Hand it to somebody else with the same price history. If they come back with a different set of trades, it was not a strategy — it was a description, and the gaps were filled in by your judgement without either of you noticing.
| What is said | What is missing | Can it be tested? |
|---|---|---|
| “I sell puts on good companies” | Which companies, which strike, which expiry, when you close | No |
| “I sell 16 delta puts, 45 days out” | How wide, what minimum credit, what closes it, on which names | Not yet |
| “16 delta, 45 DTE, $5 wide, minimum $0.50 credit, close at half or at 21 days, on these fifty names” | Nothing | Yes |
A strategy that cannot be handed to somebody else is not a strategy. It is a habit you are good at.
Not quite. A system is usually taken to include position sizing and the rules for running the whole account, not just the individual trade. See strategy vs setup vs system.
To be tested here, yes. Anything that needs a judgement call on the day cannot be replayed against history, so it cannot be compared with the alternative.
As few as possible and no fewer. Every extra rule is another thing that could have been fitted to the past rather than discovered in it.
The exit. Entries get described in detail and the exit gets one clause, which is the half that moves the result more.
Every test on this site publishes its strategy in full before the numbers — the structure, the four numbers, the universe, the period and the exit.
No result is published yet, so there is nothing to link to. The specification of each queued test is on its article now, written before the run rather than after it.
