I was trying to build out the example in the paper to understand how it works better (https://www.casact.org/sites/default/files/2021-07/Munich-Chain-Ladder-Quarg-Mack.pdf), but I’m spending more time than I want trying to follow the notation and correctly replicating some of the pieces.
Does anyone have a working example of MCL in Excel that they would be willing to share?
If you haven’t got an answer by the 2nd, DM me on here and I should be able to clear data out of my old template and send it to you. I’m travelling and don’t have access to my work drive right now.
Thanks, Mountainhawk. If you can share, I’d appreciate it.
As a side note, Copilot has been really helpful the past month with random questions and coding issues, but it totally failed on this one. While it did eventually point me to Munich Chain Ladder as a method that incorporated both paid and incurred data, it was convinced that the calculation was the average of the paid and incurred chain ladder estimates.
OK, I found it and was able to make a template with no data in it, but since it’s on my work PC I can’t upload here – if you can give me an email I’ll send it over. I did make one minor modification to the method where I cap the sigma/rho ratio to 10, because I found it made the method a little more stable.