Hi All, When computing VaR on historical data, what historical data should be stored for a full revaluation approach? 1) For instance, I could store two previous years bootstrapped curves, take the ratio from one day to the next in the term structure, shock the current zero curve to each of the ratios, revaluate and take the 95th percentile of my returns... or 2) I could store two previous years of overnight rates, eurodollar futures, and swap rates. Take the ratio from one day to the next of each of those, shock the current rates, and rebootstrap at each time step, value, and then take the 95th percentile. I suppose the same question could apply to historical monte-carlo counterparty credit scenarios - should the parameters be based on zero rates or the inputs (overnight rates, eurodollar futures, and swap rates) with rebootstrapping?Thanks!