Keeping a trade journal: what to record, how to review it by R and by setup, and how to avoid hindsight bias
Markitel guides (AI-assisted)
Drafted with AI from the source reports listed below. Informational only — not investment advice.
A broker statement tells you what happened. A trade journal tells you why it happened, and whether you did what you planned. This guide from the Markitel guides desk covers which fields to log, how to measure results in units of risk ("R"), how to compare setups, and how to stop your memory from rewriting your trades after the fact.
A journal won't make anyone profitable. It's a way of measuring, and that's worth having because the odds are hard for retail traders, especially with leveraged products. When the European Securities and Markets Authority (ESMA) restricted CFDs, it cited national regulators' analyses showing that 74-89% of retail accounts typically lose money on their investments, with average losses per client ranging from €1,600 to €29,000. Those measures were agreed by ESMA's Board of Supervisors on 23 March 2018. If you trade, you should be able to tell, from evidence, whether you're in that majority and where the losses come from.
Why a journal, not just a broker statement
Your broker records fills, fees and profit and loss. It doesn't record your plan, your stop before you moved it, the setup you thought you were trading, or how you felt when you clicked. Those are the things you can actually change.
A journal also gives you a record made at the time, before you knew the outcome. As the section on hindsight bias explains, that record is the main protection against a well-documented memory distortion.
The fields worth logging
Keep it short enough that you'll fill it in on every trade. A field you skip half the time is worse than no field at all, because it skews your review. Split the fields into three groups by when you fill them in.
Before entry (filled in before you place the order)
- Date and time of the decision.
- Instrument and direction (long or short).
- Setup tag: one short, consistent label from a fixed list you define, such as "breakout", "pullback to moving average" or "range fade". Don't invent new tags on the fly.
- Thesis: one or two sentences on why this trade, and what would prove you wrong.
- Planned entry, stop and target prices.
- Position size and planned risk in money. This is your 1R for the trade (explained below).
- Risk as a percentage of the account.
- Signal source, if any. If you use a screener or signal service, note it. If it's a Markitel signal, record it as the output of Markitel's rule-based ten-factor technical engine. That lets you check later how you used the signal, kept separate from your own discretionary calls.
- Context: for example, scheduled news, the session, or an unusually wide spread.
- Screenshot of the chart at entry.
During the trade
- Actual fill price and any slippage from the plan.
- Every stop or size change, with the time and the reason.
After exit
- Exit price, time and reason (stop hit, target hit, manual exit, time-based exit).
- All costs: commissions, spread, financing or overnight fees.
- Net profit or loss in money.
- Result in R (net P&L divided by planned risk).
- Rule adherence: a simple yes or no. Did you follow your own plan?
- Notes: emotions, mistakes, anything you'd do differently.
Measuring results in R
An R-multiple is a trade's profit or loss divided by its initial risk, R: the amount the trade would lose if price reached the initial stop. The idea comes from trading coach Van Tharp, whose definition is that "All profits can be expressed as a multiple of the initial risk (R)." A trade that hits its stop scores −1R. A trade that makes twice what it risked scores +2R.
Why bother? Because R-multiple makes trades comparable across assets, position sizes, and account sizes. In one phrasing, dollar results rank trades by size. R-multiples rank them by decision quality. That matters a lot for small accounts. A $3 loss and a $30 loss can be the same mistake at different sizes.
The key rule: R is fixed when you open the trade: stop distance multiplied by position size. Don't recalculate it after moving your stop.
A worked example (hypothetical, round numbers)
Say you have a hypothetical $100 account and you risk $2 per trade.
- You buy 2 shares at $20.00 with a stop at $19.00. Risk per share is $1, so 1R = $2.
- Trade A: you exit at $22.00. Gross profit is $4. Before costs, that's +2R.
- Trade B: price gaps through your stop and you get out at $18.50. Loss is $3, which is −1.5R. The stop was no guarantee.
- Costs: if Trade A cost $0.50 in fees and spread, net profit is $3.50, or +1.75R.
On small accounts, costs can be a big share of 1R, so always log R after costs. If $0.50 of costs comes off every trade, that's a quarter of your risk unit gone before price moves at all.
