Skip to content

How the numbers work

These definitions come from the same metric registry as the journal. Each formula states when its result is unavailable.

A worked example

Suppose a USD account closes a trade with $150 gross P&L, $10 in costs, zero swap and $100 of initial risk. Its net result is $140, or 1.40R.

Across three settled trades of +$100, −$50 and $0, if the first two fall outside the breakeven band, win rate is one win divided by two decided outcomes, or 50%.

Average holding time

average holding time across settled closed trades with valid instants

When the metric is unavailable
No eligible holding times.

  • Samples below 30 are labelled small.

Average loss

Σ(losing net_pnl) / losses

When the metric is unavailable
No losing trades.

  • Computed over closed trades with a settled P&L; blocked and FX-pending trades are counted separately.

Average win

Σ(winning net_pnl) / wins

When the metric is unavailable
No winning trades.

  • Computed over closed trades with a settled P&L; blocked and FX-pending trades are counted separately.

Breakeven band

account setting, else 0.1 × initial_risk, else 0.001 × balance at open, else 5
  • The basis used is shown beside the band.

Capture ratio

net_pnl / MFE

When the metric is unavailable
MFE is zero or unknown.

Expectancy

(win_rate × avg_win) − ((1 − win_rate) × |avg_loss|)

When the metric is unavailable
Every trade in the sample is a breakeven.

  • Expected result per non-breakeven trade.
  • Computed over closed trades with a settled P&L; blocked and FX-pending trades are counted separately.

Expectancy (R)

mean(r_multiple) over trades where it exists

When the metric is unavailable
No trade in the sample has an initial stop.

  • Always shown with its coverage; below 80% it is flagged as a self-selected sample (review S1).

Gross P&L

long: Σ(exit qty × exit price × multiplier) − Σ(entry qty × entry price × multiplier); short: the inverse

When the metric is unavailable
The trade is open, or its contract multiplier is unknown (it is never assumed to be 1).

  • Computed in the instrument quote currency, then converted at the rate persisted on the trade.
  • For same-currency imports, uses the broker’s posted closing amounts when every full closing fill agrees at currency precision.

Holding time

elapsed seconds between the opening and closing instants

When the metric is unavailable
The trade is still open or an opening or closing instant is invalid.

  • Includes time outside trading sessions.

Initial risk

|avg_entry − initial_stop| × quantity × multiplier

When the metric is unavailable
No initial stop was recorded, the stop equals the entry, or the multiplier is unknown.

  • Converted to the account currency at the rate persisted on the trade.

Maximum adverse excursion

long: (min_low − avg_entry) × quantity × multiplier; short: (avg_entry − max_high) × quantity × multiplier

When the metric is unavailable
No price data covers the holding window (CFDs without a proxy underlying, review S12).

  • Pending until the price backfill for the window completes.

Max daily drawdown

largest single-trading-day decline in realised net_pnl, on the account trading day
  • Day-over-day realised, not intraday floating equity: it will differ from a prop firm dashboard (review A2).

Max drawdown

max peak-to-trough decline of the cumulative realised net_pnl curve
  • Deposits and withdrawals are not in the curve.
  • The performance figure, not the risk-limit one.

Longest loss streak

longest consecutive run of losses in close order (then trade ID)

When the metric is unavailable
No settled closed trades with a valid close time.

  • Breakevens and unresolved closes break a run. Open trades are ignored.
  • Samples below 30 are labelled small.

Longest win streak

longest consecutive run of wins in close order (then trade ID)

When the metric is unavailable
No settled closed trades with a valid close time.

  • Breakevens and unresolved closes break a run. Open trades are ignored.
  • Samples below 30 are labelled small.

Maximum favourable excursion

long: (max_high − avg_entry) × quantity × multiplier; short: (avg_entry − min_low) × quantity × multiplier

When the metric is unavailable
No price data covers the holding window (CFDs without a proxy underlying, review S12).

  • Pending until the price backfill for the window completes.

Money-weighted return (annualised)

r such that Σ c_i / (1 + r)^(t_i) = 0; t_i in years (actual/365) from inception

When the metric is unavailable
No time has passed, or there is no money both in and out.

  • What the account earned on capital as it was deployed. Shown beside TWR, never instead of it (D4).

Net P&L

gross_pnl − commissions − fees + swap

When the metric is unavailable
Gross P&L is null, or the FX rate for a foreign-currency trade is still pending.

  • Swap is signed: negative when financing was charged, positive when credited.

Net P&L

Σ net_pnl, including breakeven trades
  • Computed over closed trades with a settled P&L; blocked and FX-pending trades are counted separately.

Win / loss / breakeven

win if net_pnl > band; loss if net_pnl < −band; otherwise breakeven

When the metric is unavailable
The trade is open, its P&L is blocked, or its FX rate is pending.

Planned reward:risk

|target − avg_entry| / |avg_entry − initial_stop|

When the metric is unavailable
No target or no initial stop was recorded.

Profit factor

Σ(winning net_pnl) / |Σ(losing net_pnl)|

When the metric is unavailable
There are no losing trades, so the ratio has no denominator.

  • Breakeven trades are in neither sum (review S2).
  • Computed over closed trades with a settled P&L; blocked and FX-pending trades are counted separately.

R-multiple

net_pnl / initial_risk

When the metric is unavailable
No initial stop was recorded for this trade. It is never estimated.

Return on balance

net_pnl / account balance at trade open

When the metric is unavailable
The balance at open is zero or unknown.

  • Uses the balance when the trade opened, not the current or starting balance.

Sortino ratio (per-trade R)

mean(R) / √(Σ min(R, 0)² / n)

When the metric is unavailable
Fewer than 10 eligible R trades or no downside below 0R.

  • Includes all settled closes with R. Coverage is eligible R trades divided by all settled closes.
  • Samples below 30 are labelled small.
  • The target is 0R and the denominator includes the entire sample. The ratio is not annualized.

System Quality Number

√n × mean(R) / sample standard deviation(R), where variance uses n − 1

When the metric is unavailable
Fewer than 10 eligible R trades or zero sample standard deviation.

  • Includes all settled closes with R. Coverage is eligible R trades divided by all settled closes.
  • Samples below 30 are labelled small.

Total commissions

Σ commissions
  • Computed over closed trades with a settled P&L; blocked and FX-pending trades are counted separately.

Trades

closed trades with a settled P&L, including breakevens
  • Open, P&L-blocked and FX-pending trades are counted separately, not here.

Time-weighted return

Π(1 + r_i) − 1, r_i = (ending_equity − starting_equity) / starting_equity per sub-period between cash flows

When the metric is unavailable
A sub-period starts with zero or negative equity.

  • The headline return: it measures the trader, not the funding schedule (D4).

Win rate

wins / (wins + losses)

When the metric is unavailable
Every trade in the sample is a breakeven.

  • Breakeven trades are excluded from both sides and shown as their own count.
  • Computed over closed trades with a settled P&L; blocked and FX-pending trades are counted separately.