Trading Metrics That Actually Matter
Most traders track the wrong metrics. Here's which of the 100+ available trading metrics actually inform decisions — and how to read them together.
Tip
Key Takeaways
- Most traders over-index on win rate and net profit while ignoring the risk-adjusted and consistency metrics that actually predict forward performance
- The metrics that matter fall into three categories: profitability (is the edge real?), risk (can you survive the drawdowns?), and consistency (will it keep working?)
- No single metric tells the full story — you need to read metrics together to form an accurate picture
- Composite scoring systems distill 100+ metrics into actionable ratings, but understanding the underlying metrics makes you a better decision-maker
The Metric Overload Problem
Modern trading platforms can calculate dozens — sometimes hundreds — of performance metrics for any strategy. AlgoChef alone tracks over 100 metrics across profitability, risk, consistency, distribution, trade analysis, and time-based dimensions.
This creates a problem: which ones actually matter?
The answer depends on what decision you're trying to make. But for the most common decision — "should I trade this strategy with real money?" — a relatively small set of metrics provides the vast majority of the signal. The rest are useful for deep analysis but not essential for the go/no-go decision.
This guide walks through the metrics that experienced systematic traders rely on most, organized by the three questions every metric should help answer.
Category 1: Profitability — Is the Edge Real?
These metrics answer the most basic question: does this strategy make money, and how reliably?
Profit Factor
What it is: Total gross profit divided by total gross loss (expressed as absolute value).
Why it matters: Profit factor is arguably the single most useful summary metric for strategy quality. It tells you how many dollars you earn for every dollar you lose.
| Profit Factor | Interpretation |
|---|---|
| Below 1.0 | Strategy is losing money |
| 1.0-1.2 | Marginally profitable — barely covers costs |
| 1.2-1.5 | Moderate edge — viable but not exciting |
| 1.5-2.0 | Strong edge — the sweet spot for robust strategies |
| 2.0-3.0 | Excellent edge — but verify it's not overfitted |
| Above 3.0 | Suspicious — very likely overfitted unless sample is large |
The trap: A high profit factor based on few trades is meaningless. A profit factor of 4.0 over 20 trades could easily be random. A profit factor of 1.5 over 500 trades is far more reliable.
How to use it: Profit factor is your first filter. If it's below 1.3 after transaction costs, the strategy probably isn't worth further analysis. If it's above 2.5, check for overfitting before celebrating.
Win Rate
What it is: Percentage of trades that are profitable.
Why it matters: Win rate is the most intuitive metric — the percentage of time you're right. But it's far less informative than most traders think.
The trap: Win rate tells you nothing about magnitude. A strategy with 80% win rate where wins average $100 and losses average $500 is a losing strategy (expected value: 0.8 × $100 - 0.2 × $500 = -$20 per trade). A strategy with 35% win rate where wins average $1,000 and losses average $200 is highly profitable (expected value: 0.35 × $1,000 - 0.65 × $200 = +$220 per trade).
Context matters: Win rate must be read alongside payoff ratio (average win / average loss). Together, they define the strategy's edge. Alone, win rate is almost meaningless.
| Strategy Type | Typical Win Rate | Why |
|---|---|---|
| Trend-following | 35-45% | Many small losses, few large wins |
| Mean-reversion | 55-70% | Many small wins, few large losses |
| Breakout | 40-55% | Moderate — depends on filter quality |
| Scalping | 60-75% | High frequency, small edges per trade |
Expected Value per Trade (R-Expectancy)
What it is: The average profit or loss per trade, expressed in multiples of the risk taken (R). If you risk $200 per trade and your R-expectancy is 0.35R, you expect to make $70 per trade on average.
Why it matters: R-expectancy is more useful than raw average trade profit because it normalizes for position sizing. A strategy that makes $500 per trade sounds better than one that makes $150 — until you learn that the first risks $2,000 per trade (0.25R) and the second risks $200 per trade (0.75R).
The benchmark: Any positive R-expectancy indicates a theoretical edge. Above 0.3R is solid. Above 0.5R is excellent. Above 1.0R warrants scrutiny for overfitting.
