# Why one week of risk is so noisy

Derive the noise of a variance estimate from squared shocks, averaging, chi-square shapes and kurtosis—then connect measurement to risk forecasts.

A simulated teaching example with formulas, symbol definitions and paced Zon–Coach conversation. Separate measurement uncertainty from forecast bias and information leakage.

Canonical page: https://21ifm.com/ml/why-one-week-of-risk-is-so-noisy

## Explore

- [Explore the connected chapter](https://21ifm.com/ml/learning-from-forecasts-in-markets)
- [Machine learning collection](https://21ifm.com/machine-learning)