Estimate model risk reserve

What it's for

Adds a visible capital buffer when credible models disagree about severe downside instead of hiding model uncertainty inside one chosen forecast.

What you give it

Inputs split into evidence read from your connected systems, calibration your team owns, and numerical controls that affect precision but never the result's meaning.

{
  "material_reserve": 50,
  "models": [
    {
      "id": "operating-plan",
      "model_weight": 0.5,
      "scenario_losses": [
        250,
        100,
        20,
        -30
      ]
    },
    {
      "id": "market-stress",
      "model_weight": 0.3,
      "scenario_losses": [
        500,
        180,
        40,
        -20
      ]
    },
    {
      "id": "delivery-stress",
      "model_weight": 0.2,
      "scenario_losses": [
        350,
        220,
        60,
        0
      ]
    }
  ],
  "reserve_confidence_level": 0.9,
  "scenario_probabilities": [
    0.1,
    0.2,
    0.3,
    0.4
  ],
  "tail_probability": 0.1
}

What you get back

This is the actual output of running the example above — computed by the same function the platform calls, not an illustration.

{
  "assumptions": [
    "Models are credible approved alternatives over one loss unit, horizon, and aligned scenario set; cosmetic variants do not count as independent evidence.",
    "Model weights and reserve confidence reflect governance rather than posterior truth unless separately justified.",
    "The reserve covers represented between-model tail disagreement, not shared misspecification, omitted scenarios, liquidity, or parameter uncertainty within each model."
  ],
  "configuration": {
    "model_weight_sum": 1,
    "reserve_confidence_level": 0.9,
    "scenario_count": 4,
    "tail_probability": 0.1
  },
  "decision": "model_risk_reserve_material",
  "method": "model_disagreement_cvar_reserve_v1",
  "model_diagnostics": [
    {
      "cvar_difference_from_weighted": 155,
      "cvar_loss": 500,
      "expected_loss": 90,
      "is_at_or_above_reserve_quantile": true,
      "model_id": "market-stress",
      "model_weight": 0.3
    },
    {
      "cvar_difference_from_weighted": 5,
      "cvar_loss": 350,
      "expected_loss": 97,
      "is_at_or_above_reserve_quantile": false,
      "model_id": "delivery-stress",
      "model_weight": 0.2
    },
    {
      "cvar_difference_from_weighted": -95,
      "cvar_loss": 250,
      "expected_loss": 39,
      "is_at_or_above_reserve_quantile": false,
      "model_id": "operating-plan",
      "model_weight": 0.5
    }
  ],
  "summary": {
    "material_reserve": 50,
    "model_count": 3,
    "model_risk_reserve": 155,

How it works

Statistical audit & measurement — Check whether a number is fit to decide on: coverage, timing, reconciliation, and the gaps a dashboard hides.

1. 1 Freeze credible non-cosmetic model alternatives, weights, one loss unit/horizon, aligned scenarios/probabilities, tail probability, and reserve confidence before comparison.
2. 2 Calculate each model's expected and CVaR loss, their weighted risk and dispersion, then take the governed weighted upper quantile of model CVaR as the conservative risk basis.
3. 3 Reserve the positive difference from weighted CVaR and disclose that it covers represented between-model disagreement rather than shared misspecification.

Before you trust it

Every tool in the catalog ships with the conditions under which its answer is meaningful — and the conditions under which it should abstain instead of guessing.

Calibrating it to your org

Calibration workflow

1. 1 Define the management decision, target outcome, aggregate unit, privacy boundary, cadence, and prediction/intervention horizon for this organization.
2. 2 Build a tenant-scoped historical cohort using only information available before each prediction or decision; preserve zero periods, censoring, assignment probabilities, and unresolved outcomes when the method requires them.
3. 3 Estimate statistical parameters on training history, but obtain costs, utilities, risk tolerance, practical-effect thresholds, capacity, and policy constraints from accountable decision owners.
4. 4 Validate on later time windows or held-out aggregate units at the deployment grain, against a simple baseline and the function-specific validation strategy.
5. 5 Deploy only if the returned decision clears evidence, overlap, calibration, robustness, and guardrail checks; warning, unsupported, schema-gap, and fallback decisions are abstentions.
6. 6 Monitor realized outcomes, data drift, coverage, and decision regret; recalibrate at a governed cadence or after a detected regime/definition change, never merely because a stakeholder dislikes the result.

Call it from your AI

You don't wire up 388 tools in your MCP client. The GitRevio MCP server exposes 18 tools, three of which let an agent search the catalog, read a tool's schema, and run it — so the assistant finds this one on its own.

gitrevio_capabilities_search
  { "q": "calculate an explicit modelrisk reserve from" }
  → finds "estimate_model_risk_reserve"

gitrevio_capability_describe
  { "capability_id": "estimate_model_risk_reserve" }
  → returns the input schema and agent guidance shown on this page

gitrevio_capability_run
  { "capability_id": "estimate_model_risk_reserve", "arguments": { ... } }
  → returns the result shown above

Works in Claude Desktop, Claude Code, Cursor, Cline, Continue.dev, Goose and Aider. See the MCP server.

Related tools

See every tool in Finance & unit economics →