Loading market quotes…

Subscribe free
Market Fieldnotes13 minFreeGBC Research

Trading journal: what to record, plus a free template

In short

A trading journal records every trade twice. Before the order: the reason and its source, known events with dates, entry, the level where you admit you were wrong, share count, dollar risk and holding period. After the fill: prices and quantities from the broker confirmation, costs, net result and the result in multiples of risk (R). The original note is never rewritten; changes of mind become dated lines. Each month, count recurring errors and fix one.

In 1972, before President Richard Nixon’s trips to Beijing and Moscow, psychologists Baruch Fischhoff and Ruth Beyth asked a group of participants to estimate the probability of a range of possible outcomes of the two visits. The estimates went down on paper.

Some time after Nixon returned, without warning, they asked the same people to recall the probabilities they had given. For events they believed had happened, the remembered numbers were generally higher than the ones actually written down; for events that hadn’t happened, lower. Few of them remembered being very surprised. The study came out in 1975 under the title “I knew it would happen.”

Trades run on the same mechanism. After a win you remember calling the rally; after a loss you remember having doubts from the start. A trading journal keeps the original sheet of paper, written before the market tells you how it ended. In a moment we will run the experiment on you, with a trade you have already closed.

What is a trading journal, and what is it for?

A trading journal is a log where you record every trade at two moments: before the order, with the plan and the reasons, and after the fill, with what actually happened. Its job is to let you reconstruct a decision without relying on memory.

A screenshot of a green profit can’t do that. It shows one number and stays silent on everything else: why you got in, what you expected, where you would have admitted you were wrong, whether the size was right. A useful journal answers those questions six months later, when the chart has changed and so has your opinion.

An Excel file or a Google Sheet is enough, as long as every row has the same fields and the original note stays untouched.

What to record before you enter a trade

Memory distorts the pre-order part most, so fill it in first and then leave it alone. It has four blocks.

The reason, with its source

Write down why this trade and why now: an earnings report, a contract, a management change, a sector rotation. Add the link to the source. “Looked strong” is impossible to check a month later; “raised revenue guidance, link to the release, volume three times the average” takes two minutes to verify.

Known events until the planned exit

Earnings, dividends, shareholder meetings, macro data that fall inside the planned holding period, each with a date and a source. If the field is empty because you didn’t check, the note isn’t ready. One missed event turns a five-day swing trade into an unintended bet on the earnings gap.

Entry, stop level and share count

The price where you plan to enter, the price where you admit you were wrong (your stop level), and the number of shares. Those three numbers give you the dollar risk: (entry − stop level) × shares. That is the yardstick you will judge the result against, and it depends on how you chose your position size.

Planned holding period and the chart as it looked

How many days you expect to hold, plus a screenshot of the chart as it looked before the order. The chart you reopen a month from now already contains the ending, and it will make Monday’s uncertainty look obvious.

What to record after the fill

After the order come the facts: quantity filled, actual prices, times, commissions, any currency conversion, and the result. Copy them from the broker’s confirmation. In the United States, SEC Rule 10b-10 requires the broker to send that confirmation at or before completion of the transaction, showing the date, price, number of shares and the broker’s compensation.

The order you meant to send and the order that filled are two different events. An immediate-or-cancel order, as FINRA explains, fills whatever it can right away and cancels the rest: you wanted 50 shares and may get 30. A limit order may never fill, so it pays to know your order types before you use them. The journal gets the quantity on the confirmation.

Costs go in their own column, in dollars: actual commissions and, if you buy dollar stocks from a euro account, currency conversion, which can bite on both the buy and the sell. Our guide to trading costs lists them one by one.

How do you calculate the result of a trade with partial exits?

Take a hypothetical trade. On Monday you buy 50 shares at $40, $2,000 in total. The stop level is $38 and the planned holding period five days, so the planned risk is ($40 − $38) × 50 = $100. On Wednesday you sell 20 shares at $42, on Thursday the other 30 at $39. Assume a $1 commission per order.

