Running the book

Picking a good trade and running a good book are different skills, and the second one is where accounts are lost. Nine articles below, all of them working off a single simulated year: $50,000 in cash, six names, 46 trades, and one six-week cycle where five of the six positions were assigned at once.

Position sizing for option sellersThe 2 percent max-loss rule imported from stock trading permits zero contracts on all six names in a worked $50,000 book. A secured-cash rule permits nine. A stress-decline rule permits thirty-four. Which one survives contact with a real drawdown, and why the popular one is unusable.
Single-ticker concentration riskSix positions looks diversified until you weight them. In a worked $50,000 book one ticker was 39.6 percent of the capital at work and three names in one sector were 78.5 percent. The sector fell 22 percent and five of six positions assigned in the same six weeks.
Correlation risk in an options bookThree short puts each with a 27 percent chance of assignment should all land together 1.9 percent of the time. At the 0.80 correlation semiconductors actually carry, it is 14.6 percent. The same correlation makes quiet quarters more likely too, and both facts come out of the same calculation.
Maximum loss on a cash-secured putThe maximum loss on one worked $113 put is $11,022, reached if the company goes to zero. That number is true and useless. The two-standard-deviation loss is $1,933, the three-sigma loss is $3,302, and what actually happened was $1,422 on a 1.67 sigma move.
Picking up pennies in front of a steamrollerA 0.008 delta put on a $118 stock pays $6, and $4.70 after commissions. That is $37 a year on $8,200 of secured cash, against $344 for leaving the same cash in a money market. One 40 percent drawdown costs 30 years of the premium.
Margin against a cash account for option sellersSix puts that need $26,800 of cash need $3,754 under exchange margin rules. That lets a $50,000 account hold thirteen copies of the same book. When the worked crash cycle arrived, five copies produced a $6,296 margin call and the cash account was never asked for a dollar.
The options income ledgerA worked year of 46 trades collected $4,598 of premium, which reads as 9.2 percent on a $50,000 account. The account made 14.3 percent. The gap is the share column, and the most common ledger error inflates the same year by $588 by counting the premium twice.
Did the covered calls beat buy and holdA worked $50,000 premium book returned 14.3 percent against 10.8 percent for buying the same shares. It won, and the two positions that were called away still handed over $3,900 of upside, which is 84.8 percent of the entire year's premium. Why one good year proves nothing.
Option seller psychologyFour ordinary reactions to one bad six weeks, each costed against the mechanical ledger that did nothing. Closing the tested put cost $3,763. Sitting out two cycles cost $1,051. Concentrating in the loudest premium turned a 2.8 percent hit into 11.4 percent.

Read them in this order

If you are sizing a book for the first time: the three sizing rules and what each one permits, then why six positions can be two bets, then why they fail together. Those three decide almost everything before you place an order.

If you already run a book and want the pages that change decisions: what leverage does to the same six puts, the ledger error that flatters a year by $588, and the four reactions to a bad cycle, priced.

One account, nine articles

Every number here comes off a single illustrative year. $50,000 in a cash account, six names, eight 46-day cycles, cash-secured puts around 0.22 delta and covered calls above cost basis after assignment, with no rolling and no early closing anywhere. It collected $4,598 of premium across 46 trades and finished at $57,168, up 14.3 percent. Five of 29 puts were assigned, two of 17 calls were exercised, and two cycles produced no trade at all on one name because no strike above its basis was worth ten dollars. Every price and probability was computed through Black-Scholes at 4.2 percent rather than asserted.

What this series will not tell you

That the strategy won because it is good. It beat buying the same six stocks by $1,758 over this year, and the audit page publishes that and then takes it apart, because the result turned on one stock collapsing and on assignments landing near a low. The ten-year index record already on this site says the opposite.

It also will not tell you that a high win rate is safety. The two positions that were called away handed over $3,900 of upside above their strikes, which is 84.8 percent of the entire year's premium, and that happened in the year the strategy won.

And it will not explain how the OptionsKing confidence score is computed. Sizing and concentration are arithmetic and belong in public. What the engine does stays private: the score feeds a ranking and the app shows you the highest-ranked handful of what it found, and how it works covers what that guarantees.

Questions people actually ask

How many option contracts should I sell at once?

Size so a 25 percent decline in one name costs no more than about 5 percent of the account, and so the same decline across everything at once costs no more than about 15 percent. The 2 percent max-loss rule people import from stock trading permits zero contracts on every name in a worked $50,000 book.

Why do several of my positions always go wrong together?

Correlation bunches the outcomes. At the 0.80 pairwise correlation a sector basket carries, three short puts with a 27 percent assignment chance each land together 14.6 percent of the time rather than 1.9 percent, and the odds of a completely quiet cycle go up at the same time.

Is premium collected the same as my return?

No, and the gap is the point of the ledger page. The worked book collected 9.2 percent of the account in premium and returned 14.3 percent, and one position produced six winning trades and still lost money.

Do I need a margin account to sell puts?

No. Exchange minimum margin would carry the same six puts for about a seventh of the cash, which is 7.1 times the leverage on an identical obligation. In the worked crash cycle the cash account was never asked for a dollar and five leveraged copies of it produced a $6,296 call.