How to Build a Daily Trading Journal (With Examples)
Build a trading journal that actually improves your trading: what to log per trade, how to grade decisions vs outcomes, and how to turn notes into rules.
Most trading journals are diaries. They record what happened ("shorted gold at 4,346, lost $90, felt frustrated"), and then they get abandoned after three weeks because they never change anything.
A journal that actually improves your trading works differently. It grades what you decided against what the market did, it separates good decisions from lucky ones, and it forces the mistakes you keep making into written rules. That last job is the one that matters. A mistake you notice but never write down as a rule will be back next week wearing a different setup, and you'll "discover" it all over again.
Here is what to log, how to review it, and how to close the loop, using the actual review process behind Open Market Journal's daily session reviews.
A journal is a loop, not a log
The loop has four stages. Skip any one and it stops working:
- Prepare. Write down what you expect before the session: regime, key levels, bias, and the one thing that would prove you wrong. (How to prepare for a session covers this.)
- Trade. Execute against the plan.
- Review. After the close, grade the prep against real candles and audit every decision.
- Feed forward. Turn the review's findings into tomorrow's rules.
If your journal stops at step 3, you're collecting evidence and never using it. Everything useful happens at step 4.
Log conditions required vs conditions met
The most useful thing in a journal is not the outcome. It's a small table that makes you honest about whether the setup you planned to take was actually there when you clicked buy.
For every trade, write the conditions your plan required, then mark which ones were genuinely present:
| # | Condition the plan required | Met? | Evidence |
|---|---|---|---|
| 1 | Price at a level tested 2+ times | YES | 4,346 rejected twice last week |
| 2 | Rejection candle confirming the level held | NO | Entered on arrival, no rejection yet |
| 3 | Higher-timeframe trend agrees | PARTIAL | D1 down, H1 turning |
Now the verdict is obvious. In that example you entered with one of three conditions fully met. Win or lose, you took a trade your own plan didn't sanction. That's the finding, and it's more valuable than the P&L.
This table catches the single most expensive habit in retail trading: arriving at a good level and treating the arrival as the signal. A losing gold short in the source desk got summed up this way, and it's worth pinning above your screen:
The level is not the trade. The reaction at the level is the trade. Price arriving at a sell zone during a strong rally is not a sell signal. It's a place to start watching for one.
Grade the decision and the outcome separately
This is the idea that changes everything, and almost nobody does it. Score the decision and the result on two different axes.
A trade can be a good decision that lost, or a bad decision that won. If you only track P&L, the bad-decision-that-won trades quietly reinforce the exact behaviour that eventually empties the account. So put every trade in this grid:
| You took it | You skipped it | |
|---|---|---|
| Rules said take it | Correct execution | Execution miss: why did you hesitate? |
| Rules said skip it | Rule violation, flag it even if it won | Correct skip |
Watch the bottom-left cell. A rule violation that happened to win is the most dangerous entry in the whole journal, because your brain files it as a success and orders seconds. Log it loudly.
A real review of a break-even day in the source desk signed off with "Decision quality: GOOD, Outcome quality: NEUTRAL." That's a healthy entry. It tells you the process held even though the P&L didn't move.
Journal your no-trade days too
Most people only write in the journal when they trade. That's backwards. Zero trades is not zero data. Most of the improvement hides in the setups you didn't take.
Every day, walk the candles and ask what could have triggered, whether you saw it, and if you passed, whether the pass was right. Both answers teach you something:
- You skipped a valid setup out of fear. That's an execution miss to work on.
- You skipped correctly and the "setup" would have lost. That's a correct skip, and logging it builds the nerve to keep skipping.
If you only journal on days you trade, you train yourself to believe trading is the productive act. Plenty of days, the best trade was the one you didn't take.
Grade your preparation, not just your trades
Your trades are only as good as the map you drew beforehand. So audit the map. After the session, pull the levels and bias you wrote that morning and check them against what price did.
The metric worth tracking is "useful when reached." A key level only counts as useful if price actually got there and it did what you predicted: held as forecast, swept and recovered as flagged, or broke cleanly in a scenario you'd named. A level price never reached is neither right nor wrong. Drop it from the count.
A 12-day audit of exactly this in the source desk produced a baseline worth knowing:
Of every 10 levels a prep lists, roughly 5 get reached, and about 4 of those produce a usable signal.
