When to Roll Options vs Close: A Decision Framework

Roll a tested short option only if you can do it for a net credit and you'd happily open the new position today. Close it when the only roll available costs a debit, or when the reason you sold it is gone.

If you'd be fine owning the shares at the put strike, or selling them at the call strike, take assignment instead.

Four things decide which of the three it is: the roll credit, how much extrinsic value is left in the option you'd buy back, how many days are left, and whether you still want the stock. Each gets a section below, followed by three worked examples with hypothetical numbers.

What "tested" means

A short option is tested when the stock moves to or through the strike. For a cash-secured put, the stock has fallen toward your strike. For a covered call, it has risen toward it.

Tested isn't the same as lost. A put that's slightly in the money with weeks to go can still expire worthless if the stock recovers. It just means assignment is now a real outcome, so you need a plan for it.

The four checks

1. Can you roll for a net credit?

A roll is two trades done as one order: buy back the current option and sell a later one.

Net roll credit = price you get for the new option − price you pay to buy back the old one

Multiply by 100 for one contract. If the result is positive, you're being paid to wait longer.

If it's negative, you're paying for more time on a trade that's already going against you.

The roll-for-credit rule is simple: don't roll for a debit.

A debit roll means you're adding money to a losing position to avoid a result (assignment or a realised loss) that you could take today.

Roll out in time first. Only move the strike (down for a put, up for a call) if the roll still pays a credit after the strike change.

Enter the roll as a single spread order with a limit price. Two separate orders leave you exposed if the stock moves between fills.

2. How much extrinsic value is left?

An option's price is intrinsic value plus extrinsic (time) value.

Put intrinsic value  = max(strike − stock price, 0)
Call intrinsic value = max(stock price − strike, 0)
Extrinsic value      = option price − intrinsic value

The extrinsic value is what you can still earn by holding.

When a tested option has little extrinsic value left, holding it has little upside. It also becomes more likely that the buyer exercises early, because they give up almost nothing by doing so. The early assignment guide covers when that happens.

Low extrinsic value is the point where you have to choose. Roll, close, or let assignment happen. Waiting mostly just adds risk.

3. How many days are left?

More days on the new contract means more extrinsic value, so a larger roll credit. That's why the credit rule tends to push each roll further out.

The trade-off is time. Every roll ties up the same capital (or the same shares) for longer, and the stock can keep moving against you. Near expiration there's a second issue: gamma is highest, so small moves in the stock cause big swings in the option's price. A decision you put off in the last week can get more expensive by the day.

4. Do you still want the underlying?

Ask one question: would I open this exact position today, with fresh money, at this strike?

If yes, rolling or taking assignment are both reasonable. If no, close.

A roll on a stock you no longer want is just a way of avoiding a realised loss. The credit doesn't change the thesis.

For a covered call the question flips.

Are you fine selling the shares at the strike? If you are, assignment is the trade working as intended.

One more check before any roll: position size. If the position ties up more of your buying power than you'd put into one new trade, closing it reduces that concentration.

The decision in one table

Worked examples (hypothetical)

These numbers are made up to show the arithmetic. Real quotes depend on the stock, IV and bid-ask spread.

Example 1: tested put, roll out for a credit

Say you sold a $50 put for $1.20. The stock falls to $47 with 8 days left.

• Buying the put back costs $3.20. Intrinsic value is $50 − $47 = $3.00, so extrinsic is $0.20.
• The $50 put five weeks further out is bid at $4.10.
• Net roll credit = $4.10 − $3.20 = $0.90, or $90 per contract.

Total credit collected is now $1.20 + $0.90 = $2.10, so your breakeven if you're assigned on the new put is $50 − $2.10 = $47.90.

Rolling down to the $48 put in the same later expiration would pay, say, $2.60. That's $2.60 − $3.20 = a $0.60 debit, which fails the rule. So the roll is out, not down.

The alternative is to hold and let it assign at expiry. Your cost basis would be $50 − $1.20 = $48.80 on a stock trading at $47. If you want the shares anyway, that's a legitimate choice. Rolling earns more premium but keeps the cash tied up for five more weeks with the same assignment risk at the end.

Example 2: tested covered call, let it assign

Say you own 100 shares bought at $50 and sold the $55 call for $1.00. The stock is at $57 with 4 days left.

• Buying the call back costs $2.10: $2.00 intrinsic ($57 − $55), $0.10 extrinsic.
• Rolling out to next month's $55 call at $3.00 would be a $0.90 credit.
• Rolling up and out to a $57.50 call at $1.60 would be a $0.50 debit.

With only $0.10 of extrinsic value left, buying the call back means paying almost all intrinsic value for the right to keep the upside. If you're fine selling at $55, let it assign. You sell at $55 and keep the $1.00, so you get $56 per share, a $6 gain per share over your $50 purchase ($600 per contract).

