In October 2014, France's AMF examined four years of CFD and forex trading at the country's main brokers. The verdict fits in one line: a majority of clients lost money, with a significant median loss. How many of them could have explained why? That is what a trading journal is for first and foremost: knowing.

Write down the hypothesis before you enter

Most journals are filled out in the evening, after the position has been closed. That is too late. Once the outcome is known, memory rewrites the story: the winning trade becomes “obvious,” the losing trade “unlucky.” So record the essentials before you press the button. One minute is enough.

  • The setup, under a fixed name: “Asian range breakout,” “pullback to the moving average.” Always use the same name, or you will not be able to sort anything.
  • The reason, in one sentence. If it cannot fit in one sentence, the trade is not ready.
  • Entry, stop, target, at the exact price.
  • The risk in euros. This is your unit of measurement: R. A full loss is −1R; a gain equal to twice the risk is +2R. In R terms, a gold trade and an EUR/USD trade become comparable.
  • The context: a release on the economic calendar within the hour, a quiet session, a trending or range-bound market.
  • Your state, from 1 to 5. Fatigue, rushing, the urge to win back losses: these can be measured too.

Risk is set with the stop and position size, never with leverage. The European caps—30 to 1 on major currency pairs, 20 to 1 on gold and major indices—tell you what the broker allows you to use. Not what you should commit.

Then give the verdict, without leniency

After the exit, complete the record.

  • The result in R, including fees.
  • Plan followed: yes or no. One box, but the most useful one in the journal.
  • Any mistake, chosen from a short, fixed list: early entry, stop moved, premature exit, trade outside the plan, excessive size.
  • The extreme excursions. How far did price move against you before the exit (the MAE, in the terminology used by English-speaking traders)? And how far in your favor (the MFE)? The first tells you whether your stops are too tight. The second tells you whether your targets leave money on the table.
  • Two screenshots, at entry and exit.

Always separate the decision from the outcome. An out-of-plan trade that wins is still a mistake. A compliant trade that loses is still a successful trade.

Trader reviewing printed trade screenshots on a table, with powered-off screens in the background

Five numbers, no more

A spreadsheet can spit out thirty indicators. Five are enough to guide your trading.

Win rate. The percentage of winning trades. On its own, it says nothing: an attractive win rate often hides small gains and large losses.

Gain/loss ratio. The average win divided by the average loss, in R. It is the counterpart to win rate: you should never read one without the other.

Expectancy. What a trade returns on average, in R. Multiply the win rate by the average win, then subtract the loss rate multiplied by the average loss. If it is positive, the method makes money over time, provided you apply it. If it is negative, no amount of discipline will save it.

Maximum drawdown. The largest decline in the account from a high to the next low. It measures what you will have to endure psychologically. In a prop firm, it is also the limit that can terminate the account.

Longest losing streak. It prepares you mentally. A setup that wins less than half the time will sooner or later produce long losing streaks. If you know that in advance, you will not change methods at the worst possible time.

Calculate these figures by setup, not just for the account as a whole. A good setup can hide a bad one for months.

How many trades before you judge?

One good week proves nothing. One bad week does not either. And years of experience do not guarantee skill. In their study covering all day traders in Taiwan, Brad Barber, Terrance Odean, and their co-authors conclude that only a very small proportion of practitioners are predictably profitable. In Brazil, a study of all beginners who traded index futures during their first day-trading period between 2013 and 2015 concluded that the vast majority of those who persisted lost money. Persistence alone is not enough. You have to measure.

Set your sample size before drawing conclusions—for example, several dozen trades per setup, taken under the same conditions. Below that, you are observing. Beyond it, you are making a judgment. And every time you change a rule, reset the count to zero: you are now testing a different method.

One week tells a story. A sample delivers a verdict.

The four-step weekend review

Block out a fixed time slot, with the broker platform closed. Less than an hour is enough.

  1. Start with the out-of-plan trades. Count them and add up their result in R. This is often where a significant share of the money leaks away.
  2. Update the five figures, setup by setup.
  3. Look for one recurring pattern: a time of day, an asset, an emotional state, or a type of mistake. Just one—the most costly.
  4. Write down one decision. One only, again. Three resolutions per week means zero resolutions kept.

From numbers to rules

A good rule reads like a test: “if…, then…”. It concerns an action you control, can be checked in the journal, and includes a review date. “Be more patient” is not a rule. Here are three:

  • “If a major release is due within the hour, I open nothing on the affected asset.”
  • “If I have taken two full losses during the session, I close the platform.”
  • “Once placed, the stop never moves farther away.”

Let us see how a review leads to a rule.

The setup that won most often was not the one that made money. Without a journal, this trader would probably have increased exposure to B and abandoned A in the middle of its worst losing streak.

All of this can be done by hand, with a spreadsheet, or with a free trading journal template. It is tedious, and that is often where people drop off. Edgyx imports trades, calculates these statistics by setup, and its Ora coach flags recurring patterns during the review. The tool matters less than consistency.

What to remember

  • Record the setup, the reason, the stop, and the risk before entering: afterward, memory cheats.
  • Measure everything in R and check on every trade whether the plan was followed.
  • Track five figures per setup: win rate, gain/loss ratio, expectancy, maximum drawdown, and longest losing streak.
  • Do not judge a method until you have a sufficient sample, and reset the count to zero whenever you change a rule.
  • Take one rule from each review, written as “if…, then…”, with a review date.