Leverage and R
With margin, CFDs or futures, work out R from stop distance times full position size, not from the margin you put up. Leverage works on the full position, so small price moves turn into large account moves. Schwab notes that "Depending on the product, forex and futures leverage can be at 20-to-1 or even 50-to-1 compared to equities' overnight margin of 4-to-1." Take a hypothetical case at 20-to-1: a 1% move against you equals 20% of the margin posted.
Losses can also go beyond your deposit. The SEC warns that when you day trade with funds borrowed from a firm or someone else, you can lose more than the funds you originally placed at risk. The journal helps here. If a loss shows up worse than −1R, the planned risk wasn't the real risk.
As one guide puts it, losers should cluster at -1R. Anything worse points to a stop that failed or was moved. Count your trades that come in clearly worse than −1R. That count is often the most useful number in the whole journal.
Average R (expectancy)
Tharp's framework holds that a trading system can be characterized as a distribution of the R-multiples it generates. Expectancy is simply the mean, or average, R-multiple generated. Add up your R results and divide by the number of trades.
Treat that figure as a description of past trades, not a forecast. Small samples are noisy, and markets change. As one prop-trading firm puts it, tracking results this way is not, on its own, a strategy that makes anyone profitable; it's a measurement tool.
Reviewing by setup
One average across all your trades hides a lot. The setup tag is what lets you break it down.
Once a month, or every 20–30 trades, group your journal by tag and work out four things for each group:
1. Number of trades 2. Average R after costs 3. Share of trades where you followed your rules 4. Number of losses worse than −1R
Here's a purely hypothetical summary to show the layout:
| Setup (hypothetical) | Trades | Avg R | Rules followed | Losses worse than −1R |
|---|---|---|---|---|
| Breakout | 20 | +0.3R | 18 of 20 | 0 |
| Range fade | 15 | −0.4R | 9 of 15 | 4 |
| Unplanned ("saw it moving") | 10 | −0.6R | n/a | 3 |
How to read a table like this:
- Don't trust small samples. Ten or twenty trades can be swung by luck. A setup that looks good over 15 trades may look very different over 100.
- Separate the setup from your execution. In the hypothetical "range fade" row, rules were broken on 6 of 15 trades. Before deciding the setup doesn't work, check the trades where you did follow the rules.
- Name your unplanned trades. An "unplanned" tag is uncomfortable but useful. It turns impulse trades into a figure you can track.
- Slice by other fields too, such as instrument, time of day, signal source or news days. If you tag signal-based trades, you can compare how you acted on a Markitel engine output against your own calls. That's a review of your own process, not evidence that any signal is profitable.
Avoiding hindsight bias
Hindsight bias is the common tendency for people to perceive past events as having been more predictable than they were. Psychologist Baruch Fischhoff documented it in 1975, building on Kahneman and Tversky's research on judgment under uncertainty. In his experiments, people told which outcome had occurred rated it as more likely than those who were not told — and they were largely unaware that the outcome information had shaped their judgment. Even when asked to ignore what they knew, participants could not disregard the outcome knowledge even when instructed to.
For traders, this has a direct cost. A behavioural-finance text notes that one detriment of hindsight bias is that it can prevent learning from mistakes. After a winning trade, you remember being sure. After a loss, the warning signs look obvious. Either way, you learn the wrong lesson.
Knowing about the bias isn't enough. One summary of the research notes that warning people about hindsight bias did not reduce it. You need habits built into how you keep the journal:
- Write the thesis before you enter. A thesis written after the trade is reconstructed memory, not a record.
- Timestamp everything, and never overwrite. If you change your view, add a dated note below the original entry. Most spreadsheets and journaling apps can log edit history.
- Screenshot at entry. Review the chart as it looked then, not the finished chart that shows what came next.
- Grade the decision separately from the outcome. Before you look at P&L, score each trade on process: was the setup valid, was size correct, was the stop placed as planned? A rule-following loss can be a good trade. A rule-breaking win is still a mistake.
- Review with outcomes hidden. Hide the P&L and R columns for the first pass, judge the decisions, then reveal the results.
