Hi, A quick question from me about trending individual losses to be representative of the period on which the new rates will apply, and also developing losses to an ultimate position. If we inflate the historic losses, and also apply a development pattern to them, is there not a risk that inflation is being double counted for? If we assume a constant rate of inflation in the past and in the future, then when we develop our losses with our development pattern would this not account for inflation as well (because our link ratios reflect how incurred amounts change from one period to the next, and this will account for inflationary changes in incurred amounts)? Thanks, Joe
Hi Joe This isn't double counting as you are allowing for two different impacts of inflation. Lets say we have data from the 2019 underwriting year at the end of 2020 and we want to use it to calculate premium rates for business written in the 2021 underwriting year. The 2019 underwriting year is unlikely to be fully developed, so we project it to ultimate using some appropriate development pattern. This will implicitly allow for inflation on the amounts that settle after 2020 (as typically the longer they take to settle the greater the impact of inflation). These amounts still relate to the 2019 underwriting year and they help us produce an ultimate claims cost for the 2019 underwriting year. Now to trend the data to make it applicable for the business we are planning to write in the 2021 underwriting year, we need to allow for claims inflation between the 2019 underwriting year and the 2021 underwriting year - on average all of the payments (those we have already made by the end of 2020 and those we will make in the future) will now be two years later in time, so we need to adjust for a further two years' worth of inflation.
Thank you, After I asked the question I started thinking something similar myself, and it's nice to have it confirmed. Regards, Joe
