Rolling is a close and an open, not a rescue

"I rolled it for a credit" is the most comforting sentence in options and the one that hides the most. A roll is two orders sent as one: buy back the option you're short, and sell a different one on the same stock, usually further out in time and often at a different strike. Combining them on one ticket doesn't change what each one is.

The two halves

You sold a $45 put a month ago for $1.20. The stock has fallen to $44 and the put now costs $2.60 to buy back. If you close it, you realize a loss of $1.40 per share, $140. That number exists whether or not you open another trade.

Now you sell a $45 put expiring six weeks later for $3.10. The roll ticket reads: net credit $0.50. That's correct arithmetic and a misleading summary. What happened is that a $140 loss was booked and a new trade was opened that collects $310 and takes on six more weeks of the same obligation. The $0.50 is the difference between those two things. It isn't a profit.

What "for a credit" tells you, and what it doesn't

A net credit means the new option's premium exceeded the cost of closing the old one. That's all. It doesn't mean the position is ahead: cumulatively you've collected $1.20 plus $3.10 and paid $2.60, so $1.70 in premium against a put that's now at the money. And it says nothing about whether the new trade is any good on its own.

The test that keeps rolls honest: if you had no position at all, would you sell this exact put today, at this strike and expiration, for this premium, on this stock? If yes, roll. If the only reason to sell it is that it lets you avoid writing down a loss, you're paying six weeks of risk to keep a number off the ledger.

Out, down, and both

Rolling out, same strike and a later date, collects more time value and keeps the same breakeven per share, minus the loss you booked. Rolling down and out moves the strike lower for less premium, sometimes for a debit, and lowers the price at which you'd be assigned. On covered calls the mirror image is rolling up and out to keep pace with a rising stock. Every version obeys the same rule: the old option closes at a realized gain or loss, and the new one is a fresh trade with its own numbers.

When a roll is the right move

Rolls make sense when the view on the stock hasn't changed and the new position is one you'd take fresh. They also make sense purely as a way out of expiration risk: closing a put with three days left and reselling a month out is often cleaner than sitting through pin risk on Friday afternoon. What rolls are bad at is fixing a position that was wrong. A put on a stock that has fallen 30% on real news doesn't become right because the calendar moved.

How to write one down

Two lines. The first: the original put, opened at $1.20, closed at $2.60, realized −$140. The second: a new put, opened at $3.10, with its own open date and its own outcome still to come. A log that shows one line reading "rolled, +$0.50 credit" will overstate your results every time and, worse, will teach you that rolling works.