Reference

Glossary

Plain-English definitions of the terms you’ll meet in backtest reports and in my articles.

Returns

Net profit: Total profit minus total loss over the test, after any costs that were modelled. If costs weren't modelled, the real figure will be lower.
Gross profit and gross loss: The sum of all winning trades, and the sum of all losing trades, kept separate.
Profit factor: Gross profit divided by gross loss. Above 1 means the strategy made money over the test; 2 means winners earned twice what losers cost.
Average trade: Net profit divided by the number of trades. The amount a typical trade made or lost.
Expectancy: What you can expect to win or lose per trade on average, combining win rate with the size of wins and losses. Positive expectancy is the minimum requirement for an edge.
Payoff ratio: Average winning trade divided by average losing trade. Also called the payout ratio or win/loss ratio.
CAGR: Compound annual growth rate. The steady yearly growth that would turn the starting capital into the final equity over the test period.
Buy and hold: What simply buying at the start of the test and holding to the end would have returned. A useful benchmark that a strategy should beat after costs, or justify with lower risk.

Maximum drawdown: The largest fall in equity from a peak to a later low, in money or as a percentage. The worst loss you would have had to sit through.
Drawdown duration: The longest time the strategy spent below a previous equity high. Also called stagnation. Long flat periods test patience as much as losses do.
Return to drawdown: Net profit divided by maximum drawdown. How much the strategy earned for each unit of its worst loss.
Sharpe ratio: Average return divided by the volatility of returns, usually annualised. Higher is better. Different platforms calculate it differently, so compare Sharpe ratios only from the same tool.
Sortino ratio: Like the Sharpe ratio, but only downside volatility counts as risk, so large winning trades aren't penalised.
Risk of ruin: The probability that the account falls to a chosen loss level (for example, half the starting capital). Usually estimated with Monte Carlo simulation.
MAE (maximum adverse excursion): How far a trade moved against you before it closed. Shows how much heat your trades take, and whether stops are too tight or too loose.
MFE (maximum favourable excursion): How far a trade moved in your favour before it closed. Comparing it with the final result shows how much profit the exits give back.

Win rate: The percentage of trades that made money. Meaningless on its own; a low win rate can be very profitable if winners are much bigger than losers.
Number of trades: The sample size. A few dozen trades can't separate skill from luck; the more trades, the more a result means.
Consecutive wins and losses: The longest winning and losing streaks. Plan for losing streaks longer than the backtest shows.
R-multiple: A trade's profit or loss divided by the amount risked on it. A trade that risked 100 and made 250 is +2.5R.
SQN (system quality number): Van Tharp's measure of a system, based on the average and spread of R-multiples and the number of trades. It rewards consistent results over many trades.
Z-score: A runs test that checks whether wins and losses cluster into streaks more than chance would explain. Strongly negative means streaky; strongly positive means wins and losses alternate.
Long and short split: Statistics for long trades and short trades measured separately. A strategy that only works in one direction can look healthy in the overall numbers.
Commission and slippage: The cost of trading and the difference between expected and actual fill prices. Leaving them out flatters every backtest, especially for strategies that trade often.

Overfitting: Tuning a strategy so closely to past data that it describes the past rather than an edge. The most common reason a great backtest fails live.
In-sample and out-of-sample: In-sample data is what a strategy was built or optimised on. Out-of-sample data was kept aside and never used for building. Only out-of-sample results say anything about the future.
Walk-forward testing: Repeatedly optimising on one window of data and testing on the next, then stepping forward. It simulates how the strategy would have been rebuilt and used over time.
Monte Carlo simulation: Reshuffling or resampling a strategy's own trades thousands of times to see the range of outcomes the same edge could have produced, including worse drawdowns than the one that happened.
What-if analysis: Re-running the record under changed assumptions, such as without the best trades, at a different position size, or long only, to find hidden dependencies.
Parameter sensitivity: How much results change when inputs change slightly. Robust strategies have a broad range of good settings, not one sharp peak.
Market phase: A period with a distinct character, such as trending or ranging, or high or low volatility. Testing by phase shows when a strategy works and when it doesn't.
Correlation: How closely two strategies' results move together. Combining strategies with low correlation can reduce a portfolio's drawdown.

Payout: The fixed percentage returned on a winning binary option, for example 80%. A losing option costs the whole stake.
Expiry: How long a binary option runs before it settles. Results are often very sensitive to the expiry chosen.
Break-even win rate: The win rate needed to break even at a given payout. It's 1 ÷ (1 + payout); at 80% payout that's 55.6%.
Martingale: Increasing the stake after each loss to recover it with the next win. It raises the win rate per series but risks very large losses in a long losing streak.

JSK TradeProof calculates them from a TradingView or broker export, inside Excel or Google Sheets.