Hypothetical trade: one buy, two sells, three commissions
TransactionSharesPriceAmountResult
Buy50$40.00$2,000
Exit 120$42.00$840+$40
Exit 230$39.00$1,170−$30
Commissions3 orders$1.00$3−$3
Total50$40.20 average exit$2,010 received+$7

A screenshot of the first exit shows +$40, a 5% gain on the shares sold. The whole trade made $10 gross and $7 net, 0.35% of the $2,000 invested.

More useful still is the ratio of the result to the planned risk, called the R multiple. Here it is $7 ÷ $100 = 0.07R, seven cents earned for every dollar you were willing to lose. R multiples let you compare trades with different prices and sizes, and their average over many trades is your expectancy, the subject of our guide to win rate and risk-reward.

The screenshot says +$40. The account says +$7.The screenshot says +$40. The account says +$7.. Hypothetical trade: 50 shares bought at $40 and sold in two piecesFIG. 1 · TWO EXITS, ONE RESULTGB.CAPITALThe screenshot says +$40. The account says +$7.Hypothetical trade: 50 shares bought at $40 and sold in two piecesExit 1: 20 shares at $42+$40Exit 2: 30 shares at $39−$30Gross result+$10Commissions, 3 orders−$3Net result+$7Hypothetical example. Assumes a $1 commission per order (three orders) and no currency cost.
Figure 1 The first exit alone shows +$40; the whole trade made $7 net.

The math checks out. Judging whether the trade was done well takes the timeline.

How do you update a journal without rewriting the original note?

On Wednesday morning, reading the news, you find out the company reports earnings on Thursday after the close. In Monday’s note the events field was empty. You sell 20 shares at $42 to cut your exposure, and on Thursday afternoon, with the stock sliding, you close the other 30 at $39 before the numbers.

The temptation is to fix Monday’s note by adding the earnings date, as if you had known. Write a new line instead, with date and time, saying what you learned, when, from which source, and what you did about it. The original note stays as it was, empty field included.

Every change of mind gets a new line and a timeEvery change of mind gets a new line and a time. Hypothetical timeline of the same trade, New York timeFIG. 2 · ORIGINAL NOTE AND UPDATESGB.CAPITALEvery change of mind gets a new line and a timeHypothetical timeline of the same trade, New York timeMON 3:30 PMOriginal noteReason and source, entry at $40, stop at $38, 50 shares, 5 days. “Known events”field: empty.MON 3:41 PMFill50 shares at $40.00: time and price copied from the broker confirmation.WED 9:50 AMDated updateEarnings come out Thursday after the close. Link to the company’s release.WED 10:05 AMActionSold 20 shares at $42 to cut exposure to the report.THU 3:20 PMActionSold the other 30 at $39, before the numbers came out.SATURDAYReviewError: earnings date missing from the note. Fix: no note is ready without thedate and source of known events.Hypothetical example for illustration.
Figure 2 Monday’s note stays as it was: every new piece of information goes in as a dated line.

The timeline separates what you knew from what you found out later, the very distinction Fischhoff and Beyth’s participants lost. At the Saturday review the error is plain, because the events check never happened. The two exits, taken on their own, were consistent with the new information.

What is the difference between a good trade and a winning trade?

A winning trade can contain the wrong share count or an earnings report held without knowing it. A losing trade can follow the plan to the letter, in which case the loss was the planned cost of an idea that turned out wrong. That is why every closed trade gets judged on two separate axes: the outcome and whether you followed the plan.

A winning trade can be a lucky mistakeA winning trade can be a lucky mistake. Judge every closed trade on two axes: the outcome and whether you followed the planFIG. 3 · PROCESS AND OUTCOMEGB.CAPITALA winning trade can be a lucky mistakeJudge every closed trade on two axes: the outcome and whether you followed the planAcceptable lossThe planned cost of an idea that was wrong.Nothing to fix.Earned winRepeat it at the same size, even wheneuphoria asks for more.Avoidable lossThe price of the mistake is already on theaccount. Find the broken rule.LuckThe market rewarded a mistake, and it staysin your process until you log it.LossGainOUTCOMEPlan: broken → followedExampleTeaching diagram. The dot is the hypothetical trade from the example: +$7 net, with the earnings datemissing from the note.
Figure 3 Judge outcome and process separately: the bottom-right quadrant is the one that fools you.

