...private. That analysis was set out in detail in a lecture I gave at the Cass Business School in April 2005, which is available on the UK Pensions Commission's website, http://www.pensionscommission.org.uk. In one crucial respect, however, our thinking evolved during last year and the conclusion we set out in our report differed from that suggested in my original lecture. I will highlight that point when we arrive at it.

My lecture last year covered issues relating to investment return risk as well as to longevity risk. It asked questions such as: Do the risks of equity investment decline with the length of the holding period? In what class of assets should pension funds be invested, and is the answer different for defined benefit and defined contribution schemes? But since today's conference is about longevity risk, I will leave aside the investment return risks and concentrate entirely on issues relating to longevity risk, and in particular four:

* How large is the uncertainty about future longevity and thus how great are the risks involved in underwriting longevity risk?;

* How much longevity risk is being absorbed by some or other agent today, and how might demand for longevity risk absorption develop in the future?;

* What are the prospects for risk absorption capacity and price?; and

* Who should bear longevity risk, and with what implications for policy?

First, then, how uncertain are our estimates of future life expectancy? And let me be clear, my focus today is entirely on uncertainty about the average life expectancy of an entire age cohort--not on variability in the individual life expectancies of individuals within each cohort--which is, of course, a significant risk, but one which is clearly statistically analyzable, manageable, and absorbable.

For it is uncertainty over the average life expectancy of entire age cohorts which is the key problem. Figure 1 sets out three projections from the UK Government Actuary's Department (GAD) of how male life expectancy from age sixty-five has developed in the past fifty years and how it might develop for the next fifty. (1)

[FIGURE 1 OMITTED]

The bottom line is the projection produced in 1983, the middle in 1992, and the top in 2003. Projected male life expectancy for 2020, for instance, has increased by five years over the last twenty years.

So the question is, how certain can we be of this latest projection, and what error margins should we assume around this projection?

At least two approaches are possible. One is to listen to alternative expert points of view about future potential medical advances and about whether a biological "limit-to-life" exists. At Cass Business School last year there were two lectures from the optimistic and pessimistic camps, and although neither Jim Vaupel nor Jay Olshansky specifically forecasted UK life expectancies, Figure 2 is a reasonable representation of their philosophies, Vaupel arguing that life expectancy at birth and at sixty-five is likely to continue rising by roughly one year every four years, Olshansky arguing that life expectancy in developed countries will soon level off. But of course these are simply two positions: the fact that they exist gives us no basis for assuming that more extreme positions are impossible. And the fact that these two points of view exist gives us no basis for ascribing to the range their views define a particular confidence level.

[FIGURE 2 OMITTED]

An alternative approach is to try to use past errors or variations in forecasts as a basis for stochastic analysis. Some of you may be familiar with the fan charts of probabilities of future inflation rates which the Bank of England publishes each quarter, looking forward over two years. And the Governor of the Bank of England, Mervyn King, suggested in a lecture to the British Academy last year that the Pensions Commission should develop similar fan charts for life expectancy projections: Figure 3 shows the first shot at that analysis by Bank of England staff. But this poses the question: Is it actually possible to calculate confidence intervals of life expectancy projections in a mathematical fashion? Are we dealing here with mathematically modelable risk or inherent uncertainty?

[FIGURE 3 OMITTED]

To consider that question, the Pensions Commission conducted an analysis of how uncertainty in mortality rate projections drives variability in life expectancy projections. In Britain over the last ten to fifteen years, mortality rate declines among age groups over sixty have been running at over 2 percent per annum (see Figure 4). If they continue at that rate, male life expectancy at sixty-five, currently estimated at nineteen, will reach about thirty by 2050. If the rate accelerates to 3 percent, life expectancy would soar to thirty-seven years. Only if it decelerates to 1 percent would the GAD's 2002-based principal projection of twenty-two years in 2050 be correct. So the GAD 2002...

NOTE: All illustrations and photos have been removed from this article.




Looking for additional articles?
Search our database of over 3 million articles.

Looking for more in-depth information on this industry?
Search our complete database of Industry & Market reports by text, subject, publication name or publication date.

About Goliath
Whether you're looking for sales prospects, competitive information, company analysis or best practices in managing your organization, Goliath can help you meet your business needs.

Our extensive business information databases empower business professionals with both the breadth and depth of credible, authoritative information they need to support their business goals. Whether it be strategic planning, sales prospecting, company research or defining management best practices - Goliath is your leading source for accurate information.