Guide · Journaling & review

The trading journal practice: grade the process, not the profit

Most traders keep a spreadsheet of entries, exits and profit or loss. That is a scoreboard. A journal is a sensor, and the difference is the whole game.

A first-week reader of any course is told to keep a journal, handed a template with columns for entry, exit and P and L, and told to fill it in. Some do. Almost none keep it up, and the ones who do rarely improve, because they built a scoreboard when they needed an instrument. This guide is about the instrument: what a journal is actually for, exactly what to log, how to grade a trade on the part you control rather than the part luck controls, the cadence that turns entries into learning, and a distinctive habit, the intervention log, that lets you treat your own improvement as an experiment. If you want the ready-made tools, the trade journal grader scores a trade against these criteria and the India journal template gives you the fields; this page is the method behind both.

1. Why the journal is the feedback loop

You cannot improve what you do not measure, and trading is unusually hostile to measurement by memory. Memory is reconstructive: every time you recall a decision, your brain rebuilds it using what you now know, so the outcome quietly rewrites the reasoning. This is the documented mechanism behind hindsight bias, the sense that a move was obvious once it has happened. In a trading context it produces a specific distortion: you remember the wins as skill, explain the losses as bad luck, and come away believing you knew more at entry than you did. Left to memory, your track record is a flattering novel, not a dataset.

A journal breaks that loop by fixing each decision in writing before the result is known, then letting you aggregate across many trades. That aggregation is where the value sits, because the errors that quietly cost the most are invisible inside any single trade and only appear in the pile: overtrading on quiet days, oversizing after a win, drifting the stop, the revenge trade that follows a loss. None of these announce themselves once. All of them are obvious across fifty entries. The journal is the only tool that makes the sample legible, and the sample is where the leaks live.

Research on expert performance is blunt about why this matters. Ericsson and colleagues defined deliberate practice as work on well-defined tasks with immediate, informative feedback and repeated chances to correct errors. Trading hands you endless repetition but withholds the feedback, because outcomes are noisy and arrive late, so raw screen time does not compound into skill. The journal supplies the missing half: a clean, immediate signal at the close of each trade, and a scheduled confrontation with your recurring errors. Without it you are practising, but not deliberately.

2. What to log, and why each field earns its place

A good entry is not long, but it is specific, and every field is there to answer a question you will ask later. The trap is logging only the numbers, entry, exit, profit, because those are the two facts that carry the least information about whether you traded well. What you want is the texture of the decision and its context, captured densely enough that a version of you three months from now can reconstruct not just what you did but why, and whether it was sound.

Split the fields by when they are written. The pre-trade fields, the thesis, the setup, the planned entry, stop, target and size, and the risk in rupees and as a fraction of equity, must be recorded before you place the order. That timing is not fussiness. It is the one guard against narrative fitting: a plan written after the outcome is a story, a plan written before it is evidence. The post-trade fields, the actual fills, the exit reason, the emotional state, and the honest verdict on whether you followed your rules, are written at the close.

3. Grade the process, not the profit

Here is the method most journals miss. At the close of each trade, give it a grade, A, B or C, for execution quality alone: did you follow the plan, size correctly, honour the stop, and resist overriding the trade mid-flight. The grade is deliberately blind to the outcome. A trade that obeyed every rule earns an A whether it made money or lost it. A trade that broke the rules earns a C whether it won or lost. You are scoring the decision, not the result.

The reason to separate them is that the outcome contains luck and the process does not. In decision science this confusion has a name: Annie Duke calls it resulting, the error of judging a decision only by how it turned out. A good decision can lose and a bad decision can win, because between the choice and the result sits a market you do not control. The process, the plan, the sizing, the stop, the restraint, is the part that is entirely yours, and it is the only part you can repeat on purpose. So it is the only sane thing to grade.

This produces the insight that reorganises a trader's whole relationship with the journal: an A-grade losing trade is better than a C-grade winning trade. It sounds backwards until you remember what a journal is for, which is training behaviour. The market pays out on the outcome, not the reasoning, so a rule-break that happens to win is a bad decision that just got rewarded, and a rewarded behaviour gets repeated. Praise the C-grade win and you are quietly teaching yourself that breaking rules works. The next time the same break meets an ordinary loss, and it will be a bigger loss, because the false confidence talked you into sizing up. Outcome-grading corrupts learning by attaching the reward to the wrong thing. Process-grading keeps the reward attached to the behaviour you actually want to build.

