Quantitative core

Performance & risk metrics

Expectancy, profit factor, drawdown, Sharpe/Sortino/Calmar, MAE/MFE and net edge — plus the live risk metrics (portfolio heat, daily loss, consecutive losses).

A strategy is judged on measured, net performance — not on a good story. These versioned metrics decide whether an idea advances through the validation gates, and the live risk metrics decide whether the next trade is even allowed.

Edge metrics

performance_metrics_v1

Expectancy

E x p ec t an cy = W in R a t e \times A v g W in - L oss R a t e \times A v g L oss

Average profit per trade. Positive net expectancy — after costs — is the bar a strategy must clear. Also tracked in R.

performance_metrics_v1

Profit factor & payoff

P r o f i tF a c t or = \frac{G r oss P r o f i t}{G r oss L oss} P a y o f f = \frac{A v g W in}{A v g L oss}

performance_metrics_v1

Breakeven win rate

B r e ak e v e nW R = \frac{A v g L oss}{A v g W in + A v g L oss}

The win rate a given payoff needs just to break even — a high win rate alone proves nothing.

Drawdown & risk-adjusted return

performance_metrics_v1

Drawdown & maximum drawdown

D D_{t} = \frac{E q u i t y _{t} - P e ak E q u i t y _{t}}{P e ak E q u i t y _{t}} M D D = t min D D_{t}

performance_metrics_v1

Sharpe, Sortino, Calmar

S ha r p e = \frac{R ˉ - R _{f}}{σ _{R}} S or t in o = \frac{R ˉ - T}{σ _{d o w n}} C a l ma r = \frac{A nn R e t u r n}{∣ M D D ∣}

Return per unit of risk — total, downside-only, and against the worst drawdown. The risk-free rate may be 0 when documented.

$R_{f}$: risk-free rate (0 if documented)
$σ_{d o w n}$: downside deviation below target T

performance_metrics_v1

Recovery factor

R eco v er y F a c t or = \frac{N e tP r o f i t}{∣ M D D ∣}

Excursion & net edge

performance_metrics_v1

MAE & MFE (long)

M A E = \frac{L o w _{t r a d e} - E n t r y}{E n t r y} M F E = \frac{H i g h _{t r a d e} - E n t r y}{E n t r y}

The worst and best unrealized moves during a trade — how much heat it took, and how much it left on the table.

performance_metrics_v1

Net edge after costs

N e tE d g e = G r oss E x p ec t an cy - A v g F ees - A v g S l i pp a g e - A v g O t h er C os t s

Clawlas prioritizes net over gross. A gross edge that disappears under realistic costs is not an edge.

Also tracked: exposure time, average holding duration, win/loss rate, expectancy in R.

Live risk metrics

risk_metrics_v1

Portfolio heat

T o t a l O p e n R i s k = i = 1 \sum m P l ann e d L os s_{i} P or t f o l i oH e a t = \frac{T o t a l O p e n R i s k}{E q u i t y}

The fraction of equity at risk across all open positions at once — the cap that throttles new entries.

Metric	Definition
Risk per trade	MaxPlannedLoss / Equity
Daily loss	StartOfDayEquity − CurrentEquity
Daily drawdown %	(CurrentEquity − StartOfDayEquity) / StartOfDayEquity
Consecutive losses	Current and maximum losing streak
Correlated exposure	Quote / chain / sector / venue / strategy concentration

quote concentration
chain concentration
sector concentration
venue concentration
market-beta exposure

Net beats gross; a high win rate proves nothing

Net performance matters more than gross, and win rate alone never proves profitability. A breach of any live limit — portfolio heat, daily loss, consecutive losses — raises a risk event with an automatic action. See Risk engine.