Historical Simulation Methods

Historical simulation is a non-parametric method for calculating Value at Risk that uses historical market data to estimate potential future losses. Instead of assuming a normal distribution, this method looks at how the portfolio would have performed during past market conditions.

This makes it particularly useful for crypto, where asset returns often exhibit fat tails and other non-normal characteristics. By using actual past data, the model can capture the complex relationships between assets and the reality of market behavior.

However, the accuracy of this method depends on the quality and relevance of the historical data used. If the past does not reflect the future, the model may fail to predict significant losses.

It is a popular approach because it is intuitive and does not require the complex mathematical assumptions of parametric models. When combined with stress testing, it provides a robust framework for assessing risk in volatile markets.

It is a cornerstone of empirical risk management.