How to read a strategy curve: five checks and one review flow
Use a practical checklist covering version, instrument, period, test mode, drawdown and trading costs.
Niubang editorial · Updated
A curve must first say under what conditions it happened
A curve shows the shape of an outcome, not a complete ledger. When you see an upward line, ask whether it comes from a historical backtest, forward test, public live record or connected account. Execution, costs and sample coverage can differ across these records. Comparing unlike records creates precise-looking conclusions without a sound basis.
Balance mainly reflects settled changes; equity also moves with unrealized profit and loss from open positions. An account with positions can have a steady balance while its equity is already in drawdown. Identify the metric behind the curve before discussing its shape.
Five pieces of evidence: each missing one is a question
One is version and environment: the EA or strategy version, terminal and symbol. Two is the sample: start and end dates, timezone, trade count and whether positions remain open. Three is the mode: backtest, forward test, simulation or live. “Historical performance” alone does not tell a reader the conditions behind the result.
Four is data and cost: the historical-data mode, spread, commission, swap, slippage and delay used in a test. Five is the risk shape: starting balance, balance and equity, maximum drawdown, losing streaks or concentrated positions. MetaTrader 5's official strategy-testing documentation says the tester can select a symbol, period, dates and forward interval, and that real-tick mode is closer to actual conditions. Keep those settings with the report instead of keeping only a return screenshot.
A ten-minute review flow
Step one: write one sentence containing the platform, symbol, sample dates, data mode, source and update time. Step two: put backtest, forward and live records in separate columns instead of combining them into one “track record.” Step three: read balance and equity together and mark periods with open positions. For drawdown definitions, continue with Niubang's drawdown guide.
Step four: write “not provided” for missing fields and list the original material needed, such as a test report, trade history or cost settings. Step five: save the review date so a later check uses the same definitions. If you need help organizing evidence, take this checklist to Niubang's technical services page and confirm scope and deliverables first; the public form does not require a password.
After reading, you should be able to say three things
You should be able to say what the curve represents, what it does not represent and which evidence is missing. For example: “This is a backtest balance curve for one symbol over a stated period; cost settings were not provided, so it is not a live result.” That is a more reliable description than “the curve looks good.”
Then open the rankings and compare similar records using the same conditions, checking source, drawdown, sample and freshness. Historical data helps explain past conditions and cannot guarantee future results. The value of this flow is knowing what to check next, not making a trading decision for you.