The trap: R-expectancy is extremely sensitive to outlier trades. A few outsized winners can inflate R-expectancy significantly. Always check whether the R-expectancy holds after removing the top 3% of trades.
Net Profit and CAGR
What they are: Net profit is total money earned. CAGR (Compound Annual Growth Rate) is the annualized return.
Why they matter less than you think: Net profit is an outcome, not a diagnostic. It tells you what happened but not whether it's repeatable. A strategy with $100,000 net profit over 5 years sounds great — but if $80,000 of that came from three trades in March 2020, the "edge" is really just pandemic volatility.
CAGR is more useful because it normalizes for time, but it still masks the path — a strategy with 25% CAGR might have had a 40% drawdown along the way.
How to use them: Net profit and CAGR are useful for comparing strategies that have already passed validation. They shouldn't be the primary decision criteria.
Category 2: Risk — Can You Survive the Drawdowns?
A profitable strategy that destroys your account along the way isn't useful. Risk metrics tell you how much pain to expect.
Maximum Drawdown
What it is: The largest peak-to-trough decline in equity, expressed as a percentage or dollar amount.
Why it matters: Max drawdown is the most important risk metric for most traders because it represents the worst experience you'll have trading this strategy — the maximum amount of money you'll lose before recovering.
The reality check: Whatever your backtest max drawdown is, your live trading max drawdown will almost certainly be worse. Backtests underestimate drawdowns for multiple reasons: no slippage during volatile periods, no partial fills, no correlation breakdown. Plan for 1.5-2x the backtested max drawdown.
| Max Drawdown | Risk Level |
|---|---|
| Below 15% | Conservative — suitable for most risk profiles |
| 15-25% | Moderate — acceptable for experienced traders |
| 25-40% | Aggressive — requires significant risk tolerance |
| Above 40% | Extreme — most traders can't psychologically survive this |
Warning
The Drawdown Survival Question: Can you watch your account drop by the max drawdown amount without making emotional decisions? If your strategy has a 30% max drawdown and you have $100,000, that's a $30,000 decline. If seeing -$30,000 in your account would cause you to panic-sell, you need either a less volatile strategy or more capital (to reduce the drawdown percentage at the same position size).
Sharpe Ratio
What it is: The ratio of excess returns (above the risk-free rate) to the volatility of those returns. Higher is better.
Why it matters: Sharpe ratio is the gold standard risk-adjusted return metric in quantitative finance. It answers: "How much return am I getting per unit of risk?"
| Sharpe Ratio | Interpretation |
|---|---|
| Below 0.5 | Poor risk-adjusted returns |
| 0.5-1.0 | Acceptable — most strategies fall here |
| 1.0-1.5 | Good — genuine edge with reasonable risk |
| 1.5-2.0 | Very good — strong risk-adjusted performance |
| Above 2.0 | Exceptional — verify it's not overfitted |
The trap: Sharpe ratio assumes returns are normally distributed. Trading returns rarely are — they tend to have fat tails (more extreme events than a normal distribution predicts). Strategies with high Sharpe ratios but fat-tailed return distributions carry more risk than the Sharpe ratio suggests.
Complement with: Calmar ratio (return/max drawdown) or Sortino ratio (return/downside deviation) for a more complete risk picture.
Calmar Ratio
What it is: Annualized return divided by maximum drawdown. A Calmar of 2.0 means you earned $2 for every $1 of worst-case drawdown.
Why it matters: Calmar ratio directly connects return to the worst pain — making it one of the most practical risk-adjusted metrics for traders who care about drawdown (which should be everyone).
The benchmark: Above 1.0 is decent. Above 2.0 is strong. Above 3.0 warrants overfitting scrutiny.
Ulcer Index
What it is: Measures the depth and duration of drawdowns. Unlike max drawdown (which captures only the worst point), Ulcer Index penalizes strategies that spend a lot of time in drawdown — even if the drawdowns are moderate.