Grading needs a rubric you apply the same way every time, or it drifts into a mood. Keep it simple and criterion-based, so any given trade lands on a grade by its facts, not by how you feel about the money.

Over time the number that matters is your grade distribution, the share of A, B and C trades, not your win rate. A win rate is an outcome statistic that swings with luck over any short run, so it cannot tell you whether you executed well this week. A grade distribution is a process statistic: it becomes legible over far fewer trades because it does not wait on the market to settle, and it measures exactly the thing you can change on purpose. You still track outcomes over a large sample, but you steer on the grades.

4. The review cadence

A journal that is written but never read is a diary. The learning lives in the review, and the review works on three clocks, each answering a different question.

At the close of each trade, you grade it, while the detail is still warm. This is the sensor reading: fast, immediate, and the raw material for everything above it. Weekly, in about half an hour, you read the week's grades and every rule-break, look for the error that showed up more than once, and choose the single highest-leverage change to test next week. Monthly, you aggregate across a larger sample and ask the questions a week is too short to answer: which setups actually earn their place, which times of day help or hurt you, which position sizes precede your worst grades, and which emotional states, poor sleep, a prior loss, boredom, cluster with your C trades. These cross-trade patterns are inaccessible inside any single entry and obvious across a month of them.

The per-trade grade is the sensor, and the weekly and monthly reviews are where a sensor reading turns into a decision. Skip the reviews and the entries pile up producing nothing; keep them and your mistakes get a scheduled, deliberate confrontation instead of a vague resolution to do better. That scheduled confrontation is the immediate, informative feedback the deliberate-practice research says skill actually requires.

5. The intervention log

Here is the practice that separates a trader who is measuring from a trader who is experimenting. Do not log only trades. Log the changes you make to your own process, each with a date. When you add a rule, cap your size after two losses, or impose a cooling-off period following a stopped-out trade, you write the intervention down as its own dated entry. Then you watch your grade distribution over the following weeks to see whether it moved.

This turns self-improvement from a vague intention into an experiment with a clean before and after. Most traders "try to size down" or "work on their discipline" and have no idea, a month later, whether it helped, because there was never a marked start line. An intervention log gives you one. You can look at the weeks before a size cap and the weeks after and read the effect off your own grades: did the share of C trades fall, did the oversizing entries stop appearing. If the change helped, you keep it; if it did nothing, you drop it and try another, rather than accumulating a stack of rules you no longer remember the reason for.

Read as a whole, the practice is a loop you run on yourself: the journal measures the process, the review finds the leak, the intervention tries a fix, and the next stretch of grades tells you whether the fix worked. That upstream discipline, deciding the edge, the invalidation level and the size before the trade and then holding yourself to them, is exactly what the method we teach is built around, because the order type and the entry are the easy half; grading and correcting your own execution is the half that compounds.

6. The failure modes, and their antidotes

Three habits quietly defeat a journal, and each has a specific fix. Confirmation journaling writes rich, flattering entries for the winners and dismisses the losers as bad luck, which reinforces your self-image and blinds the sensor; the antidote is to grade every trade to the same standard and to log at least one loss per session in full detail. Narrative fitting reconstructs a tidy reason for the trade after the outcome is known, so the entry becomes fiction that favours you; the antidote is the pre-trade fields, written before the result exists and therefore impossible to retrofit. Performative journaling writes for an imagined audience, a future student or a memoir, and the entries grow long, polished and dishonest; the antidote is to keep the journal private and write in short, blunt fragments.

Where this fits

The stakes are not abstract. SEBI's updated study in September 2024 found that 93 percent of individual traders in equity derivatives lost money across FY22 to FY24, with aggregate losses exceeding 1.8 lakh crore rupees, and its July 2025 study found that individual traders' net losses widened further to about 1,05,603 crore rupees in FY25, roughly 91 percent of them in the red. Those are the odds against undisciplined trading, and no journal changes the market. What a journal changes is whether your own repeated errors stay invisible or become findable, and whether the errors you find get corrected on a schedule or forgotten by the next session.

Read plainly, the journal does not make you money and this guide does not claim it will. It is the instrument that lets deliberate practice happen in a domain that otherwise withholds the feedback: it fixes the decision before the outcome can rewrite it, grades the part you control, and forces a scheduled confrontation with the leaks. The template you use is incidental. The loop, log, grade, review, intervene, is the practice.