Improving longevity is clearly a problem on the liabilities side of the balance sheet. Martin Steward looks at it as an opportunity on the assets side, both to generate return and offset risk

When I'm 64, asks the Beatles song, will you still be sending me a Valentine, birthday greetings, bottle of wine? Well, I've looked at the actuarial tables. Birthday greetings and the Valentine I can do. But the bottle of wine? Forget about it. It's not when we're 64 that's the problem - it's when we're 94. With every year that goes by my savings are forced to last another year, too.

It is not news that pension funds are sensitive to improvements in longevity, but very few choose to do anything about it. In some ways this makes sense. The rule of thumb is that a fund with 60% in equities and 40% in bonds will see the same negative effect on its funding level from a five-year increase in life expectancy, a 30% fall in equity markets or a 0.7% shift in either long-term rates or inflation expectations. Clearly, those last two sensitivities are the most acute and tackling them should be the priority.

But of course as you do de-risk, the more prominent longevity becomes. If five years more life is equivalent to a 30% fall in equities for a 60/40 fund, it becomes the equivalent of a 45% crash at 60/40. Moreover, it is dangerous to underestimate the scale of unexpected changes in longevity. The range of probabilities is widening dramatically: even as many anticipate a tailing-off of life expectancy as our sedentary lifestyles and unhealthy diets take their toll, a recent Time magazine cover proclaimed 2045 as the "The Year Man Becomes Immortal" thanks to exponential advances in technology. A 2010 paper by Leslie Mayhew and David Smith, ‘Human survival at older ages and the implications for longevity bond pricing', finds that life expectancy in England and Wales is increasing at a faster rate than ever before, with no evidence of any upper age limit, and concludes that longevity bonds are an effective hedge against this risk.

Long mortality, short longevity
If life expectancy is constantly being revised upwards, in market terms longevity risk is (probably) high and the longevity risk premium, which you get paid if people live longer than expected, is (probably) low. Longevity as an asset class is over-priced and you should short it. Conversely, the mortality risk premium paid for taking the risk that people die sooner than expected is high - and you should go long mortality as an asset class.

This is the strategy of insurance-linked securities (ILS) specialist Securis Investment Partners in the two-thirds of its portfolio dedicated to life risk. Andrea Cavalleri, who sources the fund's life risk, points out that for all the medical arguments that longevity improvements will tail off, statistical analysis of the ‘golden cohort' of UK citizens born in the 1930s and 40s reveals nothing of the sort.

"We ended up favouring mortality risk because we think that life settlements and UK longevity swaps, as examples, are mispriced," he says. "Life settlements are mispriced because they are working from the wrong mortality assumptions. [And] the swaps are priced based on uncertain projections of mortality rate improvements." In other words, Securis won't write a longevity swap for your pension fund - but it will buy one alongside you. "If longevity is mispriced, the protection buyer is getting out at a very good level," as Cavalleri puts it. "It's a great business."

Pension funds are waking up to this long mortality/short longevity opportunity in the context of their hedging programmes. Michalis Ioannides and Mattias Eng at BNP Paribas reckon that about £25bn (€29bn) of longevity liabilities have been transferred, mainly from UK schemes but with Denmark, the Netherlands and Sweden following that lead.

There are various ways of doing this. The most ‘complete' solution is a buyout, which involves transferring the assets and liabilities associated with some or all of a scheme's members to an insurance company or specialist buyout provider. A ‘buy-in' is a similar solution that involves exchanging scheme assets for annuities (which remain a part of the scheme's portfolio). These are expensive, which is why we see many schemes with buyout ambitions trying to ‘tidy-up' liabilities with LDI and other de-risking programmes. But that begs the question: if the scheme is going to immunise so much risk itself anyway, why not try to deal with all of it in-house? That way, the scheme retains the upside potential of its assets - and remember, in terms of life risk, the upside from mortality assets could be considerable.

