The wheel is a real and popular way to trade. What it is not, as usually written down, is testable. Here is the gap, item by item.
| The step | What is usually said | What a test needs |
|---|---|---|
| Sell a put | “Below the price” | Which delta, which tenor, what minimum credit |
| If it expires | “Sell another one” | Same day, or wait? Same delta after a move? |
| If assigned | “You own the shares” | At what cost basis, counted how |
| Sell a call | “Against the shares” | Which delta — and will you sell below your cost? |
| Repeat | “Until called away” | For how long, and what ends the loop |
Will you sell a call below your cost basis? Answer that and you have specified two different strategies.
None of that makes the wheel a bad idea. It makes it an idea that has not been pinned down — which is the difference between something you can argue about and something you can settle.
No. We are saying that as usually described it cannot be tested, so nobody — on either side — is entitled to a confident answer yet.
Most of them. The wheel gets named here because it is the clearest example and the most argued about.
A specification, a long enough sample, and a result that holds on data it was not chosen on. That is the whole of it.
Yes, and its specification is published on the article before the run. The cost-basis rule is named there rather than left to the reader.
The wheel on high-volatility names is written and unrun.
Its specification names the delta, the tenor and the cost-basis rule, and the result will be reported against buying and holding the same names — because a wheel number without that comparison does not answer the question anyone is asking.
