How to build a method of your own

A method isn't a secret strategy or a formula: it's a set of decisions written down before money is at stake. And there's psychology research explaining with some precision why writing them works and why keeping them in your head doesn't.

What a method has to define to count as one

Five things. Miss one and what you have is an intention, and intentions don't survive the first moment of pressure.

The five

What you look at. The universe of assets and the type of situation. "S&P 500 companies in the week of their earnings" is a method. "Whatever comes up" isn't, because it can't be evaluated: there's no way to tell whether it worked or whether you simply chose well what to look at afterwards.

What makes you enter. The specific criterion, written so another person could apply it and reach the same conclusion. If only you understand your criterion, it isn't a criterion: it's an intuition with vocabulary.

How much you put in. It comes from the previous chapter's risk sheet: the percentage you risk divided by the distance to your exit point.

What makes you exit. And both are needed here: the level at which you accept you were wrong, and the criterion for closing when you were right. The second gets forgotten constantly and decides half the outcome, because without it closing a gain gets decided by the previous session's impulse.

What would make you stop using it. The one fewest people write and the one that protects most. Without it, a method gets abandoned through fatigue rather than through evidence.

The test for whether it's a method

Very simple: could another person read it and make the same decision you would in the same situation?

If the answer is no, it isn't a method. It's a way of deciding that depends on who's in front of the screen that day and how they feel, and that can't be evaluated or improved, only repeated with variable luck.

Why writing it works and thinking it doesn't

There's solid evidence here, and it doesn't come from finance: it comes from behavioural psychology.

The finding

Peter Gollwitzer and Paschal Sheeran published a meta-analysis in 2006 covering 94 independent tests with over 8,000 participants, on what their field calls an implementation intention: a plan taking the form "if situation Y arises, then I will do X," specifying the when, where, and how in advance.

They found a medium-to-large effect on goal attainment. And there's a second result even more relevant here: the effect on preventing a goal already underway from derailing was of similar magnitude, if anything slightly larger.

Why this matters for investing

Because it describes precisely this chapter's problem. Almost nobody fails through not wanting to: they fail because at the critical moment the decision gets made under pressure, and under pressure the first session's biases win.

A plan written in if-then form doesn't depend on willpower in that moment. It moves the decision to an earlier point, when no money is moving and your head works properly.

Applied, the difference is this:

• Resolution: "I'm going to be disciplined about losses." It has no when, no where, no what.
• Implementation intention: "If this stock closes below €47.50, I sell at the next open." It has all three.

The first is what almost everyone has. The second is what the research found works.

And why this validates the course's three artifacts

It isn't accidental that this course produces three documents the reader fills in rather than three lists of advice. The profile sheet, the risk sheet, and this plan are exactly that: decisions specified in advance, with their when and their what.

Writing them doesn't guarantee following them. What the evidence says is that the probability of following them rises measurably against keeping them in your head, and that's the whole argument.

What goes in a trading logbook and why

A trading logbook isn't a record of trades. Your broker already has that and it teaches nothing, because it only contains what you bought and at what price. A logbook records reasoning, which is why it gets written beforehand.

What goes in, and when

Before opening the position, four things:

1. What I expect to happen and why, with the specific data I'm relying on.
2. What would have to happen for me to be wrong. It's Chapter 4's invalidation point.
3. How much I'm risking, as a percentage of capital, and where the exit sits.
4. How I'm doing today. It sounds out of place in a financial document and it's among the most useful things to record: if you're trading in a hurry, angry, or recovering a loss, knowing that afterwards explains an enormous amount.

After closing it, two:

5. What happened, without decoration.
6. Whether the reasoning was correct given the information available at the time, which is a different question from whether it worked out.

Why the order is the only thing that matters

Because writing the thesis after knowing the result is worth precisely nothing.

It's the chart chapter's hindsight bias applied to your own decisions: knowing how it ended, any reasoning rewrites itself to fit. The logbook only works if the earlier record exists and can't be edited.

From which comes this tool's one hard rule: what was written before doesn't get edited afterwards. If it needs qualifying, add a new line with its date.

What it's genuinely for

To answer a question nothing else answers: do my failures come from bad reasoning or bad execution?

They're two different problems with opposite solutions. If the reasoning is sound and execution fails — you don't honour your own exits, you open positions that weren't in the plan — the problem isn't analytical and studying more won't fix it. If execution is flawless and results don't come, then the criterion does need reviewing.

