August 30, 2026
Most writing about selling options is about how. This one is about the situations where the answer is no, which is a shorter list than it should be and a more useful one.
This is first because it's the only one that can end the account rather than dent it.
A cash-secured put requires strike times 100 in cash, set aside for the whole life of the contract. If that money is earmarked for something else, or if you're relying on being able to close the position before assignment, the trade isn't cash-secured — it's a margin trade with an optimistic plan attached. As covered in what actually happens when you get assigned, you don't control the timing, and a broker facing an unfunded assignment will liquidate on their schedule at whatever price is available.
The test: could I write a cheque for the full strike amount tomorrow morning without moving anything else around? If not, the contract count is too high.
Premium is proportional to expected movement, and expected movement usually has a cause. When one strike pays conspicuously more than the comparable strike on a similar company, that difference is information you haven't decoded yet.
Sometimes it decodes to something you're happy to underwrite — a general market selloff, a sector out of favour. Sometimes it decodes to a court date, a trial readout, an auditor's resignation, or a refinancing that may not happen. The premium looks identical in both cases.
Not knowing isn't a reason to avoid the trade forever. It's a reason not to place it today.
Earnings, FDA decisions, merger votes, court rulings. These share a structure: a single scheduled moment after which the price is somewhere quite different, and the distribution of outcomes is two lumps rather than a smooth curve.
Implied volatility prices this correctly on average, which is the problem. You collect the elevated premium in exchange for the full gap risk, the outcome is decided while the market is closed, and there is no managing the position through it. The trade is a coin flip with a fee attached, and the fee is roughly fair.
People do sell into earnings deliberately, and the volatility collapse afterwards is a real effect. But it's a distinct strategy with its own sizing rules, not an ordinary premium sale that happens to pay more. Drifting into it because the premium looked attractive is the version that goes wrong.
Look at the bid-ask spread before the premium. An option quoted $1.05 bid, $1.35 ask has a $30 spread on a contract that might pay $120. Selling at the bid and eventually buying back at the ask costs a quarter of the credit before anything happens.
Liquidity also decides whether you can act at all. The position you can't close in a hurry is the one you'll be holding through the move you wanted to avoid, and thin option chains widen dramatically in exactly those conditions. Open interest in the low double digits and a spread wider than a few percent of the premium are both reasons to pass.
Selling puts wins most months. That means position sizing gets calibrated on a long run of wins, and the calibration is being set by the easy outcomes.
The useful check is to price the bad case explicitly before entering: if this stock drops 30% and I'm assigned, what fraction of the account is now sitting in this one name? If that number would stop you sleeping, the size is wrong now, while the position is still hypothetical and free to change.
Related: several puts on different companies in the same sector, or on names that move together, is one position wearing four tickets. Correlation collapses diversification at precisely the moment you were counting on it.
The standard justification for selling a put is that you'd be happy to own the shares at the strike. When that isn't true — when the strike was chosen because the premium was good and the strike was far away — the trade has no fallback. Assignment isn't the plan working; it's an unwanted position acquired at an unfavourable moment.
And the assignment arrives on a bad day, not a neutral one. The honest question is whether you'd buy at that strike on a day the stock is trading well below it with a reason attached.
The last one is about you rather than the market, and it's the most reliable predictor of a bad trade in the whole list.
After a loss, the arithmetic of recovery is tempting: one slightly larger sale, or a slightly closer strike, and the account is level again. The strategy accommodates this beautifully, because a closer strike really does pay more and really does win most of the time. So the recovery trade usually works, which means the sizing that produced it never gets corrected, and it's still in place when the trade that doesn't work arrives.
The tell is that the size or the strike was chosen with reference to a previous loss rather than to the position itself. If a trade's justification includes a number from your own history, it's the wrong trade.
Low volatility is not, by itself, a reason not to sell. It's a reason to expect less, and to size accordingly, and possibly to do nothing because nothing is worth doing. But the impulse to reach for compensating size or closer strikes when premiums are thin is how a quiet market ends up producing the same risk as a loud one, with worse pay.