Cómo llevar un diario de trading que cambie tu forma de operar

Most sheets in this survey ask you to check claims against regulators, calculators, or published numbers. This one builds the instrument for checking claims against yourself. A trading journal is not a diary and not homework — it is the place where your opinions about your own trading go to be checked, and it is the only tool in the entire curriculum that can tell you which problem you personally have. The version specified here costs about two minutes per trade. Here is how to build it, field by field, and what to do with the rows once you have them.

Why memory fails as a record

Ask a trader without a journal how last month went and you will get a story — usually a fair one about bad luck, near misses, and a method that mostly works. The story is sincere. It is also constructed by a memory that stores vivid things over representative things and quietly reframes whatever stings. The wins stay sharp, the losses become anomalies, and the violations become reasonable exceptions. Nothing in that machinery is dishonest; it is simply not a measuring device.

Every bias in the survey's psychology region survives on exactly this gap. Recency says the method is broken; nothing argues back. Overconfidence says size up; nothing shows what happened last time. The journal is the argument back — a record written at the moment of action, before the self-editing starts, by the only witness who saw every trade. The survey's report on why most traders lose money names the absence of process as a documented cause of failure; the journal is the cheapest piece of process that exists, and the only one that doubles as evidence about the other causes.

Step one: set up the ten fields

Open a spreadsheet and make ten columns. This is the minimum set that catches the problems worth catching — fewer hides the behaviour, more kills the habit:

Field 7 needs the one piece of arithmetic this sheet asks of you. R is the outcome divided by the planned risk: a trade risking $50 that makes $75 is +1.5R; the same trade hitting its stop is −1R. The unit is symmetric by construction — winners and losers are measured against the same planned risk — and it survives account growth and size changes, so rows from January remain comparable in June.

Field 9 looks soft and is not. One word about your state, written at entry — “calm,” “bored,” “urgent,” “tired” — costs three seconds and builds the dataset no platform can export: the correlation between how you felt and how you traded. A month in, sorting trades by state note routinely produces the journal's most uncomfortable finding, usually some version of “every trade tagged urgent lost.” Field 10 borrows the pre-commitment trick from the biases sheet: one line on what would prove the trade wrong, written while you can still afford to believe it.

Step two: write planned versus actual, honestly

Fields 4 and 5 are the journal's engine, and the rule is mechanical: planned numbers are written before the trade, actual numbers at the close, and neither is ever edited afterwards. The gap between them is where every behavioural problem in this survey becomes visible. Entered eight pips above the planned level — that is chasing. Exit happened forty pips below the planned stop — the stop was moved. Target was 1.5R, closed by hand at +0.4R — a winner cut. Each gap is one of the biases from the previous sheet, caught on camera and timestamped.

This is also why the journal records ▲realized P/L — closed, final outcomes — and treats ▲unrealized P/L as scenery. An open position's running number changes by the second and trains exactly the wrong reflexes; the row is completed when the trade is closed and the outcome is a fact.

Two honesty rules keep the engine running. First, the journal is written at the moment of action — entry fields at entry, exit fields at exit — never reconstructed on Sunday, because reconstruction is memory, and memory is the problem being solved. Second, violations get logged with particular care, in plain words, without the cushioning adjectives. A journal that flatters is strictly worse than no journal: it carries the authority of a record while containing the fiction of a story.

Step three: run the weekly review

Rows nobody reads are a ritual, not an instrument. The review that turns them into information is deliberately small — thirty minutes, once a week, same time every week:

1. Read every row from the week. No analysis yet — just restore the week as it actually was, before memory's edit ships.
2. Answer three questions from the rows: Which trades followed the plan? Did plan-followers outperform violations? Which single rule was broken most often, and under what state note?
3. Change at most one rule, in writing, dated. One change keeps cause and effect legible; three changes is a new plan every month, which is no plan at all.

The discipline of one change is harder than it sounds — a bad week generates an urge to renovate everything. Resist it with the same logic as everywhere in this survey: the review is a procedure precisely so that the week's feelings do not get to redesign the system that has to outlive them.

Step four: turn rows into numbers

After a month or two, the journal starts paying in a second currency: statistics about you. Win rate is the share of rows that closed positive. Average win and average loss come straight from the R column. Together they make the one number every later sheet in this survey leans on — expectancy, your average result per trade:

Read the symmetry in that worked pair: a small shift in the loss column, the kind a moved stop produces, is the entire difference between a system that earns and one that bleeds. At $50 risked per trade, +0.08R is about $4 per trade; −0.03R is about −$1.50. Twenty trades is far too small a sample for a verdict — treat early numbers as a sketch, not a sentence — but the direction of the arithmetic is exactly what no memory can supply. Your ▲equity curve and your worst ▲drawdown, both computed from the R column, complete the picture.

Step five: choose tools last

Start with a spreadsheet. It costs nothing, the ten columns take five minutes to set up, and computing your own win rate by hand — once — teaches more than any dashboard. Journaling software earns its place only after the habit exists; adopted earlier, configuring the tool becomes the procrastination. The habit is the product. The tool is packaging. The same logic orders the whole workflow: rows this week, the review this weekend, statistics after a month — and only then, if ever, an upgrade to something with charts.