This is the so-called ‘DIY' solution: LDI to cover rates and inflation risk, plus an indemnity insurance contract, also held as an asset, to cover unexpected longevity risk. An indemnity solution makes a lot of sense if your object is truly to hedge your longevity risk, because it remains the only way to transfer basis risk as part of the package. If you try to hedge the longevity of your scheme members with a swap based on an index of national life tables, there are bound to be discrepancies.

"In extreme scenarios - a cure for cancer, say - the match will be there, but for incremental changes in longevity the basis risk can be quite unpredictable," says Costas Yiasoumi, head of longevity solutions at Swiss Re. Insurers and re-insurers reduce that basis risk for themselves with every new pension scheme client they take on: as their pool of longevity risk grows, the more it mirrors the pool of mortality risk in their life insurance books.

So while there have been several longevity swap deals implemented by pension schemes over the past few years - including the transfer of £3bn (€3.5bn) of liabilities for the BMW (UK) Operations Pension Scheme arranged by Deutsche Bank and Paternoster in February 2010 - it is small wonder that most longevity risk transfers have involved buyout or indemnity solutions. Indeed, there isn't even a cost advantage to taking on the basis (and counterparty) risk of an index swap. Why doesn't a bespoke solution cost more? Because the risk in the off-the-peg solution goes to exactly the same place - the insurers and re-insurers. Capital markets can't find many other counterparties with the same volume of mortality risk for sale. Whichever solution you choose, you are paying a hefty insurance premium - enabling re-insurers to build up the capital to meet their regulatory requirements with regard to the longevity risk you are selling them.

Long NAV, short NPV
Problems with cost seem to derive from the fact that pension schemes have, so far, regarded their long mortality/short longevity positions as hedges. If we take long mortality/short longevity as a pure investment view and are prepared to bet that the upside of general mortality will overwhelm any basis risk against our liabilities' longevity risk, more options for exposure open up and the cost can potentially come down.

As Marcus Mollan, head of strategy in strategic investment and risk management at Legal & General Investment Management, suggests, one reason pension why schemes have been slower to implement longevity as opposed to interest rate hedges is that they are not generally required to mark their longevity risk to market. "At the moment, they only have to crystalise that gap in their funding if they want to do a deal, so it's not surprising that they largely choose not to," he observes.

Moreover, a cashflow hedging solution would be booked by a pension scheme at net present value, taking a horrible snapshot that severely under-values the asset while over-valuing the liability, especially if we believe that mortality is an under-valued asset. And the longer the duration, the worse this gets: actual incurred annual cost of insurance for a 70-year-old with 15 years life expectancy would be roughly the same as that for a 45-year-old with 45 years left to live - but booked as a cost at net present value (NPV) it looks three times bigger. So why not access mortality risk without a hedging structure, as a pure investment risk, and move from an NPV world to an NAV world?

There is no shortage of life insurance-linked securities funds. Most manage either traded life policies (buying policies from the insured and therefore carrying longevity rather than mortality risk) or mass-mortality bonds (whose risk, again, may point in the same direction as pension longevity risk: Yiasoumi notes that Spanish influenza, which killed a lot of young people, actually added years to older people's lives by forcing out the common seasonal virus). But some, like Securis, do work in the trend mortality space, and in 2010 Securis carved out a mortality risk-only portfolio from its diversified ILS strategy for a Japanese pension fund.

"We don't restrict what kind of lives go into the portfolio," says Cavalleri. "That depends on where we find opportunity, and it just happens that the fund has exposure to 70 to 90-year-old US citizens. That's clearly not because they fit the liability profile of the pension fund. The idea is not to have a hedge but a good investment, and if the trend points in the opposite direction as the investor's liabilities then that's all well and good."

