The weekly review

The daily entry captures. The weekly review looks. Thirty to forty-five minutes, at a fixed time when the market is closed, following a template, and producing at most one action. Its job is not to change the plan; rules change monthly or quarterly, on evidence. Its job is to catch execution drift before a week of it becomes a month of it.

When

Same slot every week. Sunday morning, Saturday afternoon, Friday after the close; it does not matter which, as long as it is written on the plan's REVIEW line and it happens. A review that floats will float into "next week".

Do it with the market closed. A review done during a session is done by someone who is half watching a chart.

The template

WEEK OF     7 to 11 Sep            Plan version  v1.1

COUNTS
  Trading days / filter-fail days      4 / 1
  Trades taken                         7
  Valid setups missed                  2
  Setup grades                         A: 4   B: 2   C: 1
  Plan-follow scores                   3: 5   2: 1   1: 1   0: 0

RESULTS (R)
  Realised R this week                 +2.3R
  Planned R this week                  +3.1R
  Execution gap (planned - realised)   0.8R
  Realised R, A-grade trades only      +2.9R
  Realised R, B and C trades           -0.6R

MISTAKE TAGS THIS WEEK
  early-exit x2, late-entry x1

BREAKERS
  Two-loss rule fired: Tue.  Daily/weekly stop: no.

ONE OBSERVATION
  Both early exits were on Tuesday, after the two-loss rule had
  already fired on the first two trades. I kept watching.

ONE ACTION FOR NEXT WEEK
  When the two-loss rule fires, close the platform, not just stop
  trading. Add to daily routine card.

TRADES TO RE-SCREENSHOT / RE-GRADE
  Thu trade 2: graded B at entry, re-check on screenshot.

Every number comes from the journal rows. If a number cannot be produced, the journal is missing a field.

One observation, one action

The discipline of the weekly review is the limit of one action. Not five. One thing that changes next week, written as a behaviour: "close the platform when the breaker fires", "finish pre-market by 09:20", "take the screenshot before placing the order, not after".

The action is never a change to the setup, the filter or the risk numbers. Those are plan rules, they have version numbers, and they change at the monthly or quarterly review with thirty-plus trades of evidence. The weekly action is about execution: doing what the current plan says, more exactly.

If you find yourself wanting to change a plan rule on Sunday after a bad week, write the idea in a "candidates for review" list at the back of the journal, with the date and the reason. If it still looks like a good idea at the monthly review, with the numbers in front of you, it gets considered then.

Key idea: The weekly review produces one execution action. Plan rules do not change on a Sunday because of a bad week; ideas go on the candidates list and wait for the evidence.

Reading the A-grade split

The most useful line on the template is realised R for A-grade trades versus everything else. Four possibilities:

One week is far too little to conclude any of these. But tracked weekly and summed monthly, the pattern becomes visible long before the monthly stats confirm it.

Re-grading from screenshots

Pick one or two trades a week and re-grade them cold, from the entry screenshot, without looking at the outcome. If the cold grade differs from the entry grade more than occasionally, entry grading is being done with the outcome already half-known (a trade that is moving your way while you fill the row) or under pressure. The fix is to grade before the order goes in, and the checklist from Module 3 has the slot for it.

Missed setups

The counts section has a line for valid setups missed. If it is regularly higher than trades taken, the plan has a hesitation problem, and hesitation is expensive in a way that the trade rows never show. Common causes: pre-market not finished in time, second-guessing a checklist that already said yes, or fear after a losing streak. Each has a different execution fix. The weekly review names it.

Try it: Run the template on last week's trades, even if you have only partial data. Fill every line you can, mark the ones you cannot, and write one action. Then put the review time on your calendar as a repeating event, with the market closed, and treat it like a session.

Recap

• Thirty to forty-five minutes, same slot every week, market closed, following a template.
• The output is one observation and one execution action. Plan rules do not change at the weekly review.
• The A-grade versus B/C split is the most useful line; tracked weekly, it shows dilution before the monthly stats do.
• Re-grade a couple of trades cold from screenshots to check that entry grading is honest.
• Missed valid setups are counted. Hesitation is an execution cost the trade rows never show.