What should I do if I no longer want to own a stock I sold a put on?

Treat it as a position to exit, not a trade to repair. The premium you already collected is banked and irrelevant to the decision in front of you; the only live question is what it costs to stop being exposed to a company you've decided against. While the put is still out-of-the-money, closing it is usually cheap and there's little to think about. Once it's in-the-money, closing costs real money, and that cost is the price of information you didn't have when you sold — not a reason to keep holding.

Can I always roll a cash-secured put for a credit?

No, and this is the part that surprises people. Rolling down-and-out for a credit is straightforward while a put is out-of-the-money with plenty of time left. Once a low-delta put is meaningfully in-the-money, the down-and-out rolls all turn into net debits, because the strike you want to move to is worth far less than the one you're stuck with. In a modelled 100-strike put with the stock at 88, every sensible down-and-out roll cost between $188 and $534 to execute. The credit rolls that remain all keep the original strike and simply sell more time.

Is it better to buy back the put or take assignment and sell the shares?

It turns almost entirely on how much extrinsic value is left. Deep in-the-money and close to expiry, an option is nearly all intrinsic value: in a modelled position at 7 days to expiry, buying back cost $1,206 against $1,200 for taking assignment and selling the shares. At that point the two are economically the same, and the small difference buys certainty instead of a weekend of gap risk. With more time left, the extrinsic value is larger, and closing early means paying it to remove the remaining exposure.

Does the premium I already collected matter when deciding to close?

No, and treating it as though it does is the most expensive habit in this situation. The credit is already yours and it does not change what the position costs you from here. In a modelled example the credit was $37 and closing later cost $700 to $2,495 depending on how far the stock had fallen. The $37 has no bearing on whether paying $700 is sensible. Anchoring a four-figure decision to a two-figure credit is how a small position becomes a large one.

← Writing

General Option Selling

You Don't Want the Stock Any More. A Credit Roll Is Rent, Not Income.

25 July 2026 · 9 min read

Sell a cash-secured put, then decide you don't want to own the business after all, and you are no longer running an income trade. You're holding a leveraged commitment to buy something you've talked yourself out of, and the only useful question is what it costs to stop. The awkward part is that the standard advice — roll it down-and-out for a credit — mostly stops being available at the exact moment you need it.

The trade changed even though you didn't do anything

Two things go wrong on a wheel position, and they don't deserve the same response.

The first is that the price moved and nothing else did. Ordinary. You named a price you were happy to pay, the market came to it, and assignment is the plan working, not a malfunction.

The second is that you changed your mind about the company. Guidance cut, a competitor did something that matters, you read the last filing properly, or you finally admitted the position was bigger than your understanding of it. The strike is no longer a price you want to pay. It's a price you're obliged to pay.

The wheel has a well-rehearsed answer for the first case and almost nothing to say about the second. Nearly everything recommended for a tested put quietly assumes you still want the shares. Once that assumption goes, the mechanics stop being management and start being avoidance.

The arithmetic nobody puts in the diagram

A modelled position: a 100-strike put sold at roughly 0.08 delta, 30 days to expiry, stock at 112, implied volatility 30%. That collects $37 for one contract against $10,000 of collateral. Breakeven $99.63, maximum loss $9,963 if the stock goes to zero. The usual shape.

Then the thesis breaks and the stock falls. Volatility rises on the way down, as it tends to. Here's what buying that put back costs with 21 days still on the clock:

Stock	Put price	Cost to close	Delta	Extrinsic left
95	$7.00	$700	−0.66	$2.00
88	$12.90	$1,290	−0.81	$0.90
75	$24.95	$2,495	−0.96	~$0

Look at the ratio first. You were paid $37 and it now costs $700 to $2,495 to walk away. The credit was the fee for taking the position on, never a measure of its size.

The delta column is the more useful number, though. At −0.96 the option behaves almost exactly like being long 96 shares: no meaningful optionality, and effectively no premium left to decay. You're not holding an income trade that's gone temporarily offside. You're holding stock, with extra steps, in a company you've decided against.

