The five-step weekly review
A trade review turns a pile of journal entries into one decision. Block 30 minutes on the same day each week, away from live markets, and work through these steps in order:
- Completeness check. Does your journal have one row for every closed trade at your broker?
- Per-setup numbers. Win rate, average win, average loss, profit factor and expectancy for each setup.
- Process check. Results for plan-followed trades versus rule breaks.
- Worst trades. Reread the notes on your three biggest losses in R.
- One change. Write a single, testable adjustment for next week.
The order matters. You check data quality first because every number after it depends on it, and you finish with one change because five changes at once make it impossible to tell which one helped.
Step 1: check the log is complete
Count the closed trades in your broker's history for the week and compare with the number of journal rows. If they differ, find the missing trades before anything else.
Missing trades are rarely random. The ones people forget to log tend to be impulsive entries and big losses, which are exactly the trades a review needs. A journal missing its worst trades will make every setup look better than it is.
This is the step automation helps with most. A journal app that imports fills from your broker cannot forget a trade; TradeGreen connects read-only to 20+ brokers and journals each closed trade automatically. If you keep a spreadsheet, the spreadsheet guide covers importing a broker CSV.
Step 2: split the numbers by setup
Account totals mix your best and worst ideas together. Split them. Here is a worked example of one hypothetical week with eleven trades across two setups:
| Measure | Pullback | Breakout | All trades |
|---|---|---|---|
| Trades | 6 | 5 | 11 |
| Wins / losses | 3 / 3 | 1 / 4 | 4 / 7 |
| Win rate | 50% | 20% | 36% |
| Gross wins | $540 | $90 | $630 |
| Gross losses | −$210 | −$320 | −$530 |
| Average win | $180 | $90 | $158 |
| Average loss | −$70 | −$80 | −$76 |
| Profit factor | 2.57 | 0.28 | 1.19 |
| Expectancy per trade | $55 | −$46 | $9 |
| Net | $330 | −$230 | $100 |
The account made $100, a profit factor of 1.19, which looks like a modest but working week. The split tells a different story: pullbacks carried everything and breakouts lost money. Expectancy for pullbacks is 0.50 × $180 + 0.50 × (−$70) = $55; for breakouts it is 0.20 × $90 + 0.80 × (−$80) = −$46.
Formulas, so you can reproduce these: profit factor = gross wins ÷ |gross losses|; expectancy = win rate × average win + (1 − win rate) × average loss, with the average loss negative. The profit factor calculator does both.
One caution, and it is a big one: five breakout trades is far too small a sample to judge a setup. One week is a flag to watch, not a verdict. Look for the same pattern across several weeks before acting on it.
Step 3: compare plan-followed trades with rule breaks
If you mark every journal entry "followed plan: yes or no", this step takes two minutes and is often the most useful one. Average the R-multiple for each group.
In the example week, suppose 8 of the 11 trades followed the plan and averaged +0.4R, while the 3 rule breaks averaged −0.9R. Together that is 3.2R − 2.7R = +0.5R for the week, and the rule breaks gave back most of what the disciplined trades made.
This comparison separates two problems that need different fixes. If plan-followed trades lose, the strategy may need work. If plan-followed trades win and rule breaks lose, the strategy is not the issue; the execution is. Our guides to revenge trading and overtrading cover the most common execution leaks.
Convert results to R with the R-multiple calculator if your journal only stores dollars: R = (exit − entry) ÷ (entry − stop).
Step 4: reread your three worst trades
Sort the week by R and read the notes on the three biggest losses. For each, ask:
- Was the loss larger than −1R? If so, why: slippage, a gap, or a moved stop?
- Was the position size correct for the stop? Recheck it with the position size calculator.
- Was the setup one of your named setups, or an improvised trade?
- What was happening just before the entry? A previous loss, a slow morning, a news alert?
Then read your single best trade too. Knowing what a good trade felt like is as useful as knowing what went wrong.
Look for a repeated cause, not a dramatic one. One large loss from a gap may be bad luck; three losses that each followed a previous loss within ten minutes is a pattern.
Step 5: choose one change and write it down
Finish by writing one change for next week, phrased so you can check it in the next review. Some examples of well-formed changes:
| Vague | Testable |
|---|---|
| Be more patient. | No entries in the first 15 minutes after the open. |
| Cut breakouts. | Breakouts at half size until 20 more are logged. |
| Stop revenge trading. | After a loss of 1R or more, wait 30 minutes before the next entry. |
| Size better. | Use the calculator for every entry; risk stays at 1%. |
Add the change to your trading plan, and next week start the review by checking whether you kept it. Over a quarter, the list of changes you kept and dropped becomes its own record of what works for you.
Monthly and quarterly reviews
The weekly review is for execution. Longer reviews are for strategy, because they have enough trades to say something:
- Monthly: rerun the per-setup table on the whole month. Is any setup's profit factor consistently below 1.0 across several months? Is the break-even win rate for its payoff above its actual win rate? Break-even win rate = 1 ÷ (1 + payoff), where payoff = average win ÷ |average loss|. In the example, breakouts had a payoff of 1.125 and so needed about a 47% win rate; they managed 20%.
- Quarterly: review the plan itself. Retire setups that have not earned their place over a meaningful sample, and check that your risk per trade still suits your account.
No review process guarantees profits. A review only tells you, accurately, what your own trading has been doing, so that your next decision rests on records rather than memory. If you are losing and cannot see why, this diagnosis guide is a good next step.