Open any trader's spreadsheet and you can usually tell within ten seconds whether it will ever change their behaviour. If the columns stop at symbol, direction, lots and profit, it is a receipt. It records what the broker already recorded, in a font the trader chose. Nothing in it can answer the question that matters after a bad month, which is not "how much did I lose" but "which of my own rules did I break, and on which kind of setup".
The columns that answer that question are the ones no software can fill for you: the setup name, the reason you clicked, whether the trade was in the plan you wrote before the session, and what you were doing in the twenty minutes before entry. Everything else, the prices and the fees and the timestamps, is import work. That split is the whole argument between a paid journal app and a sheet you build yourself.
The fields a journal has to hold
Start from the broker export. An MT4 or MT5 statement gives you ticket number, symbol, order type, volume, open time, open price, stop loss, take profit, close time, close price, commission, swap and profit. That is enough to compute gross and net results per instrument, per weekday and per hour. It is not enough to compute anything about your process.
The second layer is arithmetic you derive rather than import. Risk in account currency at the moment of entry, taken from the distance between entry and the initial stop. The R-multiple, which is the result divided by that initial risk, and which turns a mixed book of 0.2 lot and 1.5 lot trades into one comparable scale. Maximum adverse excursion, the worst unrealised drawdown the position went through before it resolved, which tells you whether your stops are too tight or your entries too early. Maximum favourable excursion, which tells you whether you are leaving the back half of your moves on the table. If you have never measured those two, they are the highest value additions you can make this month.
The third layer is the human one, and it has to be typed. Keep it to five or six controlled values rather than free text, because free text never gets analysed. A setup tag, a plan-compliance flag, a session tag, and a one-line note is enough structure to slice a hundred trades later. Our longer piece on how to run a trading journal goes through the discipline side of that habit.
Importing is where these tools actually differ
Every journal product on the market claims automatic import. The claim is true and the edge cases are where the hours go. Partial fills arrive as several tickets against one decision, so a naive importer reports three trades where you took one. Hedged accounts holding a long and a short on the same symbol produce closes that net oddly. Broker server time is rarely your time zone, so a trade you took at the London open lands in the Asian session bucket unless you set the offset. Commission and swap are sometimes booked after the position closes, so a same-day import shows a net figure that quietly changes overnight.
Reconcile one full month by hand against the broker statement before you trust any importer. If the app's monthly net and the statement's monthly net differ by even a few units of account currency, find out why before you build a year of analysis on top of it.
What the paid apps add
Edgewonk and Tradezella are the two names that come up most often in retail forums, and they are built on the same premise: import the statement, let the trader tag, then show the tagged data back as analytics. Both accept trade history from common retail platforms, both let you attach chart screenshots to a trade, both produce breakdowns by tag, instrument and time. Their pricing and feature sets change, so check the current terms directly before you buy rather than trusting a comparison table written last year, including this one.
What you are buying is not the analysis. Anyone can compute a win rate by tag. You are buying the fact that the analysis exists on a Sunday evening when you would not have built it yourself, and that the screenshot sits next to the numbers instead of in a folder called charts final v3. That is a real product, and for traders who have already proved they will do the tagging, it is money well spent.
| Approach | Import effort | Best for | Main weakness |
|---|---|---|---|
| Broker report only | None | Tax and account reconciliation | No setup or behaviour data at all |
| DIY spreadsheet | Manual paste, one formula sheet | Under roughly 20 trades a month, one strategy | No screenshots, breaks when you change columns |
| Dedicated journal app | Automatic, needs reconciling once | Several setups across several instruments | Recurring fee, your data lives elsewhere |
The spreadsheet that beats most apps
Three tabs. The first holds the raw statement paste, untouched, so you can always rebuild. The second is one row per trade with formulas pulling from the raw tab and four hand-typed columns for tag, plan flag, session and note. The third is a pivot of the second, grouped by tag, showing count, average R, and the sum of R. Sum of R by tag is the single most useful number in retail trading, because it tells you which of your setups is paying for the others.
The weakness is real. A sheet will not hold screenshots well, it will not remind you to fill it in, and it dies the first time you change a column name mid-year. If you know you will not open it, buy the app and treat the fee as a commitment device. That is a legitimate reason to pay for software.
What changes if you are on a funded account
Prop traders have one extra requirement: the journal has to track equity against the firm's limits in the same terms the firm uses. A daily loss rule measured from the previous day's balance behaves differently from one measured from peak equity, and a journal that only reports closed profit will not warn you that your floating loss is about to breach. Read our explainer on how drawdown is measured and check which definition your firm applies before you build the tracker.
The same trader benefits from logging planned risk in R alongside the firm's percentage limits, because the two collide in ways that are obvious in hindsight. A run of four losses at 1R each is unremarkable in a strategy with a 1:2 payoff profile, and fatal if your risk per trade was set without reference to the daily cap. Trading is high risk and no journal changes that, but it does make the collision visible before it happens rather than after.
"The journal that works is the boring one you actually fill in on Friday. I would rather see a trader keep six honest columns for a year than buy a beautiful app they abandon in March."
— Alex Onta, Executive Director, SINGUARD
Key Takeaways
- The broker statement covers prices and fees, so the value of a journal sits entirely in the tags and plan-compliance fields you type yourself.
- Record results in R-multiples, not currency, so trades of different sizes become comparable across a year.
- Partial fills, hedged positions and server time offsets are the three things that silently corrupt an automatic import.
- Under roughly twenty trades a month a three-tab spreadsheet does the job; pay for an app when tag analytics across several setups is what you need.
Frequently Asked Questions
Do I need a paid trading journal app?
Not at low trade counts. If you place a handful of trades a week and run one strategy, a spreadsheet with a computed R-multiple column answers the same questions. Paid apps earn their fee when you run several setups across several instruments and want the breakdown by tag without building it yourself.
Why do my imported trades not match my broker statement?
The usual causes are partial fills counted as separate trades, hedged positions on the same symbol netted incorrectly, server time zone offsets shifting trades into the wrong session, and commission or swap booked after the close so the imported net profit is stale. Reconcile one month by hand before you trust any importer.
What should I record that the broker does not?
The setup name, the reason for entry, whether the trade was inside your written plan, the planned stop distance in R, and a note on your state at the moment of the click. Those fields are the only part of the journal that can explain a losing month, because everything else is already in the statement.