Journal and weekly review

This lesson describes the review process: when it takes place, what is read then, what decisions are made, and what specifically changes in the plan for the next period. It is for someone who already has trade entries and needs a fixed procedure so those entries translate into rule changes, rather than more notes. The method of calculating metrics — that is, which numbers you extract from the journal — is in the lesson journaling and metrics; here we deal exclusively with rhythm and decisions.

A review is an appointment in the calendar, not an occasion

A journal without a review is an archive. You record decisions in it, but nothing organises them, so you introduce corrections under the influence of the last transaction rather than the whole period. A review reverses that order: first you read the period, then you choose the change. The condition for it to work is a fixed date — entered in the calendar like any other mandatory task, with a reminder that does not depend on whether you feel like going back to the charts. What you are looking for in such a review is a repeating process error, not a single loss — their catalogue is collected in the lesson Common beginner mistakes.

The second condition is practical: the review has a predefined agenda and time limits. Without them, a meeting with yourself drifts into chart-watching and ends without a decision. The agenda below is a proposal you can adapt, but it is worth keeping the order of the blocks — decisions are made at the end, after the data.

When to review: three rhythms

The three rhythms have different jobs. The week answers the question of what to change now. The month checks whether the changes held and whether the plan’s limits were respected. The quarter does not assess transactions — it assesses the review, for example how many times the date slipped to tomorrow. The times in both tables are starting values, that is, illustrative numbers: after the first two reviews you will fit them to your way of working.

Weekly review agenda: 45 minutes in five blocks

The time limit is not decoration. Blocks two and three serve description, the fourth — decision. If after 25 minutes you have no named deviations, you end the block and move to a decision based on what you already know, instead of moving the whole meeting. A review that does not end with a decision repeats the following week in identical form.

The rule here is different from daily work: during a session you make decisions quickly, while in a review you may take your time — but only because in block four you already have the complete data, not a single candle.

What you read before the review

Preparation takes a few minutes and consists of opening four things:

1. The period’s journal — all entries, including those you do not remember. Memory suggests unusual transactions, while a review needs the complete set. Keep a CFD account separate from a brokerage account, otherwise the review mixes two sets of costs and two different sets of rules.
2. The equity curve and drawdown from peak — the curve shows the course of the period, and drawdown from peak is one number that describes the state of the account, not the impression from the last day. The context is described in the lesson expectancy and drawdown.
3. The current version of the plan — rules, limits, and setup names. Without it you cannot distinguish a deviation from a rule change.
4. The list of transactions outside the plan — if you keep it as you go, block two shortens to a few minutes. You record the result in R, so that positions of different size can be compared in one review.

During the review you do not go back to the charts looking for explanations. If the entry is not enough to understand a decision, that is a conclusion about the journal: the field was too poor. You look at charts when you are settling a hypothesis about a setup, and you record the condition, not a pretty pattern.

What decisions a review makes

The decision list is closed. You leave the review with two decisions: the first you choose from the four categories below, the second indicates what you remove from the plan:

• Keep unchanged — the rule works and there is no reason to touch it. That is also a decision.
• Restrict — you narrow the scope: fewer setups, smaller position size, or a shorter session window.
• Suspend for a sample — the condition stays in the plan, but without real capital, until more observations are collected.
• Remove — the rule or setup leaves the plan, rather than waiting for a better moment.

The outcome of a review has three elements: two decisions — one from the four categories and one removal from the plan — plus one observation for a further sample. That trio fits in a single paragraph and can be written in a few minutes. Five changes at once means no information about which of them worked — after two weeks you will not distinguish the effect of the change from ordinary volatility.

Protocol for recording a decision

1. Name the decision in one sentence in the imperative. For example: I am limiting the number of Friday entries to two.
2. Indicate the data it rests on. The review date, the number of transactions in the sample, and the period they concern. Without that, the change is an opinion.
3. Indicate the place in the plan. A rule, a daily limit, a setup definition — the specific fragment that changes. Setup definition is discussed in the lesson setup definition.
4. Set the withdrawal criterion. When you return to the previous version and after what period you check it.
5. Raise the plan version number and enter the date. The version is evidence that the change took effect and did not stay in the notes.
6. Enter the check date in the calendar. A change without an assessment date becomes permanent through neglect.

