> For the complete documentation index, see [llms.txt](https://docs.securemetrics.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.securemetrics.io/crq-pro/guides/risk-appetite.md).

# Risk appetite and ratings

Simulation output is a distribution; decision makers want a rating. The **risk appetite** bridges the two: you choose a statistic to rate against and dollar thresholds, and every result gets a **Low**, **Moderate**, **High**, or **Critical** badge.

## Configuring the workspace appetite

Go to **Settings > Workspace > Risk appetite**.

### Rate against

Choose which statistic is compared to the thresholds:

* **Expected annual loss (mean)**
* **P90 loss (1 in 10 year)**
* **P95 loss (1 in 20 year)** (default)
* **P99 loss (1 in 100 year)**

Rating against a tail percentile (the default P95) expresses "we care about bad years", while rating against the mean expresses "we care about the long-run average". Pick the one that matches how your organization talks about appetite.

### Thresholds

Set the dollar levels where the rating steps up:

| Rating                   | Default threshold |
| ------------------------ | ----------------- |
| Low                      | below Moderate    |
| **Moderate at or above** | $250,000          |
| **High at or above**     | $2,500,000        |
| **Critical at or above** | $25,000,000       |

A live preview scale shows the bands as you type, and validation keeps the thresholds in order.

## Per-scenario overrides

Some scenarios warrant their own appetite, for example a scenario scoped to a small subsidiary. In the scenario editor, switch **Risk appetite** from **Inherit** to **Custom** and enter Moderate, High, and Critical thresholds for that scenario alone.

{% hint style="info" %}
Ratings appear on the results badge, in the scenario table, and in generated reports, so aligning thresholds with your organization's actual appetite statement pays off everywhere at once.
{% endhint %}