- Note your confidence before entry, on a simple 1–5 scale. Later you can check whether your high-confidence trades actually did better, using data instead of memory.
A simple review routine
- After each trade: fill in the exit fields within the same session, while the details are fresh.
- Weekly: check rule adherence and any trade worse than −1R. Ask what happened in each case: a moved stop, a gap, wrong sizing?
- Monthly or every 20–30 trades: run the setup table. Pick one process change to test next period, not five.
- Keep the original records. Even outside of performance review, keeping clear trade records is good practice for tax and any disputes with your broker.
Key takeaways
- Log your plan, including entry, stop, size, setup tag and thesis, before you place the order. Pre-trade fields are what make a journal more than a broker statement.
- Measure every trade in R after costs, with R fixed at entry as stop distance times position size. Losses much worse than −1R point to moved stops, gaps or leverage working against you.
- Review by setup tag, and keep execution separate from the setup itself. Treat small samples with caution: average R describes the past and doesn't predict the future.
- Leverage applies to your full position, not your margin, and losses can exceed your deposit. Size and log risk on that basis.
- Hindsight bias is strong and warnings alone don't fix it. Timestamped, unedited entries, screenshots at entry and outcome-blind reviews are the practical defences.
Sources
- esma.europa.euPublication date unknown
ESMA agrees to prohibit binary options and restrict CFDs to protect retail investors
NCAs’ analyses on CFD trading across different EU jurisdictions shows that 74-89% of retail accounts typically lose money on their investments, with a...
- pineconnector.comPublication date unknown
R-Multiple Trading: 1R, Formula and Expectancy in R
An R-multiple in trading is a trade's profit or loss divided by its initial risk, R: the amount the trade would lose if price reached the initial stop...
- kenmacro.comPublication date unknown
R-multiple explained: measuring trade outcomes in risk units
R-multiple makes trades comparable across assets, position sizes, and account sizes.
- arrowalgo.comPublication date unknown
R Multiple Trading: Judge Trades by Risk - Arrow Algo
- Dollar results rank trades by size. R-multiples rank them by decision quality.
- schwab.comPublication date unknown
Basics of Margin Trading for Investors
"Depending on the product, forex and futures leverage can be at 20-to-1 or even 50-to-1 compared to equities' overnight margin of 4-to-1."
- sec.govPublication date unknown
Special Study: Report on Day-Trading Broker-Dealers
** When you day trade with funds borrowed from a firm or someone else, you can lose more than the funds you originally placed at risk.
- vantharpinstitute.comPublication date unknown
Tharp Think Trading Concepts
A trading system can be characterized as a distribution of the R-multiples it generates. Expectancy is simply the mean, or average, R-multiple generat...
- the5ers.comPublication date unknown
R-Multiple Trading Explained: Risk & Reward in R
Tracking results this way is not, on its own, a strategy that makes anyone profitable; it’s a measurement tool.
- en.wikipedia.orgPublication date unknown
Hindsight bias
Hindsight bias, also known as the knew-it-all-along phenomenon [[1]](./Hindsight_bias#cite_note-APS-1) or creeping determinism, [[2]](./Hindsight_bias...
- behavioraleconomicslab.comPublication date unknown
Hindsight Bias: Definition, Examples & Psychology
Psychologist Baruch Fischhoff documented it in 1975, building on Kahneman and Tversky's research on judgment under uncertainty.
- edu.koji.soPublication date unknown
Hindsight Bias in Course Evaluation: The I Knew That Course Was Weak All Along Problem
Fischhoff showed creeping determinism: participants told that a particular outcome had occurred judged it far more probable in hindsight than those no...
- onlinelibrary.wiley.comPublication date unknown
Hindsight Bias - Behavioral Finance and Wealth Management - Wiley Online Library
One detriment of hindsight bias is that it can prevent learning from mistakes.
- fasterthannormal.coPublication date unknown
Hindsight Bias Mental Model: Definition, Examples & When to Use
Warning people about hindsight bias did not reduce it.
Researched and drafted with AI from the sources above, then checked by a second AI review against Markitel's editorial standards. General information only — not investment advice.
Drafted by Markitel AI