The credit roll you were promised isn't on the menu

Stay with the stock at 88, where closing the 100-strike put costs $1,290. The conventional move is to roll down-and-out — lower strike, later expiry, collect a credit. Here's what the market actually offers.

Roll to	Sell for	Net on the roll	New delta
95 strike, 49 days	$1,102	−$188	−0.60
90 strike, 49 days	$791	−$499	−0.49
90 strike, 90 days	$1,019	−$271	−0.46
85 strike, 90 days	$756	−$534	−0.38
85 strike, 180 days	$1,092	−$198	−0.37

Every one is a net debit. Not marginally: between $188 and $534, on a position that collected $37 to begin with.

The reason is mechanical rather than mysterious. You're trying to swap an obligation to buy at 100 for an obligation to buy at 85 or 90. The market prices that improvement honestly and charges you for it. Premium on the lower strike isn't large enough to cover closing the higher one, and no amount of extra expiry fixes that while the stock sits below both.

So "only roll for a credit" is sound guidance with a consequence people skip past: once you're meaningfully in-the-money on a low-delta put, the credit condition usually can't be met by a down-and-out roll at all. The rule doesn't tell you which roll to pick. It tells you not to roll.

The only credit rolls left keep the thing you're trying to get rid of

There are credit rolls available at 88. They all share one feature.

Roll to	Credit	New delta	Days you stay committed
100 strike, 49 days	+$170	−0.69	49
100 strike, 90 days	+$369	−0.62	90
100 strike, 180 days	+$692	−0.54	180
100 strike, 365 days	+$1,133	−0.45	365
105 strike, 49 days	+$566	−0.77	49

To collect a credit you have to keep the strike. You're being paid $170 to stay obliged to buy at 100, for another 28 days, a stock trading at 88 that you've decided you don't want. Extend to a year and the market pays $1,133 for the same commitment. The last row is worse still: a higher strike pays more because you've agreed to pay more.

I don't call that income. It's rent. The market is charging you to store a decision you've already made and haven't acted on, and it will keep charging for as long as you keep paying. A credit for extending an obligation you regret is not the same animal as a credit for taking one on, though your account statement can't tell them apart.

It does cut delta, and I'd rather say so than pretend otherwise: −0.81 to −0.45 in the one-year case. That part is real. But look at the price of it. Eleven more months of exposure to a company you've written off, and $10,000 of collateral immobilised the whole time. Reducing the sensitivity of a position you want rid of is not the same as being rid of it.

Close it, or own it and sell the shares?

If closing is the direction, there are two routes out and the choice turns on extrinsic value. Buying the put back costs intrinsic value plus whatever extrinsic remains. Letting it assign costs intrinsic only, since extrinsic decays to nothing at expiry — but you have to survive to expiry to collect that saving.

The gap narrows fast. In the modelled position with the stock at 88 and 7 days left, the put costs $12.06 to close, or $1,206. Take assignment instead and you buy 100 shares at $100 and sell them at $88: a $1,200 loss. The difference is $6, which is the extrinsic value. Paying $6 to be finished today rather than carrying an unwanted position through a weekend and an opening auction looks obviously worth it to me. Assignment fees and the spread you'll cross on the shares tilt it further.

With 21 days left the extrinsic is $0.90, so the same choice costs $90. Now it's a genuine decision rather than a rounding error: $90 to end three weeks of exposure, against $90 of premium you could let decay while staying exposed to whatever changed your mind.

Either way, this isn't a choice between losing money and not losing money. Both branches realise a loss of roughly $1,200. All that's on the table is how much you'll pay for certainty, and how long you'll stay exposed.

What I do

My rule is short, and it's a rule precisely so I don't have to relitigate it while I'm annoyed.

Separate the two decisions. Do I want to own this business at this price? That gets answered on the business, not on the option chain and not on the size of my current loss. Only once it's answered do I look at mechanics.

If the answer is no, the position gets closed, and the credit I collected isn't part of the maths. The $37 is banked and spent. Whether paying $700 is sensible has nothing to do with it. The most expensive thing I've done in this situation is let a small credit anchor me to a large exposure, and I'd rather name the habit than keep paying for it.

