I am struggling with Ch14 and trying not to get bogged down with the detail, but keep getting dragged into it! Three questions; 1. Bootstrapping Fit model -> find residuals Create lots of new triangles with residuals Project each triangle with CL. Qu. p.18 Step 3 implies we might project the triangles with other methods to CL, I think this is wrong?? 2. ODP - Analytic This is my understanding ODP means we assume sigma = mu * phi Given data we can estimate mu and sigma of incremental claims Could assume incremental claims are logNormal and use mu and sigma to give estimated ditribution Qu. Is my understanding right? 3. ODP bootstrap Calculate fitted triangles using ODP Find residuals Create lots of new triangles with residuals Project each triangle with CL Qu. How do you calculate fitted triangles? ODP gives you an assumed distribution of inc claims how do you use this to get the fitted triangles? Also notes say this gives parameter variance? I don't understand this. Probably because I don't actually understand what we're doing! Trying to get the main points down with this chatper and move on but I'm really really struggling! Leela
