How to Follow Your Trading Plan: The Four Places Adherence Actually Breaks

By Sofia Harchich | Trading Psychologist & Behavioural Finance Writer

How to follow your trading plan is a question usually asked by someone who read their plan that morning.

That detail is the whole problem, and it is the one the standard answer cannot accommodate. The information was present. The rules were legible. You could have recited them at the moment you did something else, and some part of you was watching yourself do it.

A knowledge gap does not behave like that. Something else is going on, and it is specific enough to be located.

What the standard answer gets wrong

Almost every published answer to how to follow your trading plan resolves to the same prescription: more discipline, smaller size, print the rules, tape them to the monitor, want it more.

Every item on that list targets the same step. Rules are an enforcement mechanism — the fifth of the seven steps between a trigger and a behaviour. Enforcement works on a person who has already detected what is happening and returned to a state in which the rules can be consulted. Applied to someone at step one or two, it is a lock on a door that was never the way in.

The evidence for this is in your own record, and it takes about ten sessions to produce. Adherence is rarely uniform. It is high on some days and collapses on others, and the days it collapses on have something in common that is not willpower.

Cause one: the plan was never specified

The commonest finding, and the most quietly embarrassing.

A plan reading enter on a clean pullback with confirmation describes a feeling you have had about charts before rather than specifying anything. Two competent people reading it would take different trades, and so will you on two different mornings, which is exactly what deviation looks like from the inside.

The test: could someone else, holding only your written criteria and no conversation with you, take or reject the same entry you took? If the answer is no, adherence cannot be measured, because there is nothing definite to adhere to.

The fix: four minutes before the open, written. Three criteria, each answerable yes or no by looking at the screen. Price above the 20 on the hourly. Retest held for two candles. Stop under the swing, risk under 1%. Ugly and checkable beats elegant and interpretable.

Traders who resolve here are often relieved and slightly deflated, because it turns out the problem was clerical. That is a good outcome. It is the cheapest of the four.

Cause two: the state at the open had already decided

If the criteria are specific and deviation still happens, look at what the deviating sessions share.

Log one word before any chart is visible: settled, activated, or flat. Thirty seconds, before the platform opens. Then, at the fortnight, sort your adherence count by that word.

Most traders find their deviations cluster. Adherence at eighty or ninety per cent on settled days, forty or fifty on activated ones, with the same plan and the same person. That distribution says the operative variable is state rather than resolve, and it moves the intervention to the hour before the session, where none of the standard advice is aimed.

The three broad states — regulated, mobilised, shut down — are described by polyvagal theory, and the practically important one is the middle. Mobilisation narrows attention and shortens the interval between an impulse and an action. A plan requires you to hold a gap open. Mobilised, the gap closes before the plan is reached.

The fix, and it is unglamorous: on an activated reading, halve the size and take only the first setup. Not a rule about discipline. A rule about which instrument you are currently using to make decisions.

Cause three: the decision was right and the action was late

The third cause hides inside what looks like a discipline failure and is not one.

The plan said exit on invalidation. Invalidation occurred at 10:41. You exited at 10:48. Asked afterwards what happened, you will say you did not follow the plan — and the record shows you followed it, seven minutes late, at a materially worse price.

That is a latency, and it is measurable in a way that I lack discipline is not. Two timestamps: when the invalidation condition occurred, and when you acted. Nothing else. The median of those gaps across ten sessions is one of the most informative numbers a discretionary trader can hold, and almost nobody has it, because end-of-day reconstruction cannot produce it — timing is precisely what memory does not keep.

Latency responds to a different intervention from the first two causes. What shortens it is a routed action rather than a decision: signal arrives, written criteria get read, the criteria decide. Four seconds. The routing is the point, because a trader in the gap is being asked to reason at the moment reasoning is least available.

Cause four: the plan is followed and the deviation is in size

The fourth is the best disguised, because on paper the record looks clean.

Entries meet the criteria. Exits happen at the stated level. Every box is ticked, and the account still behaves as though there were no plan — because the plan was followed at 1.4 times the intended size on three of the ten days, and those three carried the month.

Size drifts more quietly than entry selection does, and it drifts in a pattern: upward after a win or two, upward again after a loss that felt unjust. Neither feels like breaking a rule. Both are.

The test: a column for planned size and a column for actual size, on every row. Sorting the days where those two numbers differ reveals whether the drift follows the last outcome or the state word.

