Once you have decided to sell a put, exactly one decision remains: which strike. Everything else — the stock, the expiration, the size — you have already settled. And this one decision determines your premium, your cushion, your odds of assignment and your annualised return simultaneously, in opposite directions.

There is no best answer. There is a curve, and you pick a point on it. Here is the curve.

Why delta, and not dollars

You could pick strikes by distance — “5% below the current price” — and plenty of people do. The problem is that 5% means something entirely different on a utility stock than on a biotech. On one it is a wall; on the other it is a Tuesday.

Delta already contains the stock’s volatility and the time remaining. A 0.16 delta put is approximately one standard deviation out of the money on anystock, at any price, for any expiration. That portability is why the whole industry speaks in deltas, and why “sell the 16 delta” is a complete instruction while “sell five dollars below” is not.

Reading the trade-off honestly

Move the volatility slider up in the widget and watch every row’s premium grow. Move the expiration out and watch annualised return usually fall even as the premium rises. Both are worth sitting with — but the central pattern is the one across the rows.

Notice that annualised return generally increases as you move toward the money. If annualised return were the only thing that mattered, everyone would sell 0.45 delta puts. The reason nobody sensible does is that the metric does not contain the losses — it prices the win and ignores the frequency and size of assignment.

Three filters, applied in order

First: could you own it? Multiply the strike by 100. If that number is more than you want in one company, stop — no delta fixes a position that is too large. This filter kills more candidate trades than the other two combined.

Second: would you want to own it there? If the stock fell to your strike, would you be buying because you like the price, or would you be catching a company whose story has changed? Delta cannot tell you this. Nothing on the option chain can.

Third: which delta? Only now does the widget above become the relevant question, and by this point it is the easiest of the three.

What experienced sellers tend to settle on

There is a broad convention around 0.16 to 0.30 delta at 30 to 45 days. It is not a rule and nobody can prove it optimal, but the reasoning is coherent: it sits in the steep part of the theta curve, it keeps assignment to something like one trade in four or five, and the premium is large enough to survive commissions and the bid-ask spread.

Below 0.10 delta the premium often does not clear the transaction costs, and you are taking genuine tail risk for a few dollars. Above 0.40 you are not really selling premium any more — you are buying the stock with extra steps, and should probably decide whether you want to own it outright.

What can go wrong

Optimising for annualised return. It systematically pushes you toward the money, which is systematically toward assignment.

Selling far out of the money and thinking it is safe. A 0.05 delta put still gets assigned roughly one time in twenty, and when it does the move that caused it was large. You collected very little for that.

Using delta on a stock with an event inside the expiration. Delta assumes a smooth diffusion. An earnings date is a scheduled gap, and the model has no way to represent it.

Letting the delta drift unnoticed. The 0.16 you sold becomes 0.40 when the stock falls. The position you are holding is not the position you opened.