Recording the plan version has one more function: when a month later you return to the question why did I change this, the answer is in the plan, not in memory. The same discipline is visible in the session sheet described in the lesson session playbook.

How to close a week and how to close a month

Closing the week is the five-block agenda plus one record: a new plan version with a date and the next review date. Nothing more is needed. If the week brought no transactions, the review still takes place — you then check whether the plan matches market conditions, and you keep the version unchanged.

Closing the month extends the same scheme over a longer period and adds four checks:

1. Whether the changes introduced in successive weeks held until the end of the month.
2. Whether the plan’s limits, including the daily limit and pause thresholds, were respected.
3. Whether drawdown from peak fits within the range the plan considers normal — the definition is in the drawdown entry.
4. Whether any setup is collecting a sample so slowly that it will not support a conclusion over the next month — then the decision is remove or suspend.

The month closes with one document: the plan version for the next period. You do not create a separate monthly report if its content does not change the plan. A document without a decision is extra work you will stop doing after two months.

How a review changes the plan for the next period

The plan changes only as a result of a review and only to the extent that follows from the data. In practice that means three kinds of change.

The first is narrowing the scope: fewer setups, narrower trading hours, smaller position size. It takes effect immediately, because it reduces risk.

The second is widening the scope: a new setup, a longer session window, larger size. It takes effect once the criterion from the protocol is met, for example after a set number of plan-compliant transactions. Widening without a criterion is changing the rules mid-game.

The third is removing a rule that could not be executed or did not change the result. A shorter plan is easier to execute than a plan in which half the points exist only because someone added them once. You record changes in the same version of the document you use daily — a plan kept in two files stops being a plan.

When a review says do not trade

1. The review card was not completed on time. Until it is completed, you work at reduced position size, with no exceptions to the deadline.
2. The number of transactions outside the plan grows for a second period in a row. The change from the previous review was supposed to stop that, so you execute it first and only then open new positions.
3. You changed rules during the period the review concerns. First you describe the actual state and the date of the change, then you make a decision — otherwise you are comparing two different plans.
4. Drawdown exceeded the pause threshold recorded in the plan. Full size returns after the criterion in the plan is met, not after a good week.

Review card (fill in once a week)

Typical mistakes in running a review

• A review when there is time. A date that depends on mood always slips when the period was bad — that is, exactly when a review is most needed.
• An agenda without time limits. Chart-watching displaces the decision block and the meeting ends without a change.
• Five changes after one weak week. It is then impossible to establish which change worked, so the next review starts from zero.
• A change without a withdrawal criterion. The rule stays in the plan indefinitely, even though nobody checked whether it helps.
• A separate report without a decision. A document that does not change the plan will disappear from the calendar after two months.
• Going back to charts looking for explanations. Instead, you record the missing field in the journal so the next entry is more complete.

Measurable exercise

Plan a review the way you plan a session. Enter three dates in the calendar: the nearest weekly, monthly, and quarterly. Set a reminder 10 minutes before the first one. Then conduct the review according to the agenda with time limits and measure its length — if it exceeds 45 minutes, record in the card which block overran. Record the outcome as one change in the plan, raise the version number, and enter the check date. After three weeks, compare the cards: the same rule should not fail in three consecutive periods without a decision. The journal numbers you use in block one are discussed in the lesson journaling and metrics, and the full version of the plan is described in trading plan.

Common questions

How long should a weekly review take?

About 45 minutes in five blocks: data, deviations, hypotheses, decisions, and close. The time limits are part of the procedure, because without them the decision block is dropped in favor of chart-watching. Keeping the journal up to date shortens the first and second blocks.

Is a monthly review enough?

A month is a good rhythm for assessing limits and the durability of changes, but too long for correcting current deviations. A deviation from the plan four weeks ago is hard to attribute to a specific decision, so the week remains the basic rhythm.

What to do when there were no transactions during the week?

The review still takes place. You then check whether the plan matches market conditions, whether the limits are current, and whether the next review date is in the calendar. No transactions is not the same as no decisions — the plan version may stay unchanged.

Why measure something I do anyway?

The transaction record alone does not change behavior until there is a date on which a conclusion may be drawn. Without a regular review, a correction is a reaction to the last transaction rather than a decision about the whole period; then it is also impossible to distinguish the effect of a change from ordinary volatility. Measurement with a deadline and a time limit gives the point at which a change enters the plan — and that is why a review still makes sense even in a week without a result.