Guide

Rolling a covered call, measured on real quotes

Rolling a covered call means buying the short call back and selling a new one further out. Replayed on real bid and ask since 2022, when a 25-delta call on SPY was in the money ten days before expiry, rolling up to a new 25-delta call cost a net debit all 27 times (1.62% of the share price on average) and still finished ahead of being called away, +4.71% against +4.21%. It did not beat just holding the shares (+6.29%).

Short answers

How do you roll a covered call?

You buy back the call you sold and sell a new call on the same shares at the same time, usually with a later expiry and often a higher strike. Moving only the date is rolling out; moving the strike up as well is rolling up and out. What you pay for the old call against what you collect for the new one is the net debit or credit. On SPY in our replay, rolling up from a call in the money cost a net 1.62% of the share price on average, and rolling out at the same strike paid a net 0.91%.

Should I roll a covered call or let it get called away?

In our replay, rolling finished ahead of being called away on every ticker tested, because the stock had already gone up and the roll kept the shares for the next leg while the called-away cash sat idle. It cost more risk: rolling up had a worst case of about -3.5% to -4% of the share price on SPY, QQQ, AAPL and NVDA, against -0.5% to +2.8% for being called away. Neither beat holding the shares with no call.

Does rolling a covered call up and out cost money?

Almost always, at the start. The old call is in the money and costly to close, and a new call at a higher strike pays less than that. The roll up was a net debit on 27 of 27 rolls on SPY, 29 of 29 on QQQ, 31 of 31 on AAPL and 24 of 25 on NVDA. Rolling out at the same strike was a net credit every time, because the new call has more time value, but the shares were called away again 70% to 90% of the time.

What rolling does

You sell a covered call and the stock rallies through the strike. You can let the shares be called away, or roll: buy the call back and sell another. Rolling out keeps the strike and pushes the date; rolling up and out also raises the strike so the shares can climb further before they are taken. Either way you are choosing to stay in the shares. The guide to covered calls covers the basics; this page measures the roll.

A roll is a trade of its own, with a bid and an ask on each side. The call you buy back is in the money, so it is priced near its intrinsic value plus a spread, and the one you sell is further out and pays for the time you add.

How we tested it

We replayed 108 entry days per ticker on SPY, QQQ, AAPL, NVDA and TSLA from April 2022. Each entry bought 100 shares and sold the call near 25 delta about 37 days out at the bid. Ten calendar days before that expiry, if the stock was above the strike, we compared three things on the same end date, the new expiry: let the call run (the shares are called away, and the cash sits idle), roll up to a new 25-delta call about 37 days out, or roll out at the same strike. Buybacks were at the ask, new calls at the bid, with fees on every leg. Holds that spanned a split were skipped.

That left 27 rolls on SPY, 29 on QQQ, 31 on AAPL, 25 on NVDA and 23 on TSLA, about a quarter of the entries. Dividends are left out of every row. The entries overlap, so these are a picture of the window, not independent tries, and the window from 2022 to 2026 mostly rose.

What each choice earned

Average result to the end date as a share of the stock price at entry. SPY: called away +4.21%, roll up +4.71%, roll at the same strike +4.53%, shares alone +6.29%. QQQ: +5.63%, +7.29%, +6.21%, +10.12%. AAPL: +7.02%, +8.53%, +7.64%, +11.66%. NVDA: +13.43%, +21.38%, +15.45%, +33.87%. TSLA: +15.24%, +16.81%, +17.23%, +27.99%.

Rolling beat being called away on every ticker, and no covered call row beat owning the shares with no call. That is the cost of the call itself, and the roll only gave back part of it. A stock that is in the money ten days out has already risen, so these rolls are the good cases for staying long.

What the roll cost and what it risked

Rolling up paid a net debit almost every time: on average 1.62% of the share price on SPY, 2.50% on QQQ, 2.93% on AAPL, 8.35% on NVDA and 9.99% on TSLA. The new strike was higher by 4.7%, 6.6%, 9.0%, 21.3% and 26.5% on average, and the shares were called away again on the new call 37%, 38%, 36%, 40% and 17% of the time. Rolling out at the same strike was a net credit of 0.91% to 3.18% every time, but the shares were called away again 70% to 90% of the time, so it mostly postpones the same decision.

The worst single roll up was -3.63% on SPY, -3.54% on QQQ, -3.54% on AAPL, -3.97% on NVDA and -11.79% on TSLA. Letting the shares go did better at the bottom on four of the five: -0.55%, +2.84%, -2.53%, -2.06% and, on TSLA, -10.50%. A roll puts the shares back at risk of a fall after a run-up, with a bigger debit already paid.

Before you roll one

Decide before you sell the call whether you would mind losing the shares at that strike. If you would, a collar or a higher strike is a cleaner answer than a roll after the fact. If you would be happy to be called away, the data says the roll is not a free upgrade: it costs a debit, keeps the shares at risk, and did not catch up with owning them outright.

The same logic is in the guide to rolling a put. To see what a strike was worth on real quotes before you place it, use the strike check. None of this is advice for your account.

Historical replays are not a forecast. Options can lose more than the premium collected. This is education, not a recommendation to trade anything.

Test your own rule on history, free