What it is

Strategy metrics are the summary statistics that turn a list of trades into an honest verdict about a strategy. A trading record on its own - a column of wins and losses - tells you almost nothing. The metrics distil it into two questions that actually matter: does this make money over many trades? and can I psychologically and financially survive holding it? The three most important answers are expectancy, the win-rate versus payoff relationship, and drawdown.

These numbers only mean something across a sample. A single trade has no expectancy and no drawdown worth discussing; the metrics describe the distribution of outcomes a strategy produces when repeated many times.

How it works

Expectancy is the single most important profitability metric. Expectancy is the average profit or loss you expect per trade, expressed in money or in units of risk (R). A common formulation:

• Expectancy = (Win rate x Average win) − (Loss rate x Average loss)

If a strategy wins 40% of the time with an average win of 2R and loses 60% of the time with an average loss of 1R, expectancy = (0.40 x 2R) − (0.60 x 1R) = 0.80R − 0.60R = +0.20R per trade. Positive expectancy means that, on average, each trade adds value; negative expectancy means the strategy bleeds money no matter how exciting individual trades feel. Expectancy is the bottom line: a strategy with positive expectancy and enough trades will tend to grow an account; one with negative expectancy will not, regardless of any other attractive feature.

Win rate versus payoff. The win rate is the share of trades that close profitable. Beginners fixate on it, but on its own it is almost meaningless, because it trades off against the payoff (the risk-reward ratio, average win divided by average loss). The two combine to produce expectancy, and there are two healthy archetypes:

• High win rate, low payoff - e.g. a mean-reversion style that wins 70% of the time but takes small profits and the occasional larger loss.
• Low win rate, high payoff - e.g. a trend-following style that wins only 35% of the time but lets winners run far beyond the size of the losers.

Both can be excellent; both can be terrible. A 70% win rate with a payoff below break-even is a losing strategy, and a 30% win rate with a 4-to-1 payoff is a strong one. The minimum win rate you need rises as your payoff falls - the precise break-even point is where (Win rate x Average win) equals (Loss rate x Average loss). Judging a strategy by win rate alone is the most common beginner error in this whole topic.

Drawdown answers the survival question. Drawdown is the peak-to-trough decline of the account or strategy equity, usually quoted as a percentage. Maximum drawdown is the worst such fall over the test period. It matters for two reasons. First, the mathematics of recovery are brutally asymmetric: a 50% drawdown requires a 100% gain just to get back to even, and an 80% drawdown requires a 400% gain. Second, drawdown is a psychological stress test - most traders abandon a perfectly good strategy somewhere inside a deep drawdown, locking in the loss right before the recovery. A strategy you cannot sit through is, in practice, a strategy you do not have.

How to use it

Use the three together; no single number is sufficient. A practical checklist for evaluating any strategy result:

1. Compute expectancy first. If it is not clearly positive after realistic costs, stop - nothing else can rescue a negative-expectancy system.
2. Decompose expectancy into win rate and payoff. Understand which archetype you are holding, because it tells you what to expect emotionally. A 35%-win trend system will hand you long losing streaks that are completely normal; if you do not expect them, you will quit.
3. Read the maximum drawdown and the longest losing streak. Ask honestly: could I keep following the rules through a drawdown of that depth and a losing streak of that length? If not, the strategy is too aggressive for you even if its expectancy is high.
4. Relate drawdown to position sizing. Most drawdown is something you choose: larger size amplifies both expectancy and drawdown. If the drawdown is intolerable, the usual fix is to risk less per trade, not to abandon a positive-expectancy edge.
5. Check the sample size. All three metrics are noisy on small samples; a handful of trades cannot establish any of them.

Worked example: two strategies each show +0.20R expectancy. Strategy A wins 65% with a 0.7-to-1 payoff and a 12% max drawdown; Strategy B wins 30% with a 3-to-1 payoff and a 35% max drawdown. They have the same expectancy, but they are completely different to hold. A trader who hates losing streaks belongs in A; one who cannot stand giving back open profit on the winners that do not work belongs in B. The metrics let you match the strategy to the person, not just to the math.

A few supporting numbers round out the picture. Profit factor - gross profits divided by gross losses - is a quick one-glance health check: above 1.0 the strategy made money, and most robust systems land somewhere between 1.3 and 2.0, while a backtest claiming a profit factor of 5 should arouse suspicion of overfitting. The longest losing streak deserves its own line because it, more than the average, is what actually breaks discipline; a positive-expectancy system can easily string together eight or ten consecutive losers, and you must know that number before you trade so the streak feels normal rather than terrifying. It also helps to express results per unit of risk (R) rather than in raw currency, because R-multiples are comparable across instruments and across position sizes: a strategy that averages +0.3R per trade tells you something portable, whereas "it made 1,800" tells you almost nothing without knowing how much was risked to earn it.

Strengths & limits

The strength of these metrics is that they make a strategy comparable, honest, and survivable. Expectancy tells you whether the edge is real; the win-rate/payoff split tells you what the ride feels like; drawdown tells you whether you can stay on for the whole ride. Together they convert a vague sense of "this works" into numbers you can size against and stick with.

The limits are the usual ones. Every metric is a backward-looking estimate from a sample, so it carries uncertainty and can shift as conditions change - an expectancy measured in a calm bull market may not survive a bear market. Drawdown in particular is almost always understated by history: the worst drawdown you have seen is, by definition, only the worst so far, and the real one is usually still ahead. Treat the numbers as informed estimates with error bars, size as though the true edge is smaller and the true drawdown deeper than measured, and never let a single flattering metric override the others.

Key takeaway: Judge a strategy by expectancy (does it make money), then by the win-rate/payoff mix (what the ride feels like), then by drawdown (whether you can survive it) - never by win rate alone, and always assume the real drawdown is deeper than the one history has shown you.