Without a logbook, those two situations are indistinguishable. And they're exactly the two demanding opposite responses.

Which metrics evaluate a method

Four numbers, and none of them is "how much have I made."

The four

Expected value. What each trade contributes on average, combining the hit rate with the relative size of wins and losses. It's the previous chapter's opening figure and it remains the one deciding whether something makes sense repeated.

Maximum drawdown. The worst fall from a peak to the subsequent trough. It matters as much as return because of the recovery arithmetic you already know, and for one added reason: it's the metric deciding whether you'll still be invested when it arrives.

Number of trades. Not a quality metric but the one saying whether the other three mean anything. At ten trades they mean nothing, as the previous chapter's final session demonstrated with confidence intervals.

Plan adherence. The percentage of trades that followed what was written. It's the only one of the four you can measure from week one, and the only one depending entirely on you.

The one almost nobody measures and that informs most

The fourth. It's what separates a method problem from a discipline problem, and none of the other three can do that.

A method with good expected value and 60% adherence isn't a method with good expected value: it's a method that isn't being applied. Changing it would fix nothing, because the one you have hasn't been tested yet.

When to change methods and when to hold

It's the chapter's hardest decision, because both wrong answers feel like good sense.

When NOT to change

Over a losing run. You know the arithmetic: at a 55% hit rate, five losses in a row appear about twice per hundred trades. It's what's expected, not information.

Over what someone else did. That something else worked for someone this quarter says nothing about yours, for the same small-sample reasons.

Out of boredom. More common than it looks. A method that works is, almost by definition, repetitive and unexciting. Changing it so things happen is changing the objective without saying so.

When you DO change

When the criterion you wrote at the start gets met. That's why a method's fifth element is "what would make me stop using it": so this decision is taken in advance rather than in the heat of it.

When the reason it worked changes. If your method rested on a specific condition and that condition disappears, it loses its foundation even if results don't reflect it yet.

When the sample is already sufficient and the result is bad. With enough trades and a confidence interval that no longer includes what you needed, the method is disproven. That isn't a run, it's a result.

The asymmetry worth keeping in mind

Changing too soon carries a guaranteed cost: never accumulating a meaningful sample of anything, and being left hopping between systems without ever learning whether one worked.

Holding too long carries a probable but not certain cost: continuing to apply something that doesn't work.

Since the first is certain and the second only probable, the default lean should be toward holding longer than you feel like. With one non-negotiable condition: that the abandonment criterion is written down, because otherwise "holding" stops being a decision and becomes inertia.

Your one-page plan

The third and last of the documents you take from this course. It fits on a page because it has to be readable in full at the moment it's most needed, which is when something is going wrong and there's no time.

Six fields, each capturing something already decided in earlier sessions.

What kind of situation gets looked at. That's the universe. It doesn't need to be ambitious; it needs to be specific, because only the specific can be evaluated.

How often the portfolio gets looked at. Setting the frequency in advance stops the phone notification from setting it, and the previous session in this chapter measures what happens to the result when impulse sets it instead.

What has to be true to enter. Written as if someone else had to apply it. A criterion only its author understands can't be evaluated or corrected later.

What makes an exit. Both: where the error gets accepted, and by what criterion a winner gets closed. The second is the one most people leave blank.

What gets recorded for each trade. Even one line. A logbook filled in half-heartedly is still infinitely better than none, because it preserves the chronological order that's the one thing that can't be reconstructed later.

When the method gets reviewed. Setting the date now, cold, is what stops the review from happening hot right after a bad run, which is when decisions come out worst. The abandonment criterion fits here: what would have to happen to stop using it.

With this sheet filled in you have all three of the course's artifacts: what each pot of money is for and when you need it, how much you risk per trade, and how you decide. Saving them to a free account means they're still there in six months, which is when you'll genuinely want to reread them. Finishing the course doesn't require it.

What to remember

• A method defines five things: what you look at, what makes you enter, how much you put in, what makes you exit, and what would make you abandon it.
• Gollwitzer and Sheeran found across 94 tests and over 8,000 participants that a written "if Y, then X" plan has a medium-to-large effect on meeting a goal.
• The logbook only works if written beforehand, and what was written before doesn't get edited afterwards.
• Plan adherence is the metric that informs most and the only one you can measure from week one.

Published August 3, 2026. Last reviewed: August 3, 2026.

Related: how much to risk on any single trade · how to read a full event