Keep a Trading Journal
Why journaling matters
Memory is a poor record of trading because it edits, and it edits in a consistent direction.
You remember the trades that were dramatic and the ones that confirmed what you believe about your method. You forget the routine losses, the setups you skipped, and the afternoon where you took nine contracts on a plan that called for four.
Exposing patterns
Patterns are invisible at the level of a single trade and obvious across thirty. A setup that fails repeatedly looks like bad luck each time and looks like a weak rule in a column. The same is true in the other direction: the setup quietly carrying your results is usually not the one you would have named.
Removing self-deception
The most useful thing a log does is settle arguments with yourself. Did you really only take setups at marked zones? Is your stake actually constant? Did that losing week come from the method or from four trades outside it? Each of these has a definite answer that memory cannot supply and a column can.
There is a second, slower benefit that only appears after a few months. A log gives you a way to test ideas from anywhere, including this site, against your own trading rather than accepting or rejecting them on instinct. Someone recommends waiting for the candle close; you can check what your early entries actually did. Someone recommends a session cap; you can count what your long sessions cost. Without a record, every suggestion is a matter of taste, and taste is what you were trying to remove.
Building consistency
Recording a trade requires naming it, and naming it requires it to have been a defined setup. That small friction removes a share of marginal entries by itself, before any review takes place. Traders often notice their trade count drop in the first week of journalling without having changed any rule.
- The log is evidence. Memory is an argument.
- Patterns need volume. Thirty rows say things one trade cannot.
- Recording changes behaviour. Before any review happens.
A log turns opinions about your trading into countable facts. That is the entire case for keeping one.
What to record
Keep it short enough that you fill it in during the session, because a log completed afterwards is a reconstruction.
Six fields cover everything you will actually use. Longer formats get abandoned during exactly the sessions that would have been most informative.
- Time. Entry time, which makes gaps between trades visible.
- Setup name. From your written rules. If it has no name, that is itself the finding.
- Conditions met. Full or partial. This single field carries most of the value.
- Stake. The actual figure, so you can check it against your written fraction.
- Result. Won or lost, at the expiry you chose.
- State. One word: calm, annoyed, keen, restless.
Setup and reason
Naming the setup forces it to exist in your rules. If you regularly find yourself writing something vague, that is a gap in the written plan rather than a shortcoming of the log, and filling it is a rule-writing job.
Result and size
The stake column exists to catch drift. Under fixed-fraction sizing every figure should sit close to the same proportion of your balance, and any that do not mark a decision you made in the moment. Those entries are worth reading alongside the state column.
Two things do not belong in the log. Long narrative notes about what the market was doing, which take too long and are rarely read again, and any assessment written after the outcome is known, which is contaminated by the outcome. Keep the record to what was true at the moment of entry, and leave interpretation for the weekly review where you can see several trades at once.
Emotion at entry
One word, written before the contract resolves so the outcome cannot colour it. Sorted after a month, this field usually names the habit costing you most, which is a strong return on a single word per row.
Also log the setups you saw and declined. Skipped trades are the only way hesitation becomes visible, and reviewing them tells you whether your filter is too tight or your nerve is failing.
Six fields, filled in during the session, plus a line for every setup you declined.
Review the journal
The log does nothing until it is read, and reading it on a schedule matters more than reading it thoroughly.
Pick a fixed day and keep it, whether or not the week felt eventful. Reviews triggered by bad weeks produce changes driven by bad weeks.
Weekly analysis
Read the week in order rather than sorted, so the sequence is visible. You are looking for three things: trades marked partial, gaps where the time between entries compressed, and any stake that departed from the written figure. Each points at a different failure.
Spotting leaks
A leak is a repeated small loss with a common cause. Group your losses by setup name and by the conditions-met field. If most of your losses sit under one setup name, that is a rule to reconsider. If most sit under partial rather than full, the rules are fine and the following is not, which is a completely different problem with a completely different fix.
| What the review shows | Most likely cause | Response |
|---|---|---|
| Losses concentrated in one setup | That rule is weak in current conditions | Suspend it for a month and note the date |
| Losses concentrated in partial entries | Rule following, not the rules | Tighten the pre-entry check, change nothing else |
| Compressed gaps between trades | Tilt after a loss | Enforce a pause after any losing contract |
| Stake variation | Sizing by feel | Write the figure down before the session |
Resist the urge to change something every week. A review that finds nothing wrong is a useful review, and a rule set edited every seven days never accumulates enough consistent trades to be judged. If the week produced no clear finding, write that down and leave the rules alone; the discipline of not adjusting is part of what the review is training.
