Backtesting Your Setup: A Practical Process for Prop Firm Traders
Most traders talk about backtesting. Fewer do it systematically. Even fewer do it in a way that produces data they can actually trust and trade from.
Backtesting matters for funded traders specifically because the evaluation environment demands confidence in your edge. When you're under the pressure of a profit target and a drawdown limit, the last thing you need is uncertainty about whether your setups work. That uncertainty leads to second-guessing, missed entries, and the hesitation that causes more blown evaluations than any single bad trade.
A solid backtest gives you a statistical foundation to trade from. Here's how to build one.
What Backtesting Actually Is (and Isn't)
Backtesting is the process of applying your trading rules to historical data and recording the results. Done correctly, it tells you: your win rate, average winner size, average loser size, expectancy per trade, and maximum drawdown over a sample period.
It isn't a guarantee of future performance. Markets change. Setups that worked in a trending 2021 environment may perform differently in a choppy 2024 environment. The purpose of backtesting isn't to find certainty — it's to find a reasonable statistical basis for your approach and identify the conditions where it performs best.
It also isn't optimization for optimization's sake. Traders who backtest by adjusting parameters until they find maximum historical profit are engaging in curve-fitting — a process that produces impressive backtests and terrible live results. The goal is to test your rules as they actually are, not to discover what rules would have been best historically.
The Minimum Viable Backtest
A useful backtest requires a minimum sample size to be statistically meaningful. For most setups, 100 trades is a reasonable minimum — enough to see consistent patterns emerge while not so many that the process takes months.
For manual backtesting (reviewing charts historically by hand), expect to spend 4-8 hours per 100 trades, depending on the timeframe you're trading and how complex your setup is. This is a worthwhile investment before risking evaluation capital.
For automated backtesting (using platforms like TradingView Pine Script, MetaTrader Strategy Tester, or Python-based frameworks), the sample size can be much larger, but requires coding your rules precisely — which forces valuable clarity about exactly what your setup requires.
The Manual Backtesting Process
Step 1: Write out your rules in explicit detail before you start. The entry trigger, stop placement, target, and any setup conditions must be defined precisely enough that someone else could apply them without asking you any questions. Vague rules produce vague backtests.
Step 2: Choose your historical period. Aim for at least 6-12 months of data. Include different market conditions if possible — trending periods, ranging periods, high-volatility periods. If your setup only works in one type of market, you need to know that before you trade it in an evaluation.
Step 3: Scroll through the charts at your trading timeframe and mark every point where your entry conditions were met. Do not cherry-pick. Every valid setup must be recorded — including the losing ones you'd rather not count.
Step 4: Record the outcome of each trade: entry price, stop price, target price, exit price, R-multiple result (how many times your risk you gained or lost), and any notes on why the setup did or didn't work.
Step 5: Calculate your statistics. Win rate, average winner in R, average loser in R, expectancy (average R per trade), maximum consecutive losses, and maximum drawdown.
Reading Your Backtest Results
Expectancy is the most important number. A positive expectancy means your edge is real — over a large enough sample, you'll make money. A negative expectancy means you'll lose money regardless of how good your risk management is.
Expectancy formula: (Win rate × Average winner) - (Loss rate × Average loser)
Example: 45% win rate, average winner 2R, average loser 1R.
(0.45 × 2) - (0.55 × 1) = 0.90 - 0.55 = 0.35R expectancy per trade.
At 0.35R per trade and a $500 risk per trade, you expect to make $175 per trade on average over a large sample. This is a viable edge.
Maximum consecutive losses tells you how long a drawdown period you need to be psychologically and financially prepared for. If your backtest showed 8 consecutive losses in one stretch, you need to be able to survive 8 consecutive losses in live trading without abandoning your approach.
Applying Backtest Results to Evaluation Planning
Once you have a statistical baseline, you can model your evaluation outcomes realistically.
If your expectancy is 0.3R per trade, your average risk per trade is 0.5% of account, and you take 2-3 trades per day, you can project roughly: 2.5 trades/day × 0.3R × 0.5% = 0.375% expected profit per day. To hit an 8% profit target, you'd need roughly 21 trading days — well within typical evaluation windows.
You can also model worst-case scenarios: if you start with 8 consecutive losses (as your backtest showed was possible), what is the drawdown? Does it violate your daily loss limits? If yes, adjust position sizing before the evaluation starts.
Backtesting isn't just about proving your edge. It's about knowing your edge well enough to deploy it confidently under pressure. That confidence is what separates traders who pass evaluations from traders who trade their edge perfectly in demo and choke when real stakes arrive.
Jordan Blake
CashFrame