medium · FRM Part 1 Quantitative Analysis

If a risk manager is concerned about 'ghosting' effects in their volatility estimates and decides to compare an equally weighted moving average to an EWMA model using a bootstrap simulation, which result would they likely observe in the standard errors of their forecasts?

1. The equally weighted average will have a lower bootstrap standard error because it uses more data points (the entire window) in every calculation.
2. The bootstrap standard errors will be identical because both models are ultimately estimators of the same population variance.
3. The EWMA will show more stable bootstrap standard errors because it doesn't suffer from the discrete data-exit shocks of equally weighted windows.
4. The EWMA bootstrap standard error will be infinite because the decay factor λ creates a non-stationary variance process.

Sign up free to see the explanation and track your rank →