This question gave you information on paid claims for 2009, ... ,2013 and said that the business started being written mid-way through 2009. The basic chain ladder had been given and the question was to comment on it and provide a new best estimate based on changing the data or methodology. One of my key points in answering this question was that since they only started writing the business mid-way through 2009 they probably improved their underwriting standards during this period which led to poor experience in 2009 and much improved experience subsequently. As such, in my solution for changing the best estimate I kept the 2009 data as it was and ignored it for the calculation of the 2010 and later years best estimate reserve. The solution didn't mention this at all, instead it carried out linear interpolation on the data which gave a slightly better triangle (from the point of view of the statistics calculated in part i.) but still not that good. When it came to the final part which asked you to re-assess the statistics etc I didn't have anything to write as I didn't edit the data. I think my method was fine but I would have lost out on the 5 marks at the end because I didn't think of using the solution given in the notes. Would the approach I took lose marks/be penalised for not spotting a data adjustment?
Hi Shillington There was a pretty big clue in the wording of parts (iii) and (iv) of this question that the examiners were expecting you to adjust your data or methodology, so I am afraid that I suspect you would have lost some marks on this one. Part (iii) says "....by adjusting the data or methodology...." and part (iv) asks you to "....consider updated diagnostics....based on the adjusted triangles".
I did adjust the methodology though, by not using the chain ladder on everything only on the data which it was appropriate for. I know now for tomorrow I should read "or" as "and".
Following that exam question, could someone enlighten me about the rationale of adjusting the 2009 data in the triangle?
You need to adjust the 2009 data because it is at a different stage of development at the end of each year than the other underwriting years. This is because the company only started writing the business with effect from 1 July 2009. This should also have been apparent from the differences in individual year-to-year factors and the percentages developed to ultimate by year calculated in part (i).
