The 19-Column Trade Journal: Every Field Worth Logging, and Why
The short answer
A rich trade journal logs 19 fields per trade across four groups: the pre-trade plan, the execution, the post-trade outcome and costs, and the review. A thin journal that records only profit and loss tells you that you won or lost but never why. The rich schema exists so leaks become findable: which setups, sizes, times and emotional states actually cost you. The two load-bearing fields are the R-multiple, which normalises every result into units of risk, and the process grade, which judges the decision rather than the outcome. An A-grade loss beats a C-grade win.
This is the concrete, field-by-field schema. It is the companion to the practice and methodology page, which covers the habit of writing trades down and the review ritual, and to the Trade Journal Grader and the journal template, which are the tools. Here we do one thing: define each of the 19 columns, say what to log in it, and explain why it earns its place. A column that cannot answer a question you will one day ask is dead weight; every field below answers one.
Why 19 columns, not one
The instinct of a new trader is to journal the number that hurts or delights: the rupee result. That single column is the least informative thing you can keep. It confirms an outcome you already felt at the close, and it teaches nothing about the decision that produced it. A journal that is only a ledger of wins and losses is a diary of your moods, not an instrument for improvement.
The reason a broken feedback loop matters is not abstract. SEBI's July 2025 study of individual traders in the equity derivatives segment found that over 91 percent lost money in FY25, with aggregate net losses of about ₹1,05,603 crore across a client base near 96 lakh traders, wider than the previous year. A regulator does not publish a figure like that because people cannot read charts. They publish it because the retail feedback loop is broken: losses get mentally filed as bad luck, wins as proof of skill, and the self-model drifts until the account does. A rich journal is the friction that interrupts that drift, and it works only if it captures enough dimensions for a pattern to surface.
So the schema is built in four blocks, each answering a different question. The pre-trade block records the plan you committed to before the market could move: the decision, uncontaminated by its result. The execution block records what actually happened at the moment of entry and exit, so the gap between plan and reality is visible. The post-trade block records the outcome and its costs in comparable units. The review block records your judgement of the process, which is the part you can actually improve. Nineteen columns is not a number to worship; it is the smallest set that lets each of those four questions be answered honestly.
The 19 fields, one by one
Below is the full schema. Each row names the field, its group, what to log, and the question it lets you answer later. The instruments and prices in the worked rows further down are illustrative labels, not real securities or recommendations.
| # | Field | Group | What to log | Why it matters |
|---|---|---|---|---|
| 1 | Date / time | Pre-trade | Trade date and entry time in IST | The main axis for slicing later, by week, by month, and by time-of-day. Many traders discover one session window is net negative for them. |
| 2 | Instrument / segment | Pre-trade | What was traded and in which segment | Lets you separate cash from derivatives, index from single stock, so edge is measured where it actually exists. |
| 3 | Setup / strategy | Pre-trade | A tag from your own fixed list of setups | A controlled vocabulary, not free text, so trades aggregate mechanically. This is how you find the setup that is quietly break-even. |
| 4 | Market regime / context | Pre-trade | Trend, range, high or low volatility, event day | The same setup behaves differently by regime. Tagging context stops you blaming the setup for a regime mismatch. |
| 5 | Pre-trade thesis | Pre-trade | Two lines on why the trade is worth taking | A contemporaneous record of intent. It stops the after-the-fact rewrite of your reasoning. If you cannot write it in two lines, it is not yet a trade. |
| 6 | Planned entry | Pre-trade | The price you intended to enter at | Half of the plan-versus-reality comparison. The gap to the actual fill is where slippage and hesitation show up. |
| 7 | Planned stop | Pre-trade | The level that invalidates the idea | Entry minus stop defines one R, the risk unit for the whole trade. It must be the original stop, not one moved mid-trade. |
| 8 | Planned target | Pre-trade | The intended objective, a level or an R multiple | Trades logged without a target tend to be exited on emotion. The journal exposes that pattern over a sample. |
| 9 | Position size | Pre-trade | Quantity in shares, contracts or lots | Size times the stop distance is the rupee risk. It is the input the review checks against your declared risk ceiling. |
| 10 | Risk: rupees and percent | Pre-trade | Rupee risk and risk as a percent of equity | The percent, not the rupee figure, keeps sizing comparable as the account changes. It flags the drift where risk creeps up after a good run. |
| 11 | Actual entry fill | Execution | The price the order actually filled at | The honest input to the R-multiple math. Logging the intended price instead of the fill quietly flatters every result. |
