Thursday, I will be in Paris, to discuss the results we got from the pricing game. I will present 12 ou 13 models sent to me, an discuss what happened when I created a market, where the models were competing. One or two models were clearly underestimating the losses, so with the results as they were send, each time, one company goy 80% market share and over 250% loss ratio. So I decided to normalize all the premiums, so that the average premium was the same, for all the companies. Slides are now available.
Hi Arthur,
I am unable to follow the slides. Is it possible to share the codes written in R , SAS or VBA and the assumptions made for the sake of reproducibility?.
Merci.
no I can’t, there is still some research going on on those data
those are the slides of the one-hour talk, a lot of (additional) information will be given during this hour
Hello Arthur,
I will not be able to be present at the presentation in Paris.
I submitted a model and would like to know how it performed.
Could you let me know which model (A3, A4, …, A5?) corresponding to the letter F in slide 19?
Independently from market competition, it would be interesting to see how the models perform in predicting the loss per insuree.
As I understand from the conclusion, it seems that this criteria is not as relevant under competitive conditions.
Thanks,
I will send an email to everyone (individually) to let them know which one is their model.
So see how those model performed, it should be on slide 16: there is an animated graph, where you can click to see how each model did perform, on a Lorenz curve (or lift curve, I don’t know how to call it). Most of them performed similarly…
Very clear. Thanks