A trading plan is a set of decisions made while you are calm, written down so they can outvote the version of you that exists mid-trade. Its power has nothing to do with sophistication — it comes entirely from the fact that it was written before.
Why written, why before
In the middle of a live trade you are the worst decision-maker you will ever be: money is moving, and every cognitive bias trading is famous for — loss aversion, revenge, overconfidence after wins — operates at full strength. The plan's job is to move every decision it can out of that moment and into a calmer one. Unwritten plans do not survive contact with an open position; they get renegotiated in real time, and the renegotiator is biased. Written rules can be checked, followed, and — critically — graded: at week's end you can score whether you followed them, which is a different and more useful question than whether you made money.
What the page must contain
- Markets and sessions. Which pairs, which hours. "EUR/USD and GBP/USD, London morning" is a plan; "whatever's moving" is a slot machine tour.
- The setup, described so a stranger could recognize it. Conditions that must ALL be true before you are allowed to click: trend state, level, trigger. If a rule cannot be checked yes/no from the chart, it is not yet a rule.
- Risk numbers. Percent per trade, maximum concurrent theme risk, daily/weekly stop-trading thresholds. These are the plan's load-bearing walls.
- Exit logic — both sides. Where the stop goes and why; how profits are taken (target, trail, time). Entries get all the attention; exits produce the P&L.
- Process rules. When you review the calendar, when you journal, what happens after two losses in a day (most good plans: you stop). Include the news rule — which releases you sit out.
- The prohibition list. Your personal recurring mistakes, named: no adding to losers, no trades in the first five minutes after news, no re-entry within an hour of a stop-out. This list is yours alone, and it grows from the journal.
Keep it falsifiable and small
One page. A plan long enough to need scrolling is a plan nobody consults at 8:29 before the data drops. Precision beats coverage: three setups defined so tightly you take them identically every time will teach you more in a quarter than a document covering every market condition vaguely. The test for every line: could your journal prove you broke it? "Trade with discipline" fails that test; "no position 15 minutes either side of red-folder news" passes.
Changing the plan without cheating
The plan must be allowed to evolve — but never mid-trade and never mid-drawdown-tantrum. The working rule: changes are written on Sunday, with the market closed, justified by journal evidence ("eight of my ten losers this month broke rule 4"), and then frozen for a minimum period — two weeks or twenty trades — before the next revision. A plan edited nightly is not a plan; it is a diary of moods with rules-shaped formatting.
Starting tonight
Write the ugly first version now — one pair, one session, one setup, 0.5% risk, a stop rule, a two-losses-stop rule, and three prohibitions from memory of your own worst habits. It will be incomplete and partly wrong, and it will still beat the alternative, because from tomorrow every trade generates evidence against a fixed reference instead of vibes against a memory. Traders do not fail for lack of a perfect plan; they fail from operating without any fixed reference at all.