Stylized Facts of Financial Returns › Key Parameters

The key parameters for the statistical models and risk measures used in this chapter are:

• Confidence Level ($1-\alpha$): The probability that losses will not exceed the VaR threshold (e.g., 99%). Higher confidence levels result in more conservative (larger) VaR estimates.
• Degrees of Freedom ($\nu$): A parameter of the Student's t-distribution controlling tail heaviness. Lower values (e.g., 3-5) indicate fatter tails; as $\nu \to \infty$, the distribution converges to Normal.
• Lag ($k$): The time shift used in autocorrelation calculations. Significant autocorrelation at lag $k$ implies predictability based on observations $k$ periods ago.