The realbacktesting blog

The Sortino ratio isolates downside deviation, but target choice, return frequency, and loss sequencing can still flatter a prop backtest.

A confidence interval shows how uncertain a backtest expectancy really is, and whether a positive average is doing more work than the evidence.

Autocorrelation shows when ordered returns are related, making trade counts, Sharpe ratios, and drawdown estimates easier to overstate.

The deflated Sharpe ratio tests whether a selected backtest still looks credible after trial count, sample length, skew, and fat tails.

News filters can reduce event risk, but a prop backtest must prove the filter improves survival after missed trades and costs.

Market regime backtesting checks whether a prop strategy survives trends, ranges, volatility shocks, and quiet tape.

Time-to-target shows whether a prop backtest reaches a rule target before drawdown, costs, and variance break the account.

Parameter sensitivity shows whether a backtest survives small setting changes. Prop traders need a plateau, not one perfect setting.

Exposure shows how much time and capital a strategy keeps at risk. In prop backtesting, the quiet hours can fail the account.

Time-of-day filters change volatility, spread, fills, and drawdown clustering. A prop backtest must prove the clock helps.

Take-profit distance changes win rate, payoff, holding time, and rule pressure. A prop backtest must prove the target is reachable.

Stop-loss distance changes sizing, costs, MAE, and rule pressure. A prop backtest must prove the stop survives the account.

Cross-validation can leak future information in trading. Purging and embargoing keep the test set genuinely unseen.

Trade frequency changes costs, clustering, and drawdown pressure. A prop backtest must show whether the pace is survivable.

Losing streaks expose whether a prop backtest survives normal clustering. A profitable edge can still hit the rule floor.

Minimum lot size can make a small prop account trade differently from the backtest. Check it before trusting risk or drawdown.

Average trade duration shows how long risk stays open. For prop traders, that changes costs, path risk, and rule compatibility.

Maximum adverse excursion shows how much pain a trade took before closing. For prop traders, that path can matter more than the exit.

A cTrader backtest is only useful if you can reproduce the data, costs, logic, and drawdown path yourself. Here is the checklist that matters.

The trade list did not change. The denominator did. That is why win rate, profit factor, and Sharpe can move.

If a system holds through news or over the weekend, FTMO's Swing and Standard accounts are different rule sets. The backtest has to match.

Prop firms can fail an account on floating loss, not just closed loss. The difference is balance versus equity.

A backtest built only from the winners that still exist is not conservative. Survivorship bias hides the dead names and overstates the edge.

A backtest can look brilliant simply because it smuggled in future information. Here is how look-ahead bias sneaks in and how to stop it.

Fixed lot sizing looks tidy, but it makes risk drift when a prop account can least afford it. Here is why fixed-risk sizing fits drawdown rules better.

Three trades can look diversified and still be one macro bet. Correlation risk is why prop accounts break faster than the trade count suggests.

The same cTrader strategy can produce different results when broker data, spread, commission, and trading sessions change. Here is why.

A strategy can look flawless on the sample that created it and still be useless live. Here is how prop traders can spot backtest overfitting early.

Profit factor says how much you made. Sharpe ratio says how violently you made it. For prop traders, the smoother path often matters more.

One clean out-of-sample split is useful, but it can still flatter a strategy. Walk-forward testing shows whether the edge survives repeated retests.

A prop account can be green overall and still fail today. Here is how daily loss limits and max loss rules actually work together.

A backtest's worst drawdown is one path, not a ceiling. Monte Carlo drawdown shows the loss band a prop trader actually has to survive.

There is no magic backtest sample size. Here is how to tell whether a trading strategy has enough trades to be worth trusting.

Out-of-sample testing is the only part of a backtest that has not already seen your optimisation. Here is how to read it properly.

Risk of ruin is why profitable traders still fail prop challenges. The edge matters, but position size decides whether the account survives it.

A prop challenge is not passed by win rate alone. Reward-to-risk and drawdown control matter more than being right often.

Every day this site publishes a market note and a piece on trading. They are written by a specialised AI analyst — kept honest by the same rules as our backtests.

Prop firm trailing drawdown is a moving loss floor, not a static buffer. Here is why profitable traders still fail it and how to read it correctly.

A pretty equity curve is the easiest thing in trading to fake. Here is what separates a backtest you can trust from one that is quietly lying to you.