Refining rules
Change one thing, date the change, and treat the record as split at that point. Two simultaneous changes leave you unable to say afterwards which one mattered, which wastes the month that follows.
Fixed review day, one dated change. Separate rule problems from rule-following problems before you edit anything.
Turn logs into edge
Over a few months the log stops being a record and becomes the thing your decisions are based on.
At that point you are no longer choosing rules from articles. You are choosing them from evidence about your own trading on your own instruments, which is the only evidence that applies to you.
Cutting weak setups
Suspending a setup is easier than fixing one and usually more informative. Take it out for a month, note the date, and see whether the record improves. If it does, you have found something. If it does not, the setup was not the problem and you have narrowed the search without breaking anything.
Reinforcing strong ones
The setup that survives review deserves more of your attention rather than a larger stake. More attention means preparing for it better: marking its zones earlier, watching the sessions where it appears, declining the trades that compete with it. Sizing changes belong to the sizing rules and should not be used as a reward.
A monthly summary at the top of the log helps more than it sounds. Four numbers on one line: trades taken, proportion marked full, sessions within all limits, rule breaks. Twelve of those lines make a year of trading readable at a glance, and the trend in them tells you whether you are improving as a trader independently of what the market happened to do.
Measuring progress
Progress is measurable without reference to your balance, which is useful because a balance moves for reasons unrelated to your improvement. Count the proportion of entries marked full, the number of sessions that stayed within all three limits, and the number of rule breaks. All three should trend in one direction, and all three are within your control.
- Full-entry proportion. The cleanest measure of discipline.
- Sessions within limits. Counts the thing that protects you.
- Rule breaks per week. Should fall even in a losing month.
Measure discipline rather than balance. Full entries, sessions within limits, rule breaks per week.
Journal takeaways
Journalling is the cheapest improvement available in this activity and the one most consistently skipped.
Data beats memory
Your recollection of a month of trading is a story shaped by the most dramatic sessions in it. A column of thirty rows is not a story, and that is exactly why it is useful. Everything else on this site becomes checkable once you have one.
Review regularly
A short log read weekly beats a detailed one read never. If the format is taking more than a few seconds per trade, cut a field rather than skipping sessions, because an incomplete record during your worst week removes the data you most needed.
- Six fields, filled in live. Not reconstructed.
- Skipped setups logged too. Hesitation is otherwise invisible.
- Fixed review day. Regardless of how the week went.
- One dated change at a time. So the record stays readable.
Improvement is measurable
No figures are offered here about how much journalling helps, because this desk has measured nothing and any number would be invented. What is structurally true is that you cannot assess a method without a record of how it was applied, and that most traders who believe they are following their rules discover otherwise within two weeks of writing things down. You can start the log on virtual funds today, and the log will be more useful to you than the trades in it.
Start the log before you refine anything else. Nothing on this site is checkable without it.
What readers ask about this setup
What should a trading journal include?
Six fields are enough: entry time, setup name, whether every condition was met, stake, result and one word for your state at entry. Add a line for every setup you saw and declined, because skipped trades are the only way hesitation shows up. Longer formats tend to be abandoned during the sessions that would have been most worth recording.
Should I journal demo trades as well as real ones?
Yes, and the practice account is the best place to build the habit because there is no cost to the honesty. The frequency problems, over-trading, boredom entries and drift from the plan, all appear there in recognisable form. What does not transfer fully is the emotional weight of a funded loss, so keep the two records separate rather than merging them.
How often should I review my journal?
Weekly, on a fixed day, whether or not the week felt notable. Reviews triggered by bad weeks produce changes driven by bad weeks, which is how rule sets acquire conditions nobody designed. Read the entries in order so the sequence is visible, then change at most one rule and date the change.
Can I tell whether my rules or my discipline are the problem?
That is exactly what the conditions-met field answers. Group your losses by whether the entry was full or partial. Losses concentrated in full entries point at the rules; losses concentrated in partial ones point at rule following, which needs a tighter pre-entry check rather than a new rule. Confusing the two leads traders to redesign methods that were working.