Cavalleri concedes that public supply of this trend mortality risk is not huge - and that Solvency II may dis-incentivise insurance companies from securitising their life business. But against this he cites pro-securitisation rules such as the US Regulation XXX, and Securis's own strategy of actively originating deals with counterparties. "That's the logic for paying somebody like Securis to access that risk," he argues. "It's not obvious that you can do this in size if you are only looking at the public market."

Long dentures, short toothpaste
There are longevity-related cash flows available from outside the insurance world (see panel ‘Living off your home: equity release mortgages'), but if our starting assumption is that people will live longer than expected (long mortality/short longevity) and we are prepared to take some risk against our own longevity exposure, then perhaps we should consider this investment idea as a growth opportunity across other parts of our asset portfolios.

Improving longevity and an ageing demographic are not necessarily the same thing, but the trends in mortality and natality in the developed world over the past 70 years certainly mean that one will follow the other for at least the next generation. That leaves a decision to be made about how demographic trends might affect our initial asset allocation decisions (see panel ‘Bonds and the baby boomers'). But there will also be growing revenues to be had from this demographic, for which we might re-position equity portfolios: in the words of an August 2010 Goldman Sachs paper, ‘Demographic Dynamics: A case study for equity investors', ‘as the Baby Boomers live longer, they spend more on healthcare; as they spend more on healthcare, they live longer'.
But this is not as simple as sitting on a healthcare index. There will be winners and losers in this sector, as with other sectors that we might imagine to be well-positioned for these ageing consumers.

A company specialising in drugs to treat childhood diseases, or headache tablets, is unlikely to enjoy particular benefits. So far, so obvious. But even treatments for geriatric conditions are not surefire winners because the overwhelming driver of success in healthcare will be the sector's ability to cut costs for the end buyer - an individual living longer on a smaller pension, or a government trying to look after an ever-growing army of elderly citizens. Even a miracle treatment will struggle to make money if it is too expensive for governments or health insurers. "In pharma you have to be very mindful of re-imbursement policies," says Virginie Maisonneuve, head of multi-regional equities at Schroders.

But while equity managers agree on this, they can draw starkly different conclusions. Maisonneuve sees generic drug manufacturers like Novartis or Teva as beneficiaries because they help to keep prices low. Others struggle to see how increasing competition and diminishing pricing power makes for a compelling investment.

"I would avoid pharmaceuticals if it was about the ageing population," insists Henk Grootveld, manager of the Thematic Growth Fund Rolinco at Robeco. "The pressure is on to reduce costs and pharmaceuticals are the easy target." Christian Schneider, a senior portfolio manager with specific responsibility for the Global Demographics strategy managed by RCM, agrees: "Healthcare companies providing ‘me-too' products, where there will be competitive pressure, may not be positioned well for this cost-cutting environment."

Where Grootveld, Schneider and Maisonneuve do agree is on the importance of diagnostics and preventative medicine. After all, the best way to save money on treating the seriously ill is to stop them getting seriously ill in the first place. The related area of efficiency savings in long-term care of the chronically ill could also prove important. Everyone seems to like Fresenius Medical Care, a leader in kidney dialysis, for example. "The reimbursement patterns in the US are pretty good for dialysis," notes Maisonneuve.

"Governments will be very interested in companies or products that enable faster and cheaper treatment over time, utilising fewer hospital beds or maybe even enabling outsourcing," says Richard Falle, principal fund manager at LV Asset Management (LVAM). "Fresenius is a good example, as is Elekta in Sweden, whose advanced radiation oncology systems are all about innovation aimed at minimising treatment time. These businesses with more specific products are more interesting to us that the more diversified big pharma names."

At Lombard Odier Nicholas Batrell, portfolio manager on its Golden Age fund, specifically designed to exploit this demographic trend, notes that he is really only interested in about 10% of the global healthcare universe. He also picks out the preventative treatment trend. "We think that governments will start backing healthier foods and ‘nutraceuticals' for this reason," he says. The fund's six-strong scientific advisory board, drawn from academia, medicine and industry has, among other things, helped to identify opportunities in this area of food products that confer medical benefits (one member heads the newly-formed Nestlé Health Science Institute). Falle notes that one reason LVAM's European ex-UK Growth fund holds Danone is the nutraceuticals angle.

