longer. In the example, rolling the QQQ put down and out for $83 beats closing at a $1,646 loss only if QQQ finishes December above about $642.68; below that, the rolled position loses more.
What is the difference between rolling out and rolling down and out?
Rolling out keeps the same strike and moves to a later expiration. Rolling down and out also lowers the strike, which reduces the risk of a short put but usually brings in less credit. In the example, rolling out to the December $680 put collects $833, while rolling down to $665 collects $83.
Can you backtest an options rolling strategy?
Yes, and a backtest shows how a roll rule would have done in past years. But the trigger that tests best fits how quickly dips recovered in that window; in the example, the same roll ends anywhere from +$586 to −$2,914 depending only on the path afterward. Running a fixed roll rule on synthetic charts shows how it behaves when the dip doesn't come back.
Related: Cash-Secured Put, Covered Call, Bull Put Spread, Options Expiration Guide, Backtesting vs Simulated Data.
About the author: Miroslav Slavkovský is the founder of OptionsSimulator and a software developer focused on derivatives-pricing models. He built the platform's simulation engine — implementing Heston, SABR, GARCH, and Monte Carlo methods — so traders can practise options strategies on generated market scenarios rather than on historical charts.
Practice the Roll Decision Before You Need It
Want to see how often a roll beats simply closing — without risking real money? Open the simulator, sell a bull put spread on an "Essential 1 [ Steady Drift ]" chart with the "Random" scenario, and run your roll rule forward on a synthetic market that has never existed.
Open Simulator
Step-by-step guide
- Write the rule down first. Open a one-contract bull put spread in the expiration closest to 35 days, short put near 0.30 delta, long put two strikes lower. If the price closes below the short strike with at least 14 days left, roll to the next monthly expiration with the same width, choosing the lowest short strike whose new credit is at least the cost to close the old spread; if no strike gives that, close. Allow at most two rolls, and hold the last spread to the candle before it expires.
- Generate a chart. In "Model Settings", choose "Essential 1 [ Steady Drift ]" with the "Random" scenario and a length of 6m.
- Open the spread in the accepted order. Buy the long put first with the "B" button, then sell the short put with the "S" button in the same expiration.
- Step forward candle by candle. When the trigger fires, write down the cost to close the spread, then buy back the short put first, sell the long put, and open the new spread long leg first — no edits to the rule after you see the chart.
- Record four things. How many rolls happened, the final result in dollars, the worst drawdown, and the result you would have had by closing at the first trigger.
- Repeat on 10 charts. Read the spread of outcomes, not the average, and count how often rolling beat closing at the first trigger.
Frequently asked questions
What does it mean to roll an option?
Rolling an option means closing an open option and opening a new one on the same underlying with a later expiration, a different strike, or both. In the example, a QQQ $680 put is bought back for $21.49 and a December $665 put is sold for $22.32 in one decision.
When should you roll an option?
Roll when you'd open the new position today on its own merits: your view still holds, the new option pays you for the extra time, and the size still fits your account. Close instead when your view has changed or the roll only works for a debit you wouldn't pay for a new trade.
Is rolling for a credit always better than closing?
No. A credit roll keeps the original risk open for longer. In the example, rolling the QQQ put down and out for $83 beats closing at a $1,646 loss only if QQQ finishes December above about $642.68; below that, the rolled position loses more.
What is the difference between rolling out and rolling down and out?
Rolling out keeps the same strike and moves to a later expiration. Rolling down and out also lowers the strike, which reduces the risk of a short put but usually brings in less credit. In the example, rolling out to the December $680 put collects $833, while rolling down to $665 collects $83.
Can you backtest an options rolling strategy?
Yes, and a backtest shows how a roll rule would have done in past years. But the trigger that tests best fits how quickly dips recovered in that window; in the example, the same roll ends anywhere from +$586 to −$2,914 depending only on the path afterward. Running a fixed roll rule on synthetic charts shows how it behaves when the dip doesn't come back.