A Backtesting Journal for Your Backtest Trades
Log the trades from your backtests by hand or CSV on a Backtest account and analyse them like live trades. It is a journal, not a replay or automated testing engine. Updated 18 October 2026
What is a backtesting journal?
A backtesting journal records the trades from a historical test with the strategy version, the period tested and each result in R, then adds up win rate, expectancy, drawdown and the longest losing streak. It is a log and an analysis tool, not a testing engine: you run the backtest elsewhere and record it here.
The free .xlsx template for backtesting is further down.
What This Is, and What It Is Not
The Trading Terminal is a journal and analytics platform. It records and analyses backtest trades that you enter or import. It is not a backtesting engine: there is no chart replay, no historical price data and no automated testing of rules. You run the backtest in your charting platform or by hand, then log the trades here to see the results laid out the way live trades are.
The Backtest Account Type
- In the Accounts manager, set an account's type to Backtest, with the starting balance and currency you used. Profit goal and max drawdown fields are always shown for Backtest accounts.
- Backtest accounts are left out of Risk Desk, which is built for accounts with live loss limits.
- Because each account has its own trades, you can filter the dashboard, journal and analytics to the Backtest account alone, or exclude it, so tested trades do not blend into live results.
Logging Historical Trades
- By hand. The entry date and time are editable, so you can log a trade at the date it would have happened. The session is filled in from the entry time. Setup, confluences, mistakes, R:R, grade, notes and screenshots work as on any trade.
- By CSV. Import from MetaTrader, cTrader, Tradovate or a generic CSV, choose the account to import into, set the CSV timezone if needed and map any columns the importer does not recognise.
- Limits. The Free plan allows up to 30 trades per calendar month and imports count toward it. Pro has unlimited trades. R values are used for expectancy, so log your planned and actual R:R if you want it.
A Backtesting Spreadsheet Alternative
A backtest log in a spreadsheet works until you want win rate per setup, expectancy, profit factor or drawdown, which means writing and maintaining formulas. If your backtest is already in a spreadsheet, export it as CSV and import it with the generic layout. The journal then calculates win rate, expectancy, profit factor, max drawdown and the breakdowns by setup, session, direction, timeframe and grade for you, on Pro and Elite. The trading journal template shows the columns a log needs, and how we calculate sets out each formula.
Comparing Strategies With Trade Lab
Trade Lab (Elite) lets you name strategies and compare them side by side on trades, win rate, expectancy, average R:R, profit factor and net P&L. A strategy is defined by filters on setup, symbol, session, direction, grade, timeframe and minimum R:R, and sample confidence is shown as thin, directional or strong depending on how many trades matched.
Strategy filters have no account option, and Trade Lab reads your closed trades across all accounts. If a backtested setup and the same setup traded live share a name, their trades are counted together. Giving the backtest version its own setup name keeps them separate.
Reading Backtest Results Carefully
A backtest result is only as good as the trades and fills you entered. A small sample moves a lot with each trade, which is why the Edge Score counts sample size, and a backtest cannot show how it feels to trade the plan with money at risk. Treat the numbers as a description of the tested trades, not a forecast.
What should a backtesting journal track?
- Test and version. A name and a version number for the rules. Change a rule and you start a new version, so results are never mixed.
- Period tested and instrument. Which dates and which market the trades came from, so a result is never separated from its context.
- Entry, stop and exit. Enough to work out the result in R for each trade from prices alone.
- Cost assumptions. The spread and commission you assumed. A test that ignores costs flatters the result.
- Notes on conditions and judgement calls. Anywhere you had to decide whether a trade met the rules.
How do you read a backtest in R?
Because a backtest has no real account, results are best expressed in R, the multiple of the risk taken. The free template calculates the result in R from prices, keeps a running total, tracks the peak and measures the drawdown from it, and counts the current losing streak. Here are ten fictional backtest trades:
| Trade | Result | Cumulative R | Peak R | Drawdown | Losing streak |
|---|---|---|---|---|---|
| 1 | +2.0R | +2.0R | 2.0R | 0.0R | 0 |
| 2 | -1.0R | +1.0R | 2.0R | -1.0R | 1 |
| 3 | -1.0R | +0.0R | 2.0R | -2.0R | 2 |
| 4 | +1.5R | +1.5R | 2.0R | -0.5R | 0 |
| 5 | -1.0R | +0.5R | 2.0R | -1.5R | 1 |
| 6 | -1.0R | -0.5R | 2.0R | -2.5R | 2 |
| 7 | -1.0R | -1.5R | 2.0R | -3.5R | 3 |
| 8 | +2.5R | +1.0R | 2.0R | -1.0R | 0 |
| 9 | +0.8R | +1.8R | 2.0R | -0.2R | 0 |
| 10 | -1.0R | +0.8R | 2.0R | -1.2R | 1 |
The ten trades make +0.8R with a win rate of 40%, an average win of 1.7R and an average loss of 1.0R. Expectancy is 40% x 1.7R minus 60% x 1.0R, which is +0.08R per trade, and the profit factor is 1.13. The deepest drawdown from the peak is -3.5R and the longest losing streak is 3 trades. Those last two are the numbers a backtest's average hides: they say how much you would have to sit through to collect the result. Ten trades prove nothing; this shows how to read the output.
Which mistakes ruin a backtesting journal?
- Fitting the rules to the data. Adjusting rules until the past looks good tends to give a result that does not repeat. Dating each version makes the tinkering visible.
- Marking setups with hindsight. Decide whether a trade qualifies by the rules, not by how it ended.
- Ignoring costs and slippage. Record the assumptions and keep them fixed.
- Testing one period. A result from one market phase says little about another. Log the period and compare several.
- Stopping at a flattering sample. Decide the number of trades in advance.
A backtest is a description of the tested trades, not a forecast. It cannot show how it feels to trade the plan with money at risk.
Is there a free backtesting journal?
Yes. The Trading Terminal's Free plan is A$0 with no card: the full journal and dashboard for 1 account with a Backtest type, up to 30 trades per calendar month, 90 days of dashboard history, and CSV import and export. Imports count toward the 30, so a large backtest will not fit; the Analytics page, with drawdown and expectancy across your trades, is Pro (A$19 a month or A$189 a year). It is not a backtesting engine: there is no chart replay or historical price data.
The free backtesting journal template has no trade cap. It takes the test and version, the date in the data, the instrument and prices, calculates the result in R, the running R total, the peak, the drawdown in R and the losing streak, and its Summary sheet shows win rate, average win and loss in R, expectancy, profit factor, total R, maximum drawdown in R and the longest losing streak, with results by test version, setup and instrument.
Download the backtesting journal template (.xlsx)
Free, no sign-up, no macros. Row 6 is a fictional EUR/USD test trade: overwrite it. In Google Sheets use File > Import > Upload. See also the trading journal template and the free trading journal page.