That same audit graded level types by how much signal they carried, and the results are the kind of thing you only learn by keeping score. Asia-session highs and lows got reached 83% of the time and paid off almost every time (keep them). Weekly structural levels sitting more than twice the average daily range away got reached once in twelve days (stop drawing them). You cannot tell which of your own analysis is signal and which is decoration until you've measured it against candles for a couple of weeks. Almost everyone is carrying decorative levels they'd bin if they checked.
Turn observations into rules, or they don't count
A journal full of observations that never become rules is just a tidier diary. The bar for a finding is simple: specific enough to act on tomorrow, or it doesn't ship.
Compare these:
- Weak: "Lost $90 shorting gold on March 26." That's a fact, not a rule.
- Weak: "Be more patient." Too vague to execute.
- Strong: "After a US data release, wait for 3 closed 15-minute candles before entering. The first two are noise. (Cost so far: 3 losing trades on data spikes.)"
The strong one names the trigger, the action, and the evidence. You can follow it tomorrow without interpreting anything.
Don't overreact to a single trade, though. Use a 1 → 2 → 3 rule. One bad trade is a story. Two with the same root cause is a lesson worth writing down. Three is a confirmed habit that has to become a hard rule. That keeps you from rewriting your whole system after one ugly day, while making sure a genuine pattern can't hide.
Keep a pattern ledger
Here's a failure mode the daily journal alone won't fix. You notice the same mistake Monday, again Wednesday, again the following Tuesday, each time buried in that day's notes. Because it's scattered, you never see it's the same mistake, so it never gets fixed.
The fix is one running ledger: a single file where each recurring mistake gets a stable name, a one-line note on why it happens, and a dated list of every time it shows up. When the count hits your threshold, it graduates into a rule.
A real leak this caught: five gold trades over one stretch, netting a loss, where three of them were the identical short at the same supply zone, re-armed three times. In the daily notes each looked like a separate trade. In the ledger, "re-arming the same rejected level" showed up as a counted, named habit, and only then did it get a rule. A daily note is disposable. The ledger is permanent, and the count is what finally forces the fix.
What a good journal finds on its own
Log your entries, stops, targets, and the best price each trade reached before you exited, and the journal starts answering questions you didn't think to ask.
One week's review in the source desk turned up something the daily P&L completely hid. The week was roughly flat, but every single losing trade had been in profit first, some by several times the risk, before reversing to stop out. The entries were fine. The exits were the problem: stops parked inside the market's normal noise, targets too greedy to ever get paid. The fix that fell out was mechanical: take half off at a partial target instead of holding everything for all-or-nothing. Run over the same trades, that one change turned a flat week into a clearly positive one.
You'd never feel your way to that. You find it because the journal wrote down the best price every trade reached, and the pattern was sitting there in the numbers.
Your weekly review
Daily entries catch behaviour. The weekly review catches the statistics a single day can't show:
- Win rate by direction. Are your shorts dragging down perfectly good longs?
- Average win vs average loss. Your real reward-to-risk, not the planned one.
- Profit factor and expectancy. Is the system positive at all?
- Average hold time on winners vs losers. Holding losers longer than winners is the classic tell.
- Your three biggest findings, each pointed at a specific rule to change.
Every finding names the thing it changes. "Improve discipline" is a non-action. "Move the stop outside the 15-minute average range on gold" is a change you can make.
A daily template you'll actually fill in
Keep it short enough that you use it every day:
Date / Instrument:
Prep bias (from this morning):
What actually happened (Asia -> London -> NY, one paragraph):
TRADES
- Setup:
- Conditions required vs met (table):
- Decision cell (correct exec / exec miss / rule violation / correct skip):
- Entry / stop / target / exit:
- Best price reached before exit:
- Outcome (currency and % of account):
NO-TRADE SCAN
- Setups I passed on, and whether the pass was right:
PREP GRADE
- Levels reached and useful (x of y):
- Was the bias right?
LESSON (specific, or leave it blank)
- Trigger -> action -> evidence:
- Ledger tag (if it recurs):
Decision quality: GOOD / OK / POOR
Outcome quality: POSITIVE / NEUTRAL / NEGATIVE
See it live, and go deeper
This is the loop behind every session review on Open Market Journal. Each one grades that morning's session prep against what the market actually did, in public. Reading a few is the fastest way to see the format working on real days.
To build the front half of the loop, start with How to Prepare for a EURUSD Trading Session. To understand what's actually moving the market you're journaling, read What Moves EUR/USD.
If you'd rather watch the loop run before you build your own, subscribe to the free journal. Daily prep and review land in your inbox, no paywall, with the reasoning shown.