The credit roll out to the same strike is fine if you'd rather keep the shares. It keeps your upside capped at $55 for another month.

Example 3: broken thesis, close

Say you sold a $50 put for $1.20 and, three weeks later, the company cuts guidance. The stock drops to $44. The put costs $6.40 to buy back and next month's $50 put bids $6.90.

On paper that's a $0.50 credit roll.

It passes the credit rule and fails the fourth check: you wouldn't sell this put today. Close it.

The loss is $6.40 − $1.20 = $5.20 per share, or $520 per contract. The credit rule is necessary for a roll, but it isn't enough on its own.

What our data shows about closing early vs holding

I built a research job that runs the same wheel backtest two ways and stores the results. One arm runs the full wheel: it holds each short put to expiry and, if assigned, sells covered calls until the shares are called away. The other buys each short put back at 21 DTE and opens a new one. Both use 35 DTE entries 2% out of the money, from 20 September 2024 to 18 September 2026. The run dated 20 September 2026 published 40 of 45 tickers (the other 5 missed the publication thresholds).

This isn't a study of rolling tested positions.

The managed arm closes every put at 21 DTE, winners and losers alike, and then opens a new one. It says something about "close early vs hold", which is half of the decision in this article. Every ticker in the run also used some modelled pricing (a Black-Scholes fallback where historical option prices were missing), so treat these as modelled backtest results.

With that caveat, in the simulated backtest:

• Total return was lower with the 21-DTE rule on all 40 tickers. The median difference was −20.7 percentage points.
• Max drawdown was lower on 38 of 40. The median difference was −8.2 percentage points.
• Assignment rate was lower on all 40, by a median of 32 percentage points. The managed arm was never assigned, because it always closed before expiry.

So the managed arm never held shares or sold a covered call: in practice the study compares a put-only strategy with the full wheel.

Over this window 32 of the 40 tickers rose on a buy-and-hold basis. A rising market favours an arm that takes assignment and holds shares, so the return gap may be smaller in a flat or falling market. The gap didn't go away on the 8 tickers that fell, though. The rule still lost on all of them, with a median gap of −17.4 points against −23.6 for the 32 that rose.

For the tested-position decision, the takeaway is narrow.

On a wheel, closing early to avoid assignment cost return in this modelled test, and it bought a smaller drawdown. If you run a wheel and you want the shares, assignment is part of the plan. Closing or rolling just to dodge it has a cost.

Runs are published on the 21-Day Rule research page, and the 21 DTE rule article goes through the rule itself. Later runs can differ from the dated run above.

Puts vs covered calls

Tested puts. Rolling out and down often pays a credit when IV has risen with the selloff, and assignment means buying a stock you chose to sell puts on. The risk is a stock in a real decline, where each roll adds time without changing the outcome. The rolling tested puts and rolling cash-secured puts guides go deeper.

Tested covered calls. Rolling up and out often costs a debit, because you're buying back an in-the-money call and selling a lower-priced one further from the money. In the wheel, assignment on the call is the normal exit. See rolling covered calls and how to adjust covered calls when tested.

Common rolling mistakes

• Rolling for a debit. You pay more to stay in a trade that's already against you.
• Rolling without asking the fresh-money question. A credit on a stock you no longer want is still a position you no longer want.
• Losing track of total credits. After several rolls, your real breakeven is the strike minus every credit and plus every debit. The cost basis after rolls article shows how to track it.
• Rolling out forever. Each roll pushes the decision further away. At some point closing or taking assignment is the cleaner answer.
• Waiting in the last week. With little extrinsic value left, there's not much to gain from holding and the gamma risk is at its highest.

Compare later expirations before you roll

A roll out is only as good as what the later expiration pays. The widget pulls current cash-secured put quotes for a ticker at weekly, monthly and quarterly expirations.

To see how assignment risk on a single put changes with the stock price and the days left, use the stress test below.

Once a position is closed, the capital needs a new home. The scanner demo below uses sample data, not a real account.

The bottom line

Roll only for a net credit and only into a position you'd open today. Close when the roll needs a debit or the thesis is broken. Take assignment when you want the shares at the put strike or are fine selling them at the call strike. On a wheel, our modelled backtest suggests that avoiding assignment has a real cost in return, so treat assignment as a normal outcome, not a failure.

Related Articles

• The 21 DTE Rule Explained: When and Why to Close Options Positions Early
• Rolling Tested Puts: When and How to Manage Challenged CSP Positions
• Can Cash-Secured Puts Be Assigned Early? How to Roll
• Rolling Covered Calls: When and How to Extend Positions
• How to Adjust Covered Calls When Tested
• True Cost Basis After Rolls
• Early Assignment in Options
• Gamma Risk Near Expiration
• Options Risk Management: Position Sizing and Loss Controls
• The Wheel Strategy: The Complete DTE Guide