Why it matters: Two strategies can have the same max drawdown but very different Ulcer Index values. A strategy that draws down 20% once and recovers quickly has a low Ulcer Index. A strategy that draws down 15% repeatedly, spending 60% of its time underwater, has a high Ulcer Index — and is far more painful to trade despite the "lower" max drawdown.
How to use it: Ulcer Index is a "quality of life" metric. Low Ulcer Index strategies are easier to trade psychologically. High Ulcer Index strategies cause chronic stress even if the end result is profitable.
Recovery Factor
What it is: Net profit divided by max drawdown. Measures how much the strategy earns relative to its worst loss period.
Why it matters: A strategy with $50,000 net profit and $25,000 max drawdown (recovery factor: 2.0) is telling you that it earned twice what it lost at its worst point. Higher is better — it means the strategy's profitability comfortably compensates for its risk.
The benchmark: Above 3.0 is solid. Above 5.0 is excellent. Below 2.0 means the profitability barely justifies the drawdown risk.
Category 3: Consistency — Will It Keep Working?
Profitability and risk metrics describe the what. Consistency metrics describe the how reliably — and they're the best predictors of forward performance.
System Quality Number (SQN)
What it is: A statistical measure developed by Van Tharp that evaluates the quality of a trading system based on the distribution of R-multiples. It considers both the average R-expectancy and the standard deviation of R-multiples, scaled by the square root of the number of trades.
Why it matters: SQN is specifically designed to answer "how good is this system?" in a single number that accounts for both the quality and consistency of the edge.
| SQN | Van Tharp's Classification |
|---|---|
| Below 1.6 | Poor — difficult to trade profitably |
| 1.6-2.0 | Below average |
| 2.0-2.5 | Average — workable |
| 2.5-3.0 | Good |
| 3.0-5.0 | Excellent |
| 5.0-7.0 | Superb |
| Above 7.0 | Almost certainly overfitted |
Tip
The SQN Reality Check: An SQN above 7.0 is a red flag, not a badge of honor. Real trading systems rarely sustain SQN values that high. If your backtest shows SQN above 7, run the overfitting checklist before proceeding.
Payoff Ratio
What it is: Average winning trade divided by average losing trade (in absolute terms).
Why it matters: Payoff ratio and win rate together define the mathematical structure of your edge. A strategy with 40% win rate needs a payoff ratio above 1.5 to be profitable. A strategy with 65% win rate can be profitable with a payoff ratio below 1.0.
The relationship: Win Rate × Payoff Ratio should be meaningfully above 1.0 for the strategy to have edge. The higher the product, the stronger the edge — but both components matter.
Consecutive Losses (Max Losing Streak)
What it is: The longest run of consecutive losing trades in the history.
Why it matters: Max losing streak is a psychological survival metric more than a mathematical one. Most traders don't abandon strategies because of total drawdown — they abandon them because of consecutive losses. Eight losses in a row feels like the strategy is broken, even if the overall statistics are fine.
Planning for it: Calculate the expected max losing streak for your win rate: a 55% win rate strategy should expect runs of 7-9 losses over 200 trades. If your backtest shows a max of 5, the real max losing streak is probably longer — plan accordingly.
Time Underwater
What it is: The percentage of time (or trades) that the equity curve is below its previous high-water mark.
Why it matters: Time underwater captures the experience of trading the strategy. A strategy that spends 60% of its time in drawdown — even if the drawdowns are small — is psychologically draining. A strategy that spends 30% of its time underwater with similar returns is much easier to trade.
The benchmark: Below 40% is comfortable. 40-60% is manageable. Above 60% is difficult to sustain — expect to question the strategy regularly.
Reading Metrics Together: The Patterns That Matter
Individual metrics are useful. Metric patterns are powerful. Here are the combinations that reveal the most about strategy quality.
Pattern 1: High Win Rate + Low Payoff Ratio
What it means: The strategy wins often but wins are small relative to losses. Common in mean-reversion and scalping strategies.
The risk: A small decline in win rate destroys profitability. If the strategy needs 65% win rate to break even and is currently at 68%, there's only 3% margin before it becomes a loser. This is a fragile edge.
