Maria Dolores Martínez-Miranda and Miranda Thomas report on a recent talk on granular reserving given by Jens Perch Nielsen, where he advised against abandoning our existing pool of practical actuarial knowledge in favour of the technological Wild West
An enormous amount of old methodology dominates one of the single most important numbers in the risk-related financial sector.
Non-life insurance accounts for 5% of UK gross national product and this number, the reserve, is perhaps the single most important number on the balance sheet.
At a recent Cass Consulting event, Jens Perch Nielsen, professor of actuarial science at Cass Business School, argued that the most important thing for future developments in reserving is the practical knowledge that actuaries already have. He began by describing the history of chain ladder methodology. There is a growing belief that this method needs to be re-invented, he said. However, this has produced a "technical Wild West", with every new technological development hailed as the answer.
Nielsen encouraged caution on this front. What if, he suggested, all the tacit knowledge of a hundred-plus years of daily work by thousands of intelligent actuaries has created a breadth of useful knowledge? What if the best approach to developing new methodology is to first understand what we already have?
He argued that actuaries' vast pool of knowledge has not been linked to recent scientific developments in other academic fields. Rather than jumping on the latest technology bandwagon, they need to reformulate what they do so that appropriate scientists can work with it. According to Nielsen, mathematical statisticians have much to offer when tackling the old actuarial reserving problem.
With his academic colleagues, he has set out to rephrase chain ladder methodology so it makes sense to mathematical statisticians. He has achieved this in three co-authored ASTIN Bulletin papers from 2010, 2011 and 2012, which propose a double chain ladder - representing the old chain ladder but in a sophisticated, modern mathematical vocabulary.
Using these papers to rephrase the classical chain ladder technique as a method of mathematical statistics, Nielsen argues that "the sky is the limit" in terms of future developments. He believes the chain ladder as a granular method has enormous consequences for the way the profession reinvents and improves reserving methodology while preparing for Solvency II.
Delegates also heard from Malcolm Cleugh, group reserving actuary at RSA, who explained how the company has been working with Nielsen for nine years to apply this research in a summer internship programme with students from Oxford University.
The interns applied the research to real-life RSA reserving data, to assess the validity of the methodologies developed. In 2012, the research reached the point where meaningful results were generated. The methods can be used to test the validity of prior selections made by reserving actuaries.
Jens Perch Nielsen joined Cass Business School as professor of actuarial science in 2012. He gained his doctorate at UC Berkeley and has worked at various financial companies, including RSA. He contributes regularly to a number of actuarial science and mathematical statistics papers. He is also an entrepreneur, a co-owner and a board member of two Copenhagen based companies. For more examples of Professor Nielsen's work, see www.cassknowledge.com/research/author/jens-nielsen
Report by Maria Dolores Martínez-Miranda, Marie Curie senior research fellow, and Miranda Thomas, communications manager, at Cass Business School