What a journal is for
A trading journal exists to answer questions about your own trading with evidence instead of memory. Which setup pays? What does breaking your own rules cost? Are your losses the size you planned? Memory keeps the dramatic trades and loses the ordinary ones, and the ordinary ones are most of the record.
That purpose decides what to write down. Record what you will want to count, sort or compare later, and skip what you never will. A short record filled in every time is worth more than a long one filled in when you feel like it.
What to record for every trade
Every trade carries two kinds of information. The first is what the platform already knows: instrument, direction, open and close times, prices, size, commission, swap and net profit or loss. Import these from your platform's history file rather than typing them. Typing introduces errors, and an error in a price or a size quietly corrupts every statistic built on it.
The second kind is what only you know, and it has to be written at the time. These fields turn a list of fills into a journal:
| Field | Example | Why it matters |
|---|---|---|
| Setup | Pullback to 20 EMA | Lets you compare one approach with another. Choose from a fixed list of five to eight names; free text cannot be counted. |
| Planned stop | 1.0825 | Where the idea is wrong, decided before entry. Moving the stop later does not change it. |
| Planned risk | $200 | The loss if the planned stop is hit. It is the unit for R-multiples. |
| Reason | Retest of the broken 1.0840 high at the London open | One sentence, written before entry. Later it shows whether the trade matched the setup. |
| Mistakes | Moved stop | Tags from a fixed list. They make the cost of rule-breaking measurable. |
| State | Rushed | One word, optional. Useful only if you use the same few words every time. |
Planned risk is plain arithmetic. A long EUR/USD trade entered at 1.0850 with a stop at 1.0825 risks 25 pips. At 0.8 standard lots and US$10 per pip per standard lot, the planned risk is 25 × 0.8 × 10 = $200. The position size calculator runs the same sum the other way: start from the dollar risk and it returns the size.
Keep emotions brief. One word from a short list — calm, rushed, bored, frustrated — can be sorted later. A paragraph about how you felt cannot, and it is rarely read again.
Planned risk and R-multiples
An R-multiple expresses a result as a multiple of the risk you planned. R is the planned risk, and the R-multiple is the net result divided by it.
R-multiple = net P&L ÷ planned risk
Dollar results mix two things: the quality of the trade and the size you chose. R removes the size. Take three trades:
| Trade | Planned risk | Net P&L | R-multiple |
|---|---|---|---|
| A | $200 | +$460 | +2.3R |
| B | $50 | −$50 | −1.0R |
| C | $400 | +$200 | +0.5R |
| Total | +$610 | +1.8R |
In dollars, trade C looks like a solid win. In R, it earned half of what it put at risk, while trade A earned more than twice its risk. Across the three trades the average is +0.6R per trade. Once every trade is in R, you can compare a week of small positions with a week of large ones, or one setup with another, on the same scale.
R also exposes execution problems. A loss larger than −1R means the stop was not honoured, or the market gapped or slipped through it. Count how many losses go beyond −1.2R each month. A steady number of them points at execution rather than at the setup.
One rule keeps R honest: record planned risk before the result is known. Planned risk filled in afterwards tends to drift toward whatever makes the result look reasonable.
How to tag mistakes
A mistake tag records that you broke one of your own rules. It is not a label for a losing trade. A loss of −1R on a valid setup is a cost of doing business; a win taken after moving your stop is still a mistake.
Write the list before you need it, and make every tag observable — something another person could confirm from the record. For example:
- Moved stop — widened the stop after entry.
- No setup — took a trade that is not on your setup list.
- Oversized — planned risk above your own per-trade maximum.
- Early exit — closed before stop or target without a rule that called for it.
- Revenge trade — entered outside your plan shortly after a loss.
- News window — traded inside a news window you had chosen to avoid.
Tag on the day the trade closes. One trade can carry several tags. Avoid vague tags such as “bad discipline”: they cannot be checked, so they end up on every losing trade and tell you nothing.
Measuring the cost of mistakes
With tags in place, the cost of each mistake is a sum. Group the month's trades by tag and add up their net results. Here is a month of 40 trades:
| Group | Trades | Net P&L | Average |
|---|---|---|---|
| No mistake tag | 31 | +$1,860 | +$60 |
| Moved stop | 6 | −$1,140 | −$190 |
| Revenge trade | 3 | −$570 | −$190 |
| All trades | 40 | +$150 | +$3.75 |
The nine tagged trades lost $1,710 between them. The 31 clean trades made $1,860, and the month finished at +$150. That comparison is one of the most useful numbers a journal produces, because it points at behaviour you control.
Taking trades out of a record is not quite the same as not making the mistake, so use a sharper measure where one exists. For stop-related tags, compare the actual loss with the planned risk. Suppose the six moved-stop trades were all losers with a planned risk of $150 each. Honouring the stops would have capped the losses at $900. They lost $1,140, so moving stops cost $240 beyond plan, or 1.6R.
A spreadsheet can do this with a filter and a sum. Aurith Journal Pro shows it as a cost-of-mistakes breakdown by tag.
How to review weekly
Pick a fixed time when markets are closed and keep the review to about 30 to 45 minutes. Following the same steps every week stops it turning into a mood-driven reread of your worst trade.
- Complete the record. Import the week's history. Every trade needs a setup, a planned risk and its mistake tags, or an explicit “none”. Fix gaps before you look at any totals.
- Read the numbers. Net P&L, number of trades, win rate, average R and the largest loss in R.
- Sort by R-multiple. Look at the three worst trades. Decide whether each was a planned loss or a mistake, and check that the tags agree.
- Group by setup. Note the total R and the trade count for each one. Six trades of a setup tell you very little; treat small groups as notes, not conclusions.
- Add up the mistake cost for the week, tag by tag.
- Write one change for next week. Make it specific and checkable: “no entries in the first 15 minutes after the open”, not “be more patient”.
- Check last week's change. Did you keep it? A record of kept and broken changes belongs in the journal too.
Once a month, run the same review over the whole month. Decisions about keeping or dropping a setup belong there, where the samples are larger, rather than in a weekly review after a bad Friday.
Keeping it going
Journals are abandoned when they take too long. Import instead of typing. Fill in the decision fields as you trade, while the reason is still fresh. Keep the setup and mistake lists short enough to remember. When the record has some depth, work out its expectancy in R, and check a planned trade's reward against its risk with the risk to reward calculator.