What to check: How stable is the win rate historically? Calculate the rolling 30-trade win rate — if it fluctuates by more than 5% regularly, this strategy is riskier than it looks.
Pattern 2: Low Win Rate + High Payoff Ratio
What it means: The strategy loses often but wins are large relative to losses. Classic trend-following profile.
The risk: Long losing streaks. A 40% win rate strategy will regularly have 6-8 loss runs. The payoff ratio makes up for it mathematically, but psychologically, most traders can't handle sustained losing. They'll kill the strategy during a losing streak — which, ironically, is often right before a large winning trade.
What to check: Max losing streak and time underwater. Can you actually sit through the drawdown periods this strategy demands?
Pattern 3: Good Sharpe + High Outlier Dependency
What it means: The risk-adjusted returns look good, but they depend on a few outsized winning trades. Remove those trades and the Sharpe collapses.
The risk: The "edge" is an illusion. The strategy doesn't have a consistent small advantage — it has a few lottery tickets that happened to pay off. Those specific events may never recur.
What to check: Recalculate all metrics after removing the top 3% of trades. If the strategy is still profitable with a reasonable Sharpe, the edge is robust. If it turns negative, walk away.
Pattern 4: Strong Returns + High Kurtosis
What it means: Returns are good on average, but the distribution has "fat tails" — extreme events occur more often than a normal distribution would predict.
The risk: Your max drawdown in live trading will likely be worse than the backtest suggests. Standard risk metrics (Sharpe, VaR) underestimate tail risk in high-kurtosis distributions.
What to check: Run Monte Carlo stress tests. Look at the 1st and 5th percentile outcomes, not just the median. Plan capital requirements based on stress-test drawdowns, not backtest drawdowns.
Pattern 5: Declining Rolling Metrics
What it means: Key metrics (win rate, profit factor, Sharpe) have been trending downward over recent rolling windows.
The risk: This is degradation in action. The strategy's edge is eroding. The headline metrics may still look acceptable because they include the strong earlier period, but the current state is deteriorating.
What to check: Run an IS/OOS comparison — baseline period vs. recent period. If multiple metrics show significant divergence, refer to the degradation detection guide.
From Metrics to Decisions
Metrics are tools, not answers. The goal isn't to memorize threshold tables — it's to develop the judgment to read metric patterns and translate them into trading decisions.
A few principles:
Never rely on one metric. Win rate alone is misleading. Profit factor alone is misleading. Sharpe alone is misleading. The power comes from reading them together as a system.
Adjust expectations by strategy type. A 45% win rate is excellent for a trend-following strategy and terrible for a mean-reversion strategy. A Sharpe of 1.0 is strong for a daily strategy and weak for a high-frequency strategy. Context matters.
Prefer risk-adjusted metrics over raw returns. Net profit and CAGR are exciting but don't account for the risk taken. Sharpe ratio, Calmar ratio, profit factor, and R-expectancy are more useful for decision-making because they balance return against risk.
Watch for degradation patterns. Metrics are snapshots — but degradation is a trend. The most important metric analysis isn't the current values but how the values are changing over time. A strategy with "acceptable" current metrics but declining trends is more concerning than a strategy with "mediocre" current metrics that are stable or improving.
See your strategy's 100+ metrics in 60 seconds →
Dive deeper: What Is Strategy Validation?, The Complete Guide to Strategy Degradation, or learn about Monte Carlo simulation.
Related Articles
How to Validate a Trading Strategy Before Going Live
A step-by-step pre-live checklist for systematic traders. 7 validation steps from backtest to live deployment — and the discipline to follow them.
What Is Strategy Validation (and Why Most Traders Skip It)
Strategy validation is the most important — and most skipped — step in algorithmic trading. Learn what it is, why it matters, and what happens when you skip it.
Curve-Fitting Checklist: Is Your Strategy Overfitted?
A practical checklist for detecting overfitting in trading strategies. 12 warning signs, testing methods, and the discipline to reject strategies that look too good.