Rolling: Buying Time, and What It Costs
Close one leg, open another. Watch the net credit, the new breakeven and the blended return update.
Worth reading first: Cash-Secured Puts, From Cash to Assignment, Theta: Watching Time Value Bleed Out
Rolling is two trades executed as one decision: buy back the contract you sold, and sell another one further out in time, often at a different strike. It is the standard response to a short option that has gone wrong, and it is surrounded by more muddled thinking than any other move in premium selling.
The muddle is this: rolling feels like it fixes a losing position, because the account never shows a realised loss. It does not fix anything. It converts a loss you would have taken today into a loss you have financed with time. Sometimes that is exactly right. Sometimes it is how a small problem becomes a six-month problem.
Roll a put that has gone against you
You sold a $100 put for $250 and the stock has fallen. Choose the replacement strike and expiration, and watch the net ticket, the new breakeven and the blended annualised return respond.
You sold the $100 put when it was $100
Closing leg
$100 put · 5 days left
- Credit at open
- +$250
- Cost to buy back
- −$796
Replacement leg
$95 put · 30 days
- New credit
- +$493
- New breakeven
- $90.07
This roll costs more to close than the new leg brings in. That is a debit roll — it can still be the right decision, but it is a decision to pay for time, not a way to avoid a loss.
The three directions you can roll
Rolling out keeps the strike and moves to a later expiration. You are buying time and nothing else. Because the further-dated contract has more extrinsic value, this almost always collects a net credit.
Rolling down and out moves to a lower strike and a later date. You are improving your breakeven — genuinely reducing the price at which you would be assigned — and paying for it with premium. Often this is a small net credit or a small debit.
Rolling up and out is the covered-call equivalent: a higher strike, later date, used when a stock has rallied through your call and you want to keep the shares. It usually costs a debit, because you are buying back an in-the-money call.
| Roll choice | New credit | Net ticket | New breakeven | What you achieved |
|---|---|---|---|---|
| Out — $100, +30 days | $900 | +$90 | $90.60 | Bought a month, tiny credit |
| Down and out — $95, +30 days | $560 | −$250 | $91.05 | Better strike, paid for it |
| Down and out — $90, +45 days | $390 | −$420 | $94.20 | Much better strike, expensive |
| Do not roll — take assignment | — | — | $97.50 | Own 100 shares, start the wheel |
A $100 put sold for $250, now 5 days out with the stock at $92. Buying it back costs about $810.
The bottom row is the option everyone forgets to price. Taking assignment is not a failure state — on a stock you wanted, it is the strategy working as designed, and it converts a managed option position into a simple share position you can write calls against.
The arithmetic that actually matters
A roll is judged on two figures, and both are in the widget above.
Cumulative net is everything the chain has collected and paid since the first leg opened. This is the true P&L of the position, and it is the number that rolling can quietly obscure — each individual ticket can look like a credit while the cumulative figure sinks.
Cumulative net = all credits collected − all buybacks paid − feesA roll never resets this. Every buyback stays in the total.
Blended annualised return weights each leg by its own capital and its own holding period. It answers the question that actually decides whether rolling was worth it: for all the capital I tied up, across all the days I tied it up, what did this chain earn?
A chain of six rolls that ends flat has not “broken even”. It has spent six months of capital to earn nothing, and that capital had alternatives.
When rolling is genuinely the right move
You still want the stock, and you want a better price. Rolling down and out moves your breakeven meaningfully lower. If you would rather own at $91 than $97.50, and the debit is small, that is a real improvement bought at a fair price.
The credit genuinely compensates for the time. Compute the annualised return on the new leg alone. If it is competitive with a fresh trade you would open today, rolling is fine. If it is not, you are financing an old mistake at a bad rate.
You want to avoid assignment for a specific reason. Not enough cash, an unwanted tax event, an upcoming ex-dividend date on a short call. These are concrete reasons and rolling addresses them directly.
When it is not
To avoid realising a loss. The loss already exists; only its recognition is being deferred. Deferring it costs weeks of capital.
When the reason you sold the put has changed. If the company’s situation is genuinely worse, rolling down and out means committing more time to a thesis you no longer hold. Closing is cheaper than being slowly right.
Repeatedly, on the same position. Two rolls is management. Five rolls is a position that has become a relationship. Look at the cumulative net, not the last ticket.
What can go wrong
Double-counting the buyback. When the original leg’s buyback is charged to the chain and the replacement’s credit is recorded gross, it is easy to count the same cash twice and believe the chain is doing better than it is. Premium Tracker links a replacement to the leg it replaced specifically so this cannot happen.
Losing the thread across many rolls. By the fourth roll, few people can state their true cumulative P&L on the chain from memory. This is precisely where a spreadsheet quietly starts lying.
Rolling into an earnings date. Extending thirty days can put a scheduled event inside your new expiration that was not inside the old one.
Key takeaways
- Rolling is buying back one leg and selling another — it defers an outcome, it does not undo one.
- Rolling out buys time; rolling down and out buys a better breakeven and usually costs a debit.
- A credit ticket is not sufficient justification. Judge the new leg’s annualised return as though it were a fresh trade.
- Cumulative net across the whole chain is the real P&L. Individual tickets can look positive while it falls.
- Taking assignment is a legitimate choice, not a failure — on a stock you wanted, it is the strategy working.
Check your understanding
1. You buy back a put for $810 and sell a new one for $900. What did this roll accomplish?
2. A roll collects a $90 credit for 30 more days on $9,500 of capital. How should you judge it?
3. Which is the strongest reason to roll rather than take assignment?
Roll without losing the thread
Premium Tracker links a roll to the leg it replaced, so the buyback is charged once and the chain's true P&L survives across every roll.