Performance terms

These are the numbers Alltra puts on a card. Each one below is defined by the arithmetic that actually computes it, not by the usual textbook wording — because where the two differ, the arithmetic is what you are reading on screen.

Net, gross, and what costs do to them

• Net P&L — cumulative realised profit and loss over the window, after commissions and fees. It is the hero figure on every performance surface, and it is what every other figure here is derived from.
• Gross profit — the winning trades summed, as a positive magnitude. It is the numerator of the profit factor and nothing else.
• Gross loss — the losing trades summed. It is the profit factor’s denominator. Most surfaces state it as a positive magnitude, so it reads as a size and the formatter applies the minus; the Trades page’s own aggregate carries the true negative sign instead.
• Realised — settled by a closed position. Every figure in this article counts closed trades only; an open position has realised nothing, so it is excluded from all of them rather than counted as a scratch.

Because outcome keys off the NET figure, a trade whose price moved in your favour can still be a loss. That is deliberate — costs are part of the result, and a card that called it a win would be describing the chart rather than the account.

The rates

• Decided trades — the denominator every rate is computed over. On the Performance dashboard it is wins plus losses plus breakevens; on the Trades page and the daily Report it is wins plus losses only.
• Win rate — winning trades divided by decided trades, as a percentage. Unknown rather than 0% when nothing was decided.
• Loss rate and breakeven rate — the same division for the other two outcomes. Over the Performance dashboard’s denominator the three sum to 100.
• Trade-weighted — how a rate is rolled up across days. A day with one trade and a day with fifty do not carry equal weight, so Alltra weights each day’s rate by its trade count instead of averaging the rates.

That difference in denominator is real and worth knowing: including scratches lowers the rate, excluding them raises it. Excluding a scratch flatters a trader who scratches often, which is why the dashboard includes them — but if the Performance board and the Trades table quote you two win rates, this is why, and neither is broken.

The two ratios that get confused

• Profit factor — gross profit divided by gross loss. Above 1.0 the winners outweigh the losers. It is undefined — not infinite — over a period with no losing trade.
• Avg Win/Loss (the payoff ratio) — average win divided by average loss. A SIZES ratio: it answers "when I win, is it bigger than when I lose".
• Win/Loss Ratio (the count ratio) — winning trades divided by losing trades, with breakevens in neither operand. A COUNTS ratio: it answers "how often". Undefined when there were no losers.

These two routinely point opposite ways, which is the whole reason they are separate figures. Eight small winners against two large losers is a count ratio of 4.00 and a payoff ratio below 1 — a trader winning most of the time and losing money doing it. The Performance dashboard’s Win/Loss Ratio is the COUNT ratio; the tile of the same name on the Trades page carries the PAYOFF ratio, and its subtext prints the raw counts so you can tell which you are looking at.

Profit factor also carries two display rules worth recognising. At or above 10 it prints 10.00+ rather than the six-figure ratio a single lossless week produces — and a window with winners and no losers prints that same 10.00+, which states a floor that is literally true rather than claiming a figure the arithmetic does not have. Below 30 trades that won or lost — breakevens add nothing to either side of this ratio, so they are not counted into its sample — it prints the sample size beneath the value. The number is real; it is not yet settled enough to conclude from.

Expectancy

• Expectancy — mean net P&L per trade, in money — what the average trade is worth. On the Performance dashboard it divides by decided trades; on a daily report it divides by that day’s total trade count.
• Expectancy (R) — the mean realised R, over ONLY the trades that carried a risk plan. A trade with no stop has no R, and counting it as 0R would drag the mean toward zero and report a disciplined trader as a mediocre one.

The R model

R is your own risk used as the unit of measurement, which is what makes a two-contract trade and a ten-contract trade comparable. All four figures come from the stop and target you logged, and every one of them is unknown rather than zero when the plan is missing.

