Loss Event Model

Quantify a risk scenario

Turn a handful of calibrated estimates into an annual loss distribution, value-at-risk, and a loss exceedance curve. Rigorous method, minimal inputs.

Free · no account required · no usage limits · All calculations run in your browser

Threat & frequency

How often, how capable

F — Annual threat frequency 0 = never · 1 = every year

s — Years with events

n — Years observedF = (s + 1) / (n + 2)

min — Minimum plausible

mode — Most likely

max — Maximum plausible

Confidence

T — Threat strength0 to 1

P — Protection strengthceiling = 1/(1+P)

Threat — — Protection

Conditional probability protection fails · T/(T+P) = —

P(loss event) = F × T/(T+P) · F = —

—

Primary loss

90% confidence interval

L₅th — Lower bound5th percentile · you'd be surprised below

U₉₅th — Upper bound95th percentile · you'd be surprised above

Note — these are percentiles, not hard caps. About 5% of simulated draws will exceed U₉₅th. That tail is what drives VaR 99%.

Secondary loss

Fines · churn · reputation

S — P(secondary | primary)0 to ignore

L₅th — Lower bound

U₉₅th — Upper bound

Simulation

Monte Carlo

Iterations10k fast · 100k finer tail

Currency labelshown on outputs

No account No tracking No data transmission Free forever Open source

Every calculation runs locally in JavaScript. No inputs, results, or scenario data are transmitted to any server — there is no backend. Nothing is stored between sessions. This page enforces connect-src: none in its Content Security Policy, which you can verify in your browser's developer tools under the Network tab.

Questions or feedback?

Expected annual loss

—

Run the model to see results

Median

—

VaR 90%

—

VaR 95%

—

VaR 99%

—

Loss exceedance curve

—