The risk in question
Longevity risk is the danger of good news. Each year that pensioners live longer than a scheme's tables predicted is another year of payments it must find, and across a big fund a small change in life expectancy moves the liabilities by billions. So pension trustees and insurers want to hedge it, to pay a fixed, known amount and hand the uncertainty to someone else. The instrument they use is the longevity swap: the scheme pays a set schedule of payments and receives whatever its members' actual survival turns out to cost, so if the pensioners live longer than expected, the counterparty covers the difference.
It works, but only one deal at a time
The market is real and it is large. It opened in 2008, when a q-forward between J.P. Morgan and the pension firm Lucida in January was followed in July by the first capital-markets longevity swap, in which Canada Life hedged a 500 million pound annuity book, covering some 125,000 annuitants, over a forty-year contract. Since then the deals have grown enormous. The BT Pension Scheme transferred around 16 billion pounds of longevity risk in the largest such transaction yet, routing it through its own captive insurer and on to a reinsurer.
Notice what these are. Each is a bespoke, bilateral contract, negotiated privately between one scheme and one reinsurer, sized and shaped to that scheme's own members. None of them trades. There is no screen, no daily price, no way for a third party to take the other side at lunchtime.
Why the exchange never came
People tried to build one. The q-forward was meant to be the standard, tradable unit of longevity risk, a clean contract on the future mortality rate of a reference population that anyone could buy or sell, and longevity indices were published to price it against. The volume was disappointingly low, and the traded market never formed. Bespoke insurance won, and it won for the reasons this series keeps arriving at.
The first is that longevity is not fungible. A steelworkers' scheme in the north and a barristers' fund in London have different ages, jobs, health and postcodes, and so different futures; one scheme's risk is a poor hedge for another's, and a standard contract leaves both sides carrying a basis risk they can measure only roughly. The second is stranger and more fundamental. A market needs two willing sides, and longevity has only one. Almost every natural holder of the risk, every pension fund and annuity writer, sits on the same side of the trade and wants to buy protection. The only real sellers are reinsurers, who can carry it because they hold the opposite risk on their life books, that people die too soon. A market with a crowd of buyers and a short queue of sellers does not clear on an exchange. It clears in a negotiation.
The point
So the risk that most of the developed world's retirement savings depends on is priced in private, deal by deal, at whatever a handful of reinsurers will quote. The absence of a longevity market is not a failure of financial engineering; it was tried, and the standard contract exists. The absence is information. It tells you the risk is enormous and real, and also that it is too bespoke to standardise and too one-sided to trade, so its price is discovered behind closed doors rather than on a tape.
That should matter to anyone with a pension, because the number being set in those rooms, how long these people will live and what it will cost to keep paying them, is the number that decides whether the promise made to them is actually funded. It is one of the most important prices in finance, and it is one almost nobody outside the room ever sees.
Levanter Markets publishes for education. Nothing here is financial advice.