Solving an Age-Old Problem

When I started in the actuarial profession around 25 years ago pensions were relatively simple. Schemes also enjoyed healthy surpluses and contribution holidays off the back of high investment returns, and numerous new final salary schemes were set up.

How times have changed. The recent Association of Consulting Actuaries survey suggested 87% of schemes were now closed to new entrants and 18% closed to future accrual, with many others actively considering closure. I can’t recall the last time a new final salary scheme was set up.

The reasons for the transformation, including changes to legislation and regulations as well as accounting standards, are well documented. Volatility in investment markets, too, has certainly been a contributing factor for some schemes, although those who banked the surplusses and moved to hedging strategies a few years ago have largely weathered the storms.

A long life

However, one demographic issue that had crept up relatively unnoticed until the past few years has been longevity. A quarter of a century ago, a typical life expectancy might have been set at around 77-79 years of age. Then terms such as ‘cohort’ started appearing in trustees’ vocabulary, then ‘underpin’, then ‘cohort and underpin’. Now typical life expectancy assumptions might be as high 86-88 years of age, even 89, adding maybe 30% or more onto the costs of providing the pension obligation.

It is possible to discover what the mortality experience for a large scheme has been by looking backwards, but it is far from obvious what the level of future improvements will be. The Office of National Statistics recently stated that the number of people in the UK aged over 100 would almost double over the next 10 years. Currently standing at 12,000, they expect this to increase to 22,000 in 10 years and to 280,000 by 2050.

This issue is not going to go away, so what can a company or its trustees do about it? Or are they stuck with having to continually make extra funding available every three years when an actuarial valuation is carried out requiring a further allowance for longevity to be made?

One solution is to transfer the liabilities to a third party, typically through a buyout or buy-in with an insurance company. These involve the transfer of large amounts of assets, so security of the third party is therefore a key issue.

Another solution, primarily for larger schemes, is the concept of longevity swaps, whereby a provider or counterparty is introduced and exchanges cash flows with the trustees.

The counterparty pays a pre-agreed amount to the trustees for each and every pensioner until the pensioner dies. This amount is effectively paid onto the actual pensioners. This is the floating leg as it will vary depending on how pensioners are alive at the time.

In exchange, the provider has estimated what it believes the expected pension payments will be over the next 50 or more years on a month-by-month basis. This amount is fixed and this is what the trustees pay to the counterparty. The trustees have therefore fixed their future payments and neither they nor the company need to worry about the life expectancy of the pensioners. The counterparty now has the risk of pensioners living longer than expected. Some may carry some or all of the risk, but many will offload that risk to other providers such as reinsurers.

This swap operates exactly the same as an inflation or interest rate swap and it is only the difference between the cash flows that actually changes hands.

Gaining ground

Pension schemes seem to adopt a herd mentality, and markets or solutions don’t take off until someone has done something new and proved it is doable. In May 2008, the first such swaps for UK pension funds were announced and hit the headlines in the Financial Times and many other pension and risk magazines. In 2009, there were five such deals, three for Babcock International, one for RSA and one for a local authority. It takes time to transact such swaps, but we are now beginning to see more deals.

For the Babcock schemes, the trustees transacted 50-year swaps with Credit Suisse, covering 14,000 pensioners as well as their dependents with an £800m liability. The company expects to have covered around 98% of the liability by the end of the 50 years and, importantly, still have the £800m of assets covering the liabilities.

The security of the counterparty is, of course, a key issue, but the deal is highly collateralised so the trustees should be able to replace the contract in the market if there were to be a default based on the actual mortality experience of the scheme, trends in mortality of the general population and other issues.

Babcock took this route because it has grown over the past few years in the support sector and has long-term contracts with the Government and other public sector or quasi public sector organisations in a number of diverse fields. As it has grown by acquisition, it has inherited numerous pension schemes en route. The company’s pension liabilities exceed the market capitalisation of the group and, accordingly, some analysts view pensions as a drag on the share price. Interestingly, when the company announced the deal, the share price rose 16.5% overnight.

You may be wondering why it did not do a buy-in/buy-out and also transfer the investment risk. On the investment side it was constrained as it had four schemes, four sets of trustees and four sets of advisers. One scheme had hedged some years ago almost entirely, one had hedged interest rate and inflation risk to a degree and others had not. It was not possible to do a consistent and effective hedge under that governance structure.

It now has changed the governance arrangements and has one set of investment advisers and an investment sub-committee operating across all the schemes to execute the trustees’ investment strategies in a consistent and effective manner, and thus it is now embarking on a consistent hedging policy across the schemes.

Hence, while the focus initially was on longevity, when the investment hedging is added, the company will have created its own DIY or synthetic buy-in.

Will there be other such transactions? Another scheme announced recently is carrying out such a swap and I have little doubt there will be others. The market maybe limited by capacity but there certainly seems to be some interest out there.