Statistical Predictions for Chain Ladder Data (14rit206)


(University of Alberta)


The Banff International Research Station will host the "Statistical predictions for chain ladder data" workshop from to .

The proposed project concerns the analysis of so-called "chain ladder" data, the triangular form of
the data arising in predicting the reserves that insurance companies are required to keep to comply
with regulatory statutes. While several methods, even implementations in the statistical software
environment R, already exist, we were able to propose a new approach that experimentally outperforms
all of those. As there turned out to be similarities to the method developed for classification with
censored data by Delaigle and Hall (2013), the paper that contains also a good deal of theory, we
believe that we will be able to develop some more rigorous underpinning of our methodology - the
first topic we propose to work on in BIRS.

While all such problematic might be per se rather of interest of specialized actuarial literature
(where practical experience is often valued over theoretical justification), an important aspect is that
in our research on the triangular data, we found that almost isomorphic problems arise in the
biometric theory of the so-called growth curves in Biology. Strengthening this tie, and at the same
time understanding the similarities and differences is to be another objective of our investigation in BIRS.

The Banff International Research Station for Mathematical Innovation and Discovery (BIRS) is a collaborative Canada-US-Mexico venture that provides an environment for creative interaction as well as the exchange of ideas, knowledge, and methods within the Mathematical Sciences, with related disciplines and with industry. The research station is located at The Banff Centre in Alberta and is supported by Canada's Natural Science and Engineering Research Council (NSERC), the U.S. National Science Foundation (NSF), Alberta's Advanced Education and Technology, and Mexico's Consejo Nacional de Ciencia y Tecnología (CONACYT).