Hello. A general query about selected development factors please. Can we use a chi square test to test the accuracy of our selections made one at year t-1 versus the actual claims? For example, suppose that we use the CL method based on incurred claims data and we project the claims according to our selected factors. One year later, can we test these selections versus actual claims movement with a chi square test for each year? and if yes, how many df will we have ? n-1? Thank you!
Am not sure if this answers your question I havent yet gone thro' it fully ...but it looks to be on the same lines https://www.casact.org/library/astin/vol10no1/78.pdf
I think that this method is not for comparing actual development factors versus expected development factor. Could somebody please confirm?
it may not be exactly wht u wanted...but maybe u cd get some idea of how u cd proceed and find what u want!! sorry if it wasnt helpful...as i had sd i dint go thro it fully! good luck!!!