| 12 | Actual exit fill + reason | Execution | The exit price and a reason code | Reason codes, target hit, stop hit, trailed, time stop, discretionary, let the review split exits by plan from exits by impulse. |
| 13 | R-multiple | Post-trade | Result in units of initial risk | The single most comparable number in the row. It puts a small position and a large one, or two instruments, on one scale. |
| 14 | Costs | Post-trade | Brokerage, STT and other charges | Costs turn marginal winners into losers over a sample. Netting them into R keeps the edge estimate honest. |
| 15 | Holding period | Post-trade | Time from entry to exit | Reveals whether your edge lives in fast trades or slow ones, and whether you hold losers longer than winners. |
| 16 | Emotional state / notes | Review | How you felt and anything structured columns missed | Where pattern recognition begins. Fear, boredom and revenge each leave a signature the structured columns later confirm. |
| 17 | Rules-followed flag | Review | A plain yes or no: did you follow the plan | Blunt on purpose and separate from the grade, so a broken rule cannot be blurred into a soft score. It is the discipline audit. |
| 18 | Process grade | Review | A, B or C on execution, ignoring the result | The column the review reads first. It measures the decision, the only thing you control. An A-grade loss outranks a C-grade win. |
| 19 | Screenshot / chart link | Review | A marked-up chart of the setup and exit | The visual record memory cannot edit. It is what makes a monthly review of your best and worst trades concrete rather than remembered. |
These fields are not arbitrary, and the four load-bearing ones deserve their own treatment: the R-multiple (field 13), the rules-followed flag and process grade (fields 17 and 18), and risk as a fraction of equity (inside field 10). Get those four right and the journal earns its keep even if the rest are logged loosely. Get them wrong and nineteen columns of neat data still cannot tell you whether you are any good.
The R-multiple: why results are logged in units of risk
Most journals record the rupee result and stop. That is the correct number for tax filing and the wrong number for measuring skill, because a rupee result is the product of two independent things: the quality of the trade and the size of the position. Mixed together, neither can be read cleanly. A large win might be a mediocre trade in a huge size; a small win might be an excellent trade sized cautiously. The rupee figure cannot tell them apart.
The R-multiple, a concept popularised by Van Tharp, fixes this. R is your initial risk, the planned-entry-to-planned-stop distance times the position size. Every result is then expressed in units of that R. A trade that loses the full stop distance is minus one R. A trade that returns three times the risk is plus three R. Because R is dimensionless, it does not care whether the position was one lot or twenty, whether the instrument was an index future or a mid-cap, whether the capital was fifty thousand rupees or fifty lakh. A plus two R trade on a small account is exactly as valid an edge data point as a plus two R trade on a large one.
That is what makes R the right unit for a journal. The average R across a sample, called expectancy, is a stable estimate of edge that sizing decisions cannot distort. A trader averaging plus 0.3 R per trade over 100 trades has an edge; the same trader's rupee total depends entirely on whether they risked ten thousand or fifty thousand per trade, which is a sizing question, not an edge question. The journal stores both, the rupee result for the tax file and R for the review, but the column the review reads to judge the method is R.
The process grade: grade the decision, not the outcome
This is the field that changes a trader, and it is the one thin journals never have. The process grade is a plain A, B or C assigned after the trade, based only on how well you executed your plan, with the win or loss deliberately excluded from the judgement. An A-grade trade followed the plan exactly. A B-grade had a minor, acknowledged deviation. A C-grade broke the rules in a material way. Whether it made money is irrelevant to the letter.
The reason to separate the grade from the result is a documented cognitive error. Annie Duke calls it resulting: judging the quality of a decision by the quality of its outcome. Markets, like poker, contain randomness, so a good decision can lose and a reckless one can win. If you let the result assign the grade, you will reward your worst habits every time they happen to pay off and punish your best ones every time variance goes against them. The rules-followed flag (field 17) exists alongside the grade for exactly this reason: it is a blunt yes or no that cannot be softened, so a broken rule that happened to profit still reads as a broken rule.
The single most valuable cell in the journal is this one, because it is the only measure of the part you control. You do not control whether one trade wins. You control whether you followed your rules, sized from a risk budget and exited by plan. Over a large enough sample, a high share of well-executed trades is what carries an edge into results. This is why the danger case is not a loss but a specific kind of win: the C-grade win, the rule-break the market happened to reward, which teaches you to break the rule again. An A-grade loss, by contrast, is a good decision the market did not pay this time, and it is the most useful row in your book.