• Risk per unit — the distance from your entry to your stop, as a price. Unknown when no stop was logged.
• Trade risk — risk per unit multiplied by quantity and by the contract’s point value — the money that was on the line if the stop had been hit.
• Realised R — net P&L divided by trade risk. A trade that made twice what it risked is +2.0R. Unknown when there was no stop or the stop sat at the entry price.
• Planned R:R — the distance to your target divided by the distance to your stop, rendered as 1 : 2.4. It is direction-checked — a long must target above the entry with the stop below — so an incoherent plan reads as unknown rather than as a plausible negative ratio.
• Effective R — net P&L divided by the maximum adverse excursion. It measures what you made against the worst the trade actually went against you, so it needs no stop at all.

Planned R:R read beside realised R is the discipline number in Alltra. A wide gap does not say the strategy is wrong; it says you are not trading the strategy you wrote down, which is a completely different problem and a far easier one to fix.

Excursions

• MFE — maximum favourable excursion — the best unrealised profit a trade reached before it closed. How much was on the table.
• MAE — maximum adverse excursion — the worst unrealised loss a trade reached, stored as a size rather than a negative. How much heat it took.

Both are per-trade columns on the Trades table and there is no period-level total for either — a "total MFE" across a month would add together excursions that never coexisted. They also arrive only where the source supplied them; Alltra does not reconstruct them from fills.

Averages, extremes and streaks

• Average win / average loss — gross profit divided by the number of winners, and gross loss divided by the number of losers. Wherever the loss total is stated as a magnitude the average is one too, and the formatter applies the minus.
• Largest profit / largest loss — the single best and single worst closed trade in the window — usually the two that decided it.
• Hold time — entry fill to exit fill, elapsed. It is also split by outcome into average hold on wins, on losses and on scratches, which is where cutting winners early becomes visible in one line.
• Trade streak — the current run of the same outcome, signed — positive for a run of wins, negative for a run of losses.
• Max consecutive wins / losses — the longest same-outcome run inside the window. Computed over a total ordering of your trades, so the figure cannot change between two refreshes.

Drawdown and the equity curve

• Equity curve — cumulative realised net P&L across the window, one point per calendar day including days you did not trade. It is a record of closed trading, not your account balance.
• Drawdown (the underwater curve) — the equity curve minus its own running peak. It is always at or below zero and it is exactly zero at every new high, which is what makes the two charts line up peak for peak.
• Max drawdown — the deepest point that curve reached. Percent drawdown is measured against the running peak, not against the starting balance.
• Average drawdown — the mean across only the days that WERE underwater. Averaging flat days in too would report a shallower drawdown the longer you go without one — the opposite of what the figure is for.
• Recovery factor — net profit divided by the size of the max drawdown — return earned per unit of pain. Undefined rather than perfect when you have never been underwater.

The equity curve keeps untraded days in it deliberately, so the x-axis is a real calendar and an underwater stretch is as wide on screen as it was in life. It is also why an account with no trades in the window draws no line at all rather than a flat one along zero.

Consistency

• Trading day — a day that carried at least one closed trade. It is the consistency score’s denominator, and deliberately not the number of calendar days in the window — a trader who only trades Tuesdays is not 20% consistent for doing exactly what they planned.
• Consistency score — profitable days divided by trading days, as a percentage. A day counts as profitable when its net is strictly above zero, so a breakeven day does not count.

Below ten trading days the tile shows its raw fraction — 1 / 1 over the word "days" — and draws no arc, because a share of days swings by more than ten points per day at that sample size and a percentage would read as far more settled than it is. On a single-day report it is always unknown for the same reason: over one day it could only ever be 0% or 100%.

Shown, but not computed yet

Three figures have a card in the product and no answer behind them. They are listed here so that finding an empty one does not read as a fault at your end.

• Alltra Score — a blended 0–100 reading over six axes — win rate, profit factor, average win/loss, recovery factor, max drawdown and consistency. It is published only when all six resolve, and the curves that turn a raw profit factor into a 0–100 axis have not been set, so the card states that it is unavailable rather than averaging what it has.
• Overtrading index — how densely you trade against a baseline you set for yourself. The index itself has no agreed formula and is reported as not computable; the footer figures beside it — trades taken, plan max, extra trades, periods over — are exact.
• Total swap — overnight financing. Trades carry no swap column, so it is reported as not computable wherever it appears.