No more playing with time?

David Adams examines why longevity risk has only recently become such a hot topic and looks at how the market is responding to the challenges of improving mortality rates.

It's thought that life expectancy in the UK is improving at a rate of two years per decade, which works out at about five hours a day. That seems like excellent news, but trustees and sponsors of defined benefit (DB) pension schemes are only too aware of the downside: longer lives equal increased longevity risks and costs.

Some sponsors and trustees are annoyed that actuaries didn't make more of a fuss about this trend earlier, but it would have been incredibly difficult to foresee the full extent of the cohort effect, the huge increases in life expectancy experienced by the generation born between 1925 and 1940. Long-term life expectancy trends have seen improvements of one to one and a half per cent per year for 150 years, but the cohort generation has experienced improvements of four to five per cent. That only really became apparent during the last 20 years, and even then the facts were obscured by incomplete data.

"It's hard to point the finger of blame, because as soon as the evidence started to emerge the actuaries started telling people about it," says Charles Cowling, managing director at consultancy, Pension Capital Strategies.

And there's another good reason why longevity risk wasn't regarded as such a big deal more than 15 years ago. "When interest rates are very high, it really doesn't matter how long people are living," explains Kevin Wesbroom, UK lead at Hewitt Global Risk Services. "Today, with low interest rates and inflation, it makes a huge difference."

Regulation and legislation passed in the last ten years, as well as the actions of the Regulator, have also helped bring the issue to the fore. But if pension schemes are annoyed by the latter it should also be noted that they haven't always been transparent and realistic about mortality assumptions used to assess longevity risk.

The Regulator has stated that in future it will scrutinise more closely schemes using the 'short' and 'medium' cohorts, which assume that the cohort effect will cease in either 2010 or 2020 respectively, as the basis for those assumptions, than those based on variations of the ‘long’ cohort, which assumes the effect will last until at least 2040.

But are any of the three cohort assumptions really worth using?
After all, each is based on the idea that the factors that have increased the life expectancy of the cohort generation will not be repeated in later generations, something that is far from certain.

Assumptions based on scheme-specific data would almost certainly have a greater chance of being accurate, and would be able to take more account of the other variables that affect mortality, such as geographical and socio-economic factors. Schemes with differing proportions of active, deferred and pensioner members are also influenced by the cohort in different ways.

There is also an argument that although discouraging use of the two shorter cohorts seems rational, it isn't necessarily appropriate for all schemes.
"The Regulator has said that the medium and short cohorts are not appropriate, but a number of actuaries would disagree that using the long cohort is appropriate in all circumstances," warns Paul Marks, technical consultant at Gissings.

"The danger is that you end up where trustees, in order to avoid greater scrutiny, go towards using the long cohort. That undermines the whole idea of a scheme-specific funding regime."

In any case, pressure from external economic forces has persuaded many sponsors that their best course is to offload longevity risks, along with other assets, through bulk buy-out transactions with insurance companies. "The bulk buy-out market has taken off, partly because it takes away the longevity risk but also because sponsors just want this problem off their balance sheets," says Robert Gardner, partner at Redington, the asset liability management specialist firm.

"They find it hard enough just to run their company, let alone having to manage this pension scheme on the side. A lot is driven by sponsor balance sheet volatility, and/or trustees feeling they would rather have an FSA-authorised insurer as a sponsor."
This year has already seen a big rise in the number of buy-outs. "Just talking ballpark figures, the buy-out market for the last five or six years has been worth about £1.5 billion or so; we reckon this year it's going to be about £10 billion," says Cowling.

Alternatively, pension schemes can use financial instruments to trade and hedge longevity risk. Various financial providers, including JP Morgan and Credit Suisse, now offer longevity swaps. These instruments operate in a similar way to interest rate and inflation swaps, with a pension scheme paying a fixed amount to the swap provider in return for payments that continue for as long as a pensioner lives.

The market includes instruments that use an index and those that are scheme-specific. Larger schemes might be drawn to the former, while scheme specific swaps could be useful to smaller schemes that might otherwise prove susceptible to idiosyncratic mortality, where a few individuals with particularly large salaries become a very big drain on a scheme if they live for a long time.

So far, development of this market has been hindered in part by the fact that a long-term swap contract could prove vulnerable to counterparty risk, and also because of another, more fundamental problem.

"With an inflation or interest rate swap it's easy to see how you get two counterparties," says Wesbroom. "With this, I can see why people want to get rid of longevity risk, but who wants to buy the other side of it? Who wants a product that means they have to pay out more if people live longer? You'll always get speculators if there's enough money going around, but you need genuine traders like life insurers. Pharmaceutical companies and retirement homes could be interested, but it peters out a bit after that.

"A lot of trustees are nervous. They think it's only City speculators, and they get concerned that it's not a genuine product, not like a nice old annuity," he adds.

Longevity bonds
Meanwhile, PensionsFirst, a longevity security specialist firm, has developed a series of longevity bonds and derivatives, designed to be held as scheme investment assets. The aim is to allow analysis of risks, then to hedge the exposures that trustees do not wish to manage. This will enable bonds or derivatives to be tailored to the needs of each scheme.
The firm believes its methodology for projecting longevity, based on quantitative and qualitative analysis, will stimulate this market, says Timothy Lyons, founding partner at PensionsFirst. "You have to create a methodology that will allow a sophisticated investor to see the risk they are taking," he says. "The great thing about a statistical approach is that as you move forward it continues to adapt, based on experience. The model is readjusting and reprojecting all the time."

He believes this market has huge potential. "In recent years all investors have found out that their supposedly diversified portfolios were more correlated than they thought," he says.

"What's interesting about longevity is that it's not correlated with other things, it doesn't suffer from those spikes in the economy." It is also just about possible that at a future date the government might be persuaded to play an active role in this market. It will, after all, carry a large amount of longevity risk in the Pension Protection Fund (PPF).

Of course, no-one can ever be certain about what will happen tomorrow. "We're being asked to factor in future changes, but you don't know that the mortality rate is going to continue improving," says Joanne Livingstone, technical director at Punter Southall.

"We are actually way ahead of Europe and America in terms of the degree of longevity improvement we factor into pension schemes. What's changed is peoples' appreciation of how uncertain it is.

"The range of possibilities is far greater than anybody anticipated," she adds.
And at least the number of methods that pension scheme trustees and sponsors can use to remove or hedge longevity risks is increasing – giving them the chance to make positive choices that could help us all during our long, long retirement.

- Pensions Age June 2007