Risk as a fraction: the field that catches size drift
Field 10 asks for risk twice: once in rupees and once as a percent of current equity. The percent is the one that matters, and it is the one most journals omit. A fixed rupee risk is a moving fraction of a changing account. Risk two thousand rupees on a two lakh account and you are risking one percent; keep risking two thousand after the account grows to four lakh and you are now risking half a percent, quietly under-betting your edge. Far more dangerously, keep the rupee risk climbing after a good run, four thousand, then six, on the same account, and the fraction balloons without any decision that felt like a decision.
Logging risk as a percent of equity is what makes that drift visible in a monthly review. When results turn worse, the percent column answers the first question a review should ask: did the setups stop working, or did the size change? The rupee column cannot answer it, because the same rupee figure means something different at every account level. This is the field that keeps a good month from quietly becoming the setup for a bad one.
How to use it: three cadences, one goal
The columns exist so that patterns become findable, not for their own sake. A field you never review is a field you should not keep. The schema is read on three cadences, and each cadence looks at a different thing.
| Cadence | When | What you look for | Columns it reads |
|---|---|---|---|
| Per trade | Within the hour, chart still fresh | An honest process grade, the exit reason, one line on emotional state. Grade the decision before the result cools into a story. | 12, 16, 17, 18, 19 |
| Weekly | A fixed 30 minutes, same day each week | The A-grade share versus last week, the best-executed and worst-executed trades, any losing time-of-day or setup. Set exactly one goal. | 1, 3, 13, 17, 18 |
| Monthly | End of month, over the full sample | Expectancy in R, rules-followed rate, risk-percent drift, and which setups earn their place versus which are quietly break-even. | 3, 10, 13, 14, 15, 17 |
The one-goal constraint at the weekly review is not a nicety. A trader who sets five goals improves at nothing, because attention does not split five ways across behavioural change. A trader who sets one goal per week fixes one thing, and over a year that is a long series of small, compounding corrections to process. The journal's nineteen columns exist so the review has nineteen angles of attack; most weeks, three or four of them will surface the one thing worth working on next.
A rubric keeps the grade from drifting into a mood. The grade is about adherence to the plan, never about the result.
| Grade | Criteria, judged on execution alone |
|---|---|
| A | Followed the written plan exactly. Entry, stop and size as planned. Exit by rule, not impulse. Rules-followed flag is yes. The result, win or loss, does not enter the grade. |
| B | One minor, acknowledged deviation: a slightly late entry, a stop nudged within reason, a partial taken off-plan. The intent held; the execution slipped a little. |
| C | A material rule-break: no stop, size over the ceiling, a revenge entry, a stop widened to avoid the loss. Rules-followed flag is no. A profitable C is still a C. |
The point of all of this is upstream of the order ticket. A journal does not create an edge; it measures how faithfully you executed one, and where you leak. The entry has to be worth taking, the stop has to sit where the idea is genuinely wrong, and the size has to come from a risk budget rather than from whatever leverage is on offer. That upstream judgement, the edge, the invalidation level and the sizing, is the part worth learning, and it is exactly what the method we teach is built around. The nineteen columns are how you hold yourself to it.
One filled row, then the feedback loop
A single row is a record. A hundred rows, read on a cadence, are a diagnostic instrument. The whole design exists to turn one honestly filled row into a pattern you can act on, then into a change in next week's process, then back into new rows. That loop, and not any single column, is where a journal earns its place.
Put the schema to work
Reading the columns is one thing; grading a real row is another. The Trade Journal Grader walks a single trade through the process grade so the A, B or C stops being a mood and starts being a rule. The full curriculum sits behind the practice room, and the practice room is included with enrolment across the 6 stages and 30 volumes.
Take the free diagnostic Open the Trade Journal Grader View CurriculumFrequently asked questions
Why keep 19 columns instead of just profit and loss?
+A profit-and-loss-only journal records that you won or lost but not why, so it cannot tell you which setups, sizes, times or emotional states are costing you. A rich schema lets a leak become findable. Group the columns into the plan, the execution, the outcome and costs, and the review. Once each trade carries those dimensions, a weekly scan can isolate a losing time-of-day or a break-even setup that a thin journal hides completely.
What is an R-multiple and why log it?
+R is your initial risk on a trade, the distance from planned entry to planned stop times the position size. An R-multiple expresses the result in units of that risk, so a full-stop loss is minus one R and a gain of three times the risk is plus three R. Because R is dimensionless, a small position and a large one, or two different instruments, sit on one scale. The average R across a sample, called expectancy, measures edge without being distorted by how much you sized.
