What a Trading Plan Has to Decide in Advance

The short answer

Almost every trading plan you will read is a list of intentions: I will only take A-grade setups, I will respect my stops, I will not trade after two losses. These are goals. The psychology literature is unusually clear about how much work goals do on their own, and the answer is: some, but far less than people assume. Intentions account for only 20 to 30 per cent of the variance in behaviour.

The form that performs better is narrower and stranger-looking. It is an if-then pairing: an anticipated situation, named specifically enough that you will recognise it when it arrives, joined to a response you have already chosen. "Whenever situation x arises, I will initiate the goal-directed response y." Peter Gollwitzer called these implementation intentions, and the research programme behind them is about why they work when resolutions do not.

A trading plan built that way is short. It is also testable afterwards, which is the second reason to write one.

What the research actually found

Gollwitzer's argument is that implementation intentions delegate the control of a response to an anticipated situational cue, so that when the cue is encountered the response is elicited automatically rather than being decided again under pressure. The decision has already been made; what remains is recognition.

Two results give the size of the effect.

Difficult projects. University students were asked, before a Christmas break, to name two projects they intended to complete — one difficult to implement, one easy. Completion was checked afterwards. Among those who had formed implementation intentions about when and where to start, two-thirds completed the difficult project. Among those who had not, only a quarter did.

A specific, awkward task. In another study, participants were asked to write a report on how they spent Christmas Eve. Three-quarters of the implementation-intention participants had written theirs within the requested period; only a third of the control participants managed it.

And then the finding that stops this becoming a universal remedy: for the projects that were easy to implement, the completion rate was high regardless — around 80 per cent either way. Where action initiation is easy to begin with, planning this way added nothing.

That qualification matters enormously here. It tells you which parts of a trading plan are worth writing in this form: not the parts you would do anyway, but precisely the ones that are hard to initiate in the moment. Closing a position at a loss. Stopping for the day. Doing nothing when nothing qualifies.

What a trading plan has to decide

Everything on this list shares a property: it is cheap to decide in advance and expensive to decide at the screen. That is the test for whether something belongs in a plan at all.

Before entry

• What makes something tradeable at all. Written as criteria that can be checked off, not as a description. If you cannot tell afterwards whether a trade met them, they were not criteria.
• Where this idea is wrong. A level, decided from the market, before the position exists. See stop-loss discipline.
• What a mistake may cost, as a fixed amount, and therefore the size. In that order — the reasons are set out in how much to risk per trade.

During

• What would make you exit early. Specifically: "if the reason for the trade no longer holds, it is closed, profit or not" is an if-then. "I will manage my trades well" is not.
• What you will not do. Adding to a losing position, moving a stop away, re-entering immediately after a stop. Each written as a situation and a response.

Around the session

• A loss limit for the day or the week, and what happens when it is hit. This is the single most valuable if-then in most plans, because it covers the state in which every other rule is least likely to be followed. See what a losing run does to judgement.
• What would stop you trading entirely for a period, and who or what enforces it.

After

• What gets recorded, and when. Before the outcome is known, or it records your explanation rather than your decision. That is the argument of what to actually record.

The form matters more than the content

Compare these two lines, which say approximately the same thing:

I need to be more disciplined about my stops.

If price trades through the level I named before entry, the position is closed at market, with no reassessment.

The first cannot be broken, because it does not specify anything. The second can be — and being breakable is the point. A rule that can be shown, from your own records, to have been followed or not is the only kind that survives contact with a bad week.

Three properties to check in each line of a plan:

1. Is the situation recognisable without judgement? "When the market gets choppy" is not. "When the day's loss limit is reached" is.
2. Is the response a single action? Not "I will reassess" — reassessment is the thing you are trying to avoid doing under pressure.
3. Could an outsider, given your trade log, tell whether you followed it? If not, it is a sentiment.

What a plan will not do

It will not make a method profitable. Nothing in Gollwitzer's work, or anywhere else we have read, suggests that planning improves the quality of a trading idea. What it changes is the likelihood that what you decided calmly is what actually happens.

It will not survive being written once and filed. The evidence is about forming a specific plan for an anticipated situation, which implies revisiting it when the situations change.

And it will not help with the parts that are easy. If you would have done it anyway, writing it down is administration, and padding a plan with things you never struggle with is a reliable way to make it long enough to stop reading.

Limits of this evidence

Gollwitzer's research is in general psychology — student projects, health behaviours, report-writing. None of it is about trading, and we have not found a study testing implementation intentions in a trading context. Applying it here is an inference, and we label it as one.

What makes the inference reasonable is the shape of the problem rather than the domain: a decision that is easy to make in the abstract, hard to execute at a particular moment, and preceded by a recognisable cue. That is the situation the research is about. It is also, almost exactly, the situation described throughout our account of what has been measured about trader behaviour.

Nothing on this page is advice. Trading leveraged products carries a high risk of losing money quickly.