Imagine we are pricing an interest rate derivative that depends on 2 different ratesi.e we pay 5yr CMS as long as 3mL<7%Imagine we price this option on a tree that is described with a short rate model.If we are calibrating to the 3monthLibor column (i.e implied vol of caps on 3m libor), and try to calculata vega, can someoneexplain to me how the "vega calculation mechanism" works?When we move libor vol up 1%, this leads to a different parametrization of the diffusionand thus a different price? Is this how we calc vega?If yes, is it only through the shocking of whatever the calibration index might bethat the price changes, thus allowing us to calculate vega?In other words, is the only thing that matters the size of the shock on the calibration index?If that were the case then a parallel shift of the vol surface would produce the same result asa steepening or flattening of the vol surface.Can someone explain to me how the vega calc mechanism really works?