What is the difference between the process grade and the outcome?
+The outcome is whether the trade made or lost money. The process grade is whether you executed your plan well, judged independently of the result. Markets contain randomness, so a sound decision can lose and a reckless one can win. Grading the decision rather than the result is the antidote to what Annie Duke calls resulting, or outcome bias. An A-grade loss, a good decision that the market did not reward, is worth more to your development than a C-grade win.
Why is the process grade the single most valuable column?
+Because it is the only column that measures the thing you can actually control. You do not control whether any single trade wins, but you control whether you followed your rules, sized correctly and exited by plan. Over a large sample a high proportion of well-executed trades is what carries an edge into results. If you improve only one metric, raising the share of A-grade trades is the one that compounds, which is why the review reads it first.
What is the rules-followed flag for?
+It is a plain yes or no recorded after the trade: did you follow the plan you wrote before entering. It is deliberately blunt and separate from the A, B or C grade so you cannot blur a broken rule into a soft score. Cross-referenced against R-multiples, the flag answers a decisive question: are your rule-following trades outperforming your rule-breaking ones? For most traders they are, and seeing that in the data is what makes discipline feel earned rather than imposed.
Why log risk as a percent of equity, not just in rupees?
+A fixed rupee risk means a different fraction of your account as the account grows or shrinks, so the rupee figure alone hides whether your sizing is consistent. Logging risk as a percent of current equity keeps sizing comparable across time and flags the dangerous drift where risk per trade creeps up after a good run. It is the field that lets a monthly review answer whether position size, rather than the setups, is what changed.
Should I log planned levels or actual fills?
+Both, in separate columns. The planned entry, stop and target record the decision you made before the market moved. The actual entry and exit fills record what really happened. The gap between them is one of the most diagnostic numbers in the journal: it exposes slippage, hesitation and stops that were moved mid-trade. If you overwrite the plan with the outcome you lose the ability to grade the decision honestly, because you can no longer see what the decision was.
How often should I review the journal?
+On three cadences. Per trade, grade the decision and note the exit reason while the chart is fresh. Weekly, scan for patterns: the A-grade share, the best-executed and worst-executed trades, any losing time-of-day or setup. Monthly, aggregate: expectancy in R, rules-followed rate, risk-percent drift, and which setups earn their place. The columns exist so these reviews have angles of attack, not for their own sake. Set exactly one improvement goal at each weekly review.
Can I use real screenshots and named stocks in my private journal?
+Your own private journal is your record, and a chart screenshot is one of the columns. The regulatory constraint applies to educational content shared publicly, not to your personal notes. As of 1 July 2026, SEBI requires a uniform 30-day lag on market price data used for educational purposes, replacing the earlier split of a one-day sharing rule and a three-month usage rule. If you publish or teach from your journal, delay the price data by at least 30 days and keep an audit trail.
Where the facts come from
- R-multiples and expectancy. The concept of expressing a result in units of initial risk, and of expectancy as the mean R-multiple of a system, is associated with the Van Tharp Institute's trading concepts. vantharpinstitute.com
- Process versus outcome, and resulting. The error of judging a decision by its result, and the practice of a decision journal that records reasoning before the outcome is known, are set out in Annie Duke's work on decision-making under uncertainty (Thinking in Bets).
- SEBI 30-day lag for educational price data. SEBI circular HO/47/17/12(11)2025-MRD-POD3/I/11107/2026 dated 8 May 2026 prescribes a uniform 30-day lag for sharing and using market price data for educational purposes, with an audit-trail requirement, effective 1 July 2026, replacing the earlier one-day sharing and three-month usage rules. business-standard.com
- Why the feedback loop matters. SEBI's study of individual traders in the equity derivatives segment (July 2025) reported that over 91 percent lost money in FY25, with net losses of about ₹1,05,603 crore, underscoring how a broken review loop, not a lack of chart skill, drives retail losses. business-standard.com
Educational note. This guide explains the fields of a trade journal and how to review them. It is not a recommendation to trade or invest, and it is not investment advice. Bharath Shiksha is an educational publisher, not a SEBI-registered investment adviser or research analyst.
No performance guarantees. Keeping a journal does not guarantee any trading outcome. No claim is made or implied about the returns any individual will or will not generate. The instruments, prices and grades used above are illustrative labels, not real securities or results.
Historical context only. Any examples are anonymised and illustrative. Market price data used in educational content is subject to the SEBI uniform 30-day lag effective 1 July 2026. No real-time market data or specific securities are referenced.