Open an EA's results page and you get a wall of numbers: net profit, gross profit, profit factor, expected payoff, Sharpe ratio, absolute drawdown, relative drawdown, largest win, average consecutive wins, and a dozen more. Most of them are filler. A handful decide whether the strategy is worth your money, and they only make sense read in a particular order, because each one sets the context for the next.
Here's the order a serious trader reads them in, and how they hang together.
The one-line test: every number on the page should come from the same set of trades, over the same dates, net of the same costs. The moment one tile is measured differently from the tile beside it, the page stops telling a single story and starts telling several — and that's where the lying begins.
1. Net profit — after commission and swap
Net profit is the "did it make money" question, and it's the easiest one to fudge. The trap is gross versus net. MetaTrader's headline profit figure is often quoted before commission and swap — the two costs that don't appear in a trade's profit column but come straight out of your balance.
An EA reports £6,400 profit over a year on a £10,000 account. Reads well. But it's a high-frequency strategy, and once you subtract £1,900 of commission and £280 of swap on positions held overnight, the real net profit is £4,220. Still a good year, but the number that mattered was never the one on the poster.
The check: net profit should reconcile with where the equity curve actually ends. If the curve runs from £10,000 to £14,220 and the tile claims £6,400, the gap is the costs nobody subtracted.
2. Profit factor — the quick quality read
Profit factor is gross profit divided by gross loss. Above 1.0 it made money; 1.5 is healthy; below 1.0 it lost. It's the fastest read of quality — and it's carried by outliers and wrecked by small samples.
Take a profit factor of 1.5, from £12,000 of gross profit and £8,000 of gross loss. Now find the biggest single trade — £5,000, a news spike the strategy happened to catch. Strip it out and you have £7,000 against £8,000, a profit factor of 0.875. A loser that got rescued once. One trade you can't count on is holding the whole record up. Read more on profit factor →
3. Expectancy — the pounds per trade
Profit factor gives you the ratio; expectancy gives you the money. It's what you can expect to make, or lose, on an average trade, and it ties win rate to the size of wins and losses in one figure:
expectancy = (win rate × average win) − (loss rate × average loss)
Say 60% win rate, average win £80, average loss £110. Expectancy is (0.60 × 80) − (0.40 × 110) = 48 − 44 = £4 a trade. Positive, but thin — over 500 trades that's £2,000, and a small rise in the average loss wipes it out. Expectancy is where a "high win rate" strategy often quietly falls apart: win small, lose big, and 60% winners still nets pennies. Read more on expectancy →
4. Maximum drawdown — equity, not balance
Everything above is reward. Drawdown is the price. Maximum drawdown is the deepest peak-to-trough fall in the account, and the only version worth reading is equity-based, not balance-based.
Balance drawdown counts a trade only once it's closed. Equity drawdown counts open floating losses too — the ones a grid or martingale sits on while it waits for a reversal. A strategy can show an 8% balance drawdown and a 22% equity drawdown over the same stretch, and the 22% is the one that would have margin-called you.
Picture the £10,000 account: balance never dips more than £800, but at one point three open positions are £2,200 underwater. Real maximum drawdown, 22%. If your broker's stop-out sat at 20%, the balance figure would have said everything was fine right up to the liquidation. Read more on maximum drawdown →
5. Win rate — only meaningful in context
Win rate is the most quoted number and the most meaningless on its own. It means something only read next to average win and average loss. A martingale can post a 90% win rate and be one bad streak from zero. A trend follower can win 35% of the time and be excellent, because its winners are several times its losers.
Two EAs. A: 90% win rate, average win £20, average loss £260. B: 35% win rate, average win £300, average loss £90. Run the expectancy. A is (0.9 × 20) − (0.1 × 260) = 18 − 26 = −£8. B is (0.35 × 300) − (0.65 × 90) = 105 − 58.5 = +£46.50. The 90% winner loses money; the 35% winner makes it. Win rate told you the opposite of the truth. Read more on win rate →
6. Return versus drawdown — reward against pain
The last number ties the two halves together: return against drawdown. Net profit alone is meaningless without knowing what you suffered to earn it. Two common forms:
- Recovery factor = net profit ÷ maximum drawdown (in money)
- Calmar = annualised return ÷ maximum drawdown (in %)
Our EA netted £4,220 (42%) with a 22% maximum drawdown (£2,200). Recovery factor = 4,220 ÷ 2,200 = 1.9. Calmar = 42 ÷ 22 = 1.9. Below 1 means you rode a drawdown as deep as a whole year's profit — rarely worth it. Above 2 or 3 is genuinely good. This is the single number that stops a fat headline return from impressing you before you've asked what it cost.
The traps that make each one lie
Every metric above lies in the same few ways. Learn the patterns once:
- Mixing populations. A lifetime profit factor sitting next to a this-quarter net profit. Each is honest; together they describe two different strategies. Check the date range and trade count on every tile.
- Gross versus net. Costs quoted or omitted inconsistently. Reconcile to the equity curve.
- Tiny samples. A profit factor of 3 over 18 trades is noise. Under roughly 30 trades, don't trust any ratio.
- Single outliers. One trade carrying the record. Strip the biggest win and see what's left.
- Vanity metrics. Largest winning trade, longest winning streak, gross profit on its own — they fill the page and tell you nothing about whether the edge is real.
Read the six that matter, in order, from the same set of trades, and most bad EAs disqualify themselves before you reach the pretty equity curve.
The Karnek note
Karnek computes every one of these from your live trades, net of commission and swap, over a date range you choose — so the profit factor and the net profit on your screen always come from the same set of trades. It reads read-only and can never place a trade; it just makes sure the numbers reconcile before you trust them.
Comparing tools? The best EA monitoring software (2026) →
Written and reviewed by the Karnek Research team. Last updated August 2026.
Educational content only - not financial advice. Past performance does not predict future results. Trading carries significant risk.