Our +$7 trade sits in the luck quadrant. Had you skipped the news on Wednesday, you would have found out about earnings on Friday morning, with the stock already gapping at the open. The gain only tells you that this time the mistake cost nothing.

Try it now. Take the last trade you closed and, without opening the chart, write down three answers: why you got in, at what price you would have admitted you were wrong, how many days you expected to hold. Then compare them with what you wrote at the time, if you wrote anything. If any of the three answers came to you while thinking about how the trade ended, you have just rerun the 1972 experiment on yourself.

How do you review a trading journal every month?

The monthly review is there to find the error that keeps coming back and to fix one at a time. Three habits make it work.

Observable labels instead of judgments

A judgment like “poor discipline” is impossible to count. “Moved my exit after the price fell” takes a second to count. Use a closed list of labels, always the same ones, and give each trade with an error a single label: at month end one formula tells you which comes up most often.

One correction at a time

If the most frequent error is the unrecorded event, the fix is a rule you can check: no note is ready without the date and source of events in the holding period. At the next review you count the notes missing either field. That measures whether you applied the fix, regardless of what the stocks did. Changing six rules in the same month makes it impossible to tell which one made the difference.

Rejected trades count too

Log the candidates you looked at and passed on, with the reason. They show whether your selection holds steady, or loosens after a loss and turns aggressive after a win.

If you want a second pair of eyes, bring the original note, the updates and the broker confirmations to a 1:1 review of your trades: with those three documents every decision can be reconstructed in detail.

The journal records individual decisions; the account as a whole needs its own tracking on the equity curve, because three small, well-executed losses can add up to a drawdown that no single row hints at.

When results are calculated from actual fills, the spread is already included and must not be subtracted again in the cost column. You can record it separately to study execution quality. For a short, K is the average opening sale price per share and M is the total cost of buying the shares back. Also record applicable interest and stock-borrow charges.

Frequently asked questions

What should a trading journal include?

Before the order: the reason and its source, known events with dates, entry price, the level where you admit you were wrong, share count, dollar risk and planned holding period. After the fill: prices and quantities from the broker confirmation, costs, net result, result in R, whether you followed the plan, and one error label from a fixed list.

Is a spreadsheet or an app better for a trading journal?

Both work. An Excel file or Google Sheet is free, stays yours, and with a few formulas computes risk, R multiples and monthly statistics. An app imports fills automatically, but the note written before the order has to be typed by hand either way.

How often should you review a trading journal?

For swing traders, a full review once a month is a good rhythm: enough trades to spot a repeated error, soon enough to fix it. At the end of each week a few minutes are enough to check that every row has its original note, broker confirmation and label.

How do you record a partial exit?

Each exit is an update line with date, time, quantity, price and reason. The result is calculated on the whole trade: total proceeds minus purchase cost and commissions. Buying 50 shares at $40 and selling 20 at $42 and 30 at $39 gives +$10 gross, even though the first exit alone shows +$40.

What is an R multiple in a trading journal?

It is a trade’s result divided by its planned risk. If you were willing to lose $100 and made $7, the result is 0.07R; if you lost $100, it is −1R. It lets you compare trades with different sizes and prices.

Should you record trades you didn’t take?

Yes, at least the ticker, date and reason for passing. Without those rows the journal only shows trades that reached the account and can’t tell you whether your selection loosened after a loss or turned aggressive after a win.

Sources and further reading

Related research
Trading costs: commissions, spread and currency, explainedMarket Fieldnotes · 05.10.2026Earnings gap risk: why stop losses fail overnightMarket Fieldnotes · 05.10.2026Swing trading stocks: a practical processMarket Fieldnotes · 23.10.2026Stock watchlist: how to build and review your listMarket Fieldnotes · 05.10.2026Swing trading with a full-time job: a practical review routineMarket Fieldnotes · 28.10.2026 Full archive →
FREE · EVERY SUNDAY

Get the next Radar
before Monday’s open.

The first name, the reasoning, and our public research.

FREE · EVERY SUNDAYGet the watchlist