Net profit is the first number everyone looks at in a Strategy Tester report, and it's also the least informative one on the page. Two strategies can post identical net profit while one grinds out steady gains and the other survives a 40% drawdown on the way. This guide works through the metrics that actually describe a strategy's behavior — what each one measures, how it's computed, where it misleads — and then covers the numbers the standard TradingView report doesn't show you at all.
Profit factor: the ratio everyone quotes and few interrogate
Profit factor is gross profit divided by gross loss. A profit factor of 1.5 means the strategy made $1.50 for every $1.00 it lost. It's a genuinely useful compression of a backtest into one number, with two important failure modes:
- Small samples lie. A profit factor of 3.0 over 25 trades is noise; over 500 trades it's a description. Always read profit factor next to trade count.
- Outlier dependence. One enormous winner can drag profit factor from 0.9 to 1.6. A quick sanity check: recompute it with the single largest winner removed. If the number collapses, the "edge" was one trade.
As a rough descriptive convention, values between about 1.2 and 2.0 are common for strategies with realistic assumptions, and values far above that in a backtest often signal overfitting or fill assumptions that won't survive contact with live spreads.
Max drawdown: the number that describes what it feels like
Max drawdown is the largest peak-to-trough decline in the equity curve. It answers a different question than profitability metrics: what is the worst stretch this strategy put its hypothetical account through? Two things the single headline number hides:
- Duration. A 15% drawdown recovered in a week reads very differently from a 15% drawdown that took eight months to recover. The Strategy Tester's equity chart shows this visually, but the exported number doesn't carry it.
- It's a single historical path. Max drawdown is the worst outcome of one specific ordering of trades. Reorder the same trades and the max drawdown changes. Treat it as a lower bound on what's possible, not a ceiling.
Sharpe ratio: return per unit of total volatility
The Sharpe ratio divides excess return by the standard deviation of returns. Higher means more return per unit of wobble. It's the lingua franca of performance comparison, and it has one well-known blind spot: it penalizes upside volatility exactly as much as downside volatility. A strategy with occasional large winning spikes gets its Sharpe dragged down by the very trades you'd want more of.
Sortino ratio: Sharpe's downside-only sibling
The Sortino ratio fixes that blind spot by dividing return by downside deviation only — volatility from losing periods. For strategies with asymmetric return profiles (trend followers, breakout systems with rare large winners), Sortino is usually the more honest descriptor. Reading Sharpe and Sortino together tells you something neither does alone: a Sortino much higher than Sharpe indicates the volatility is mostly on the upside; the two being nearly equal indicates symmetric, choppy returns.
Win rate and average win/loss: only meaningful as a pair
Win rate on its own is close to useless — a 90% win rate strategy that risks $500 to make $20 is a coiled spring. What matters is the pairing: win rate together with the ratio of average win to average loss. The break-even relationship is simple enough to keep in your head: a strategy wins overall when win rate × avg win > (1 − win rate) × avg loss. A 40% win rate needs the average winner to be at least 1.5× the average loser just to break even before costs.
The P&L distribution: what the averages are hiding
Average win and average loss are single points summarizing entire distributions, and the shape of those distributions matters. Plotting trade returns as a histogram answers questions the averages can't: Is the P&L a smooth bell around a small positive mean, or a barbell — hundreds of near-zero scratches plus a handful of extreme outcomes in both tails? Two strategies with identical average win/loss figures can have completely different distributions, and the barbell-shaped one is far more sensitive to whether those few tail trades repeat. When you see a fat left tail — occasional losses several times the size of a typical one — that's a structural feature of the historical record worth knowing about before any other metric, because a single tail event dominates months of averages.
Capture ratio by direction: is the edge long, short, or both?
Splitting performance by direction — how much of the long-side opportunity the strategy captured versus the short side — frequently reveals that a "strategy" is really two strategies stapled together, one of which is dead weight. In sample after sample, strategies backtested through a bull-market segment show respectable combined numbers that decompose into a strong long side and a short side that only subtracts. The combined report hides this; the directional split doesn't.
What the Strategy Tester report leaves out
The built-in report is a snapshot of aggregates. Several of the most decision-relevant views are simply absent:
| Missing view | Question it answers |
|---|---|
| P&L by entry hour | Which times of day generated the results — and which subtracted from them? |
| Rolling-window stability | Was the edge persistent across the sample, or concentrated in one lucky stretch? |
| Sortino ratio | How does the strategy look when only downside volatility is penalized? |
| Distribution of trade returns | Is the P&L a smooth distribution or a barbell of scratches and outliers? |
| Multi-strategy comparison | How do variants line up on the same time axis, same metrics, same scale? |
The stability question deserves emphasis because it's the one aggregate metrics are structurally unable to answer. A profit factor of 1.8 computed over three years tells you nothing about when the profit arrived. Slicing history into rolling windows and asking "in what share of windows was this strategy (or this entry hour) net profitable?" converts a single number into a persistence measurement. We walk through that method in detail in How to Find Your Strategy's Most Profitable Trading Hours.
Getting these numbers from your own export
Everything above is computable from the trade list in a Strategy Tester .xlsx export — if you haven't exported one before, the mechanics (and the parsing traps, like open positions appearing as fake exit rows) are covered in How to Export TradingView Strategy Tester Results to Excel. Sharpe and Sortino from raw trade rows involve a bit of spreadsheet plumbing; the rolling-window and hour-bucket views involve considerably more.
If you'd rather see the numbers than build them: drop your export into the free instant report — no signup, the file is processed in memory and never stored — and you get win rate, profit factor, max drawdown, average win/loss, and P&L by entry hour immediately. Free accounts add the core charts and risk metrics tables; Pro adds Sharpe/Sortino in AI-generated statistical summaries, the Rolling Window stability heatmap, and side-by-side comparison.
Reading a report in the right order
- Trade count first. Under ~100 trades, treat every other number as provisional.
- Profit factor with the biggest winner removed. Tests outlier dependence in ten seconds.
- Max drawdown against net profit. A strategy that made 30% while drawing down 35% along the way described a rough historical ride, whatever the final number says.
- Sharpe next to Sortino. The gap between them describes the shape of the volatility.
- Then the views the report doesn't show — hourly attribution, directional split, rolling stability — because that's where single-number summaries go to die.
None of these metrics predicts the future; all of them describe the past more honestly than net profit does. The goal isn't to find the one true number — it's to make it hard for any one number to fool you.