Three of the four causes are places where the instability loop reaches the order before the plan does: a state that had already decided at the open, an activation that delays the action, and a size that changes after a run of outcomes.

How to follow your trading plan, once you know which cause it is

The four causes take different interventions, and applying the wrong one is why the question keeps getting asked.

If the finding is	The intervention is	Where it happens
Criteria not checkable by a stranger	Three yes/no criteria, written	Four minutes before the open
Deviations cluster by state word	Halve size, first setup only	The hour before the session
Invalidation acted on late	Route to written criteria, four seconds	Mid-session
Planned size ≠ actual size	Size fixed before the entry, not at it	At order entry

Notice what none of them is. None is a resolution, an affirmation, or a decision to care more. Each is a change to a procedure, at a stated time, producing a number you can check on Friday.

Adherence is an output. The useful question is what it is an output of, and that varies between traders in a way no general advice can reach.

The ten-session count

One measurement, and it should be running before any of the four fixes is attempted. Everything useful anyone can tell you about how to follow your trading plan comes out of this count rather than out of an argument.

After each session, one number: how many entries fell outside the written criteria. Not how disciplined it felt — mood tracks the last outcome rather than the behaviour, and it will tell you the day you broke every rule and won was a good one.

Ten sessions produces a baseline. Then add the state word and sort. Most traders find their cause in the first sorted column, and the whole exercise costs under ninety seconds a day.

The honest first target is eight out of ten, not ten out of ten. A zero-tolerance standard is abandoned in week two, and it takes the record with it — which is worse than the original problem, because you then have neither adherence nor evidence. Improvement to eight is visible, sustainable, and enough to change a month.

There is a second number worth adding once the first is stable: how early in each session you noticed the pattern, scored nought to five. Awareness usually improves within a fortnight. Adherence lags it, sometimes by months. The distance between those two is what reports on whether anything is actually changing, and watching it narrow is more informative than either number alone.

What the discipline framing was getting right

Fairness is owed here, because the advice is not stupid and it has not survived by accident.

Discipline is how almost every other skill is acquired. Practise the difficult thing repeatedly under conditions that punish inattention, and capability follows. That model is correct for a musical instrument, a language, a sport, and most of professional life. Traders arrive with it because it has worked for them everywhere else.

Its failure here is one of level. In those domains the practice conditions are stable enough that repetition accumulates. In a live position they are not: the trader at the moment of the decision is in a measurably different physiological state from the trader who wrote the plan, and the plan was written by the calmer one. The planning fallacy describes the general version — plans are made by a version of us who will not be the one executing them, and the mismatch is systematic rather than occasional. That mismatch is also why a trader can be profitable on demo but not live with nothing in the analysis changing.

So the discipline framing is not wrong about effort. It is wrong about where the effort has to land.

The conclusion underneath the question

Worth naming plainly, because it is usually there and almost never said aloud.

The trader asking how to follow your trading plan for the third month running has generally reached a private verdict: the problem is me. Not the criteria, not the state, not the routing — a defect of character, evidenced by repetition, which no amount of reading has fixed.

The verdict is understandable and it is doing real damage, because it points effort at the one thing that cannot be adjusted. A defect of character has no threshold, no measurement, and no Friday test. It also produces a specific behaviour: the trader stops recording, because a record of a character defect is only humiliating.

What the four causes offer instead is an address. Three of them are procedural and one is physiological, and every one of them has a number attached. That is a claim you can falsify in ten sessions, which is more than reassurance has ever offered anyone.

Count first, resolve later

Ten sessions, one number, no changes to how you trade. That is the whole of how to follow your trading plan as a first step — a measurement rather than a resolution.

Count entries outside the written criteria. Add the state word before the open. At the end, sort by the word and look at whether the deviations cluster. That single sort resolves most traders into cause one or cause two, and the other two are visible in the same data.

Then apply one intervention. One. Applying all four at once means learning nothing about which of them mattered, and the point of the count is to find out.

For the framework these four causes sit inside, read what is trading psychology; for what to record while you run the count, trading journal psychology; for the state-based reading of the second cause, emotional discipline in trading.

The pattern quiz identifies which emotional pattern your behaviour under pressure fits, and it usually predicts which of the four causes above your count will return.

---

About the Author

Sofia Harchich is a Trading Psychologist and Behavioural Finance Writer with a Master’s in Psychology. She works at the intersection of Jungian shadow work, neuroscience, and market behaviour — helping traders understand the psychology driving their decisions, not just the strategy.

Read more at thewealthmirror.com/about