Source: Institutional Investor

Longevity Experts Discover Flaws in Current Mortality Data

NEW YORK, NY--(Marketwired - Jan 15, 2014) - Institutional Investor Journals, sponsored by Prudential, recently hosted "Phantoms Never Die: Living with Unreliable Mortality Data." Professors David Blake of the Pensions Institute at Cass Business School and Andrew Cairns of Heriot-Watt University presented a fascinating forensic analysis of mortality data from England and Wales. These data are critical for insurers, and reinsurers, like Prudential, which use mortality projections for different age groups to price annuities and pension buyouts.

The analysis showed there was a puzzling pattern of mortality improvements amongst people in the United Kingdom who are now aged over 90, starting back in 1992. This was due to a combination of effects. First, there was a change in the method used to derive mid-year population estimates based on the 2001 census. Second, in 1919 in the middle of the Spanish flu, the births at the midpoint of the year were significantly lower than the average for that year (because many more people were born in the second half of the year than in the first half).

This meant that following the 2001 census, the number of people born in 1919 was overstated -- introducing "phantoms" into the data. These phantoms served to understate the estimated mortality rate for 82 year olds in 2001 (since the population was overstated). The Office for National Statistics, noticing this effect, "smoothed" the anomalies into the data -- effectively "creating" the phantoms from 1992 to 2002 and then "migrating" them out from 2002 to 2012. "Our key finding is that it is important to take into account the pattern of births each year in order to get a more accurate estimate of that year's population. Any error in estimation will follow that cohort through time," said Professor Blake.

Professor Cairns noted: "Our work has been a reminder that real-world datasets are rarely as accurate as we would like them to be. We have found that identification of anomalies and data-cleansing can only be done through the use of a comprehensive set of graphical diagnostics, identification of signature plots and rigorous statistical modelling."

"Longevity risk is real and needs to move to the top of the agenda for those managing pension fund liabilities. Through this research Professors Blake and Cairns shine a light on one aspect of flawed mortality data," states Allison Adams, Group Publisher of the Institutional Investor Journals.

Amy Kessler, senior vice president, Prudential Retirement, stressed that correcting this anomaly allowed the longevity market to go from "good to great" -- that the vast majority of mortality data and insights derived from those data are sound. Peter Nakada from RMS applauded the Cairns-Blake-Dowd team for "continuous improvement" of mortality data, and suggested that this provides a sounder footing for modeling future mortality trends as well.

To register to obtain a complimentary copy of Pension & Longevity Risk Transfer for Institutional Investors, click here.

For more information about Institutional Investor Journals please contact Erin Scanlon at escanlon@iijournals.com or (212) 2247-3255. To contact Prudential Financial, Inc. please contact Tanya Valle at tanya.valle@prudential.com or 973-802-4370.

About Institutional Investor Journals
Institutional Investor Journals
offer in-depth, original, and practical research in investment management and finance. Written and edited by world-renowned practitioners and academics, the Journals and special topic Guides are extensively read and highly regarded in the industry. Their cutting-edge analysis and strategic insights make them invaluable resources for institutional investors all over the world. Institutional Investor Journals feature articles by investment luminaries and Nobel laureates. For more information, please visit www.iijournals.com.