Rolling

A roll is two trades stapled together: buy back the short option you're losing on — that loss is real and realized the moment you pay it — and sell a new one at a later expiry, same or lower strike. Nothing is undone or extended; the second trade finances the first, and the only new value a roll ever harvests is the fresh extrinsic (time value) you sell.

The scoreboard is one number: net credit. New premium minus cost to close. Just after a breach (stock $93 vs a $95 put), rolling out 30 days collects nearly $2 and buying $5 of room costs only about fifty cents. Deep underwater (stock $78), both rolls go to a debit — a put $17 in the money is essentially all intrinsic, so there is no extrinsic left to sell. Rolls are cheap early and impossible late.

A good roll collects a net credit for more time and a better strike. A debit roll at the same strike is paying to postpone being wrong — identical risk, worse P&L, powered by hope. The one honest exception is a small debit for a much better strike, which is explicitly buying risk reduction.

Rolls don't erase a loss; they finance it. The wheel gives the losing short put a dignified exit that pure premium selling lacks: take assignment on a stock you wanted at a strike you'd pay, and let covered calls carry the campaign — or, if the thesis died, take the loss and redeploy. The lesson's roll decider prices all three choices honestly with Black–Scholes.