I don't roll a broken thesis, at any credit. If the only credit on offer keeps my strike, the roll is postponement with a fee attached. Postponement has never once improved a company's fundamentals.

If I still want the business, I take the assignment and stop fiddling. It becomes an equity position with a cost basis, and the next question is whether to sell calls against it, which is a different and much calmer conversation.

The case against my own argument

Two things could make all of that wrong, and both are worth checking before acting on it.

A thesis break is easy to confuse with a drawdown. A stock down 12% produces exactly the feeling of having been wrong, whether or not you were. If nothing has actually changed at the company and you can't name what's different beyond the price, you haven't had a thesis break. You've had a bad fortnight, and closing at a loss is an expensive way to buy relief. My test is whether I can state the new information in one sentence without mentioning the share price.

The second one cuts the other way. A cheap price is sometimes the market re-rating the business correctly, which makes the case for exiting stronger rather than weaker. If the fall is the market pricing in deterioration you hadn't seen, then the exposure you're being paid rent to keep is exposure to more of it.

Where this fits

The screener at levelbox.ai is built around one question — would you own this business at this strike — and everything above is what happens when the answer changes after you've committed. That's why the chance of assignment sits next to the premium on every candidate, computed from the option maths rather than borrowed from delta, and why the size of the commitment is shown before you make it. It cannot tell you whether your thesis has broken. Nothing can. It can make the commitment visible while you still have a choice about it, which is the only part of this that's fixable in advance.

On the numbers above. Every figure is computed from this project's Black-Scholes engine on a stated hypothetical, not taken from a live chain: strike 100, the stock price and implied volatilities as labelled, a 4% risk-free rate, and no dividend. Black-Scholes assumes European exercise, so the model prices deep in-the-money puts a shade below intrinsic value; a real American put won't trade there, because it can be exercised early. On a dividend payer the put would be worth more than modelled, and early assignment becomes a live consideration around the ex-dividend date. Treat the numbers as the shape of the problem, which is what they're for, and check your own chain before acting on any of it.

This is analysis and education, not investment advice, and none of it is a recommendation to buy, sell, or close anything. levelbox.ai is not a broker and does not place trades. What you do with a position you've changed your mind about is your judgement to exercise, on your own account.

Common questions

What should I do if I no longer want to own a stock I sold a put on?: Treat it as a position to exit, not a trade to repair. The premium you already collected is banked and irrelevant to the decision in front of you; the only live question is what it costs to stop being exposed to a company you've decided against. While the put is still out-of-the-money, closing it is usually cheap and there's little to think about. Once it's in-the-money, closing costs real money, and that cost is the price of information you didn't have when you sold — not a reason to keep holding.
Can I always roll a cash-secured put for a credit?: No, and this is the part that surprises people. Rolling down-and-out for a credit is straightforward while a put is out-of-the-money with plenty of time left. Once a low-delta put is meaningfully in-the-money, the down-and-out rolls all turn into net debits, because the strike you want to move to is worth far less than the one you're stuck with. In a modelled 100-strike put with the stock at 88, every sensible down-and-out roll cost between $188 and $534 to execute. The credit rolls that remain all keep the original strike and simply sell more time.
Is it better to buy back the put or take assignment and sell the shares?: It turns almost entirely on how much extrinsic value is left. Deep in-the-money and close to expiry, an option is nearly all intrinsic value: in a modelled position at 7 days to expiry, buying back cost $1,206 against $1,200 for taking assignment and selling the shares. At that point the two are economically the same, and the small difference buys certainty instead of a weekend of gap risk. With more time left, the extrinsic value is larger, and closing early means paying it to remove the remaining exposure.
Does the premium I already collected matter when deciding to close?: No, and treating it as though it does is the most expensive habit in this situation. The credit is already yours and it does not change what the position costs you from here. In a modelled example the credit was $37 and closing later cost $700 to $2,495 depending on how far the stock had fallen. The $37 has no bearing on whether paying $700 is sensible. Anchoring a four-figure decision to a two-figure credit is how a small position becomes a large one.