All of which underlines the prominence of political risk. The exposure to government cost-cutting overshadows everything, but some might also find it difficult to imagine public subsidies being directed to the likes of Danone or Nestlé. Obesity is set to emerge as a major cause of chronic ill-health for the coming generation of retirees, but how likely are governments to subsidise manufacturers of gastric bands?

A similar caveat might be levelled at the idea of getting involved in healthcare and nursing infrastructure. Aviva Investors likes nursing homes and private hospitals enough to make them the focus of its Quercus Healthcare fund and AEGON Asset Management's property division is keen on long-term reversionary leases on healthcare infrastructure assets. At LVAM, Falle picks out Rhön Klinikum to link this story with the overall trend to reward efficiency: "They buy inefficient state-run hospitals in Germany and use their economies of scale and better management techniques to improve them," he enthuses.

But this sector has proven a dud in the past - a number listed in the UK during the 1980s - and some believe that thinking of healthcare in terms of hospitals is looking in the rearview mirror. "Our health infrastructure was designed for a young population in the 1950s - good for treating accidents, not good for palliative care of older people," says Joseph Lu, a motality risk actuary at Legal & General. "Treatment will probably become more community-based."

Grootveld at Robeco agrees. His advice on healthcare infrastructure? "Stay out of it." Hospital stays are expensive, he points out. And one way to take advantage of the more efficient, user-friendly kit turned out by the likes of Fresenius will be to carry out more treatment in the home. "A small-cap like Mediq in the Netherlands is interesting in this regard," he suggests. "A simple supplier of wound-care products, incontinence products and other re-usables to people at home or local nurses - that's a big growth business."

But enough of infirmity - surely there is a brighter side to retirement, too? The generation retiring today enjoys considerable asset wealth, decent pensions, long experience of a globalised leisure and consumer society and good health - the source of their improving longevity. Leisure companies that cater for the frugal student backpacking crowd rather than the cruisers splashing their ‘grey euros' are going to lose out, says the received wisdom.

But we should think about how this leisure story collates with the healthcare story. The person taking their bed bath from the home nurse is clearly not cruising the Mediterranean. Or at least, they might be the same person, but getting from one to the other will take 30 years - a cohort effect that any demographics-led equity strategy will have to negotiate.

Moreover, even if these retirees do spend four weeks on safari in Kenya, that leaves 48 weeks at home, spending more on prosaic things like food, utility bills and, yes, healthcare. For the majority who will not be able to afford the dream holidays this counts double. It's a salutary reminder that investors should not neglect the basics, particularly if they believe that governments will withdraw from centralised healthcare and rely more on community networks and families. "I think that we will see a return to the model where parents look after your kids while you are working, and then you take care of your elderly parents," says Grootveld. "The dream of seeing the world and spending our hard-earned money might be more of a dream than a reality, I suspect."

This has importance for the third sector that is often picked out as a beneficiary of the ageing demographic behind healthcare and leisure: financial services. Providers of retirement products, ranging from insurance companies through asset managers to private banks and IFAs, should be able to capitalise on a clientbase growing not only thanks to ageing, but to corporate-sponsor withdrawal from pensions.

Goldman Sachs estimates that US retirement assets will grow by 40%, to around $20trn (€14.5trn), by 2014, much of it extracting higher retail fees. Annuity providers may be particularly well-positioned should an older population demand higher interest rates, boosting asset values against liabilities. "That's all on top of the growth we expect when people in countries where there isn't a decent pension system - France, Italy - find out they want to retire in five years' time but can't," says Grootveld. "Particularly as we see a low growth environment for banking, I would definitely focus on these areas."

Long ageing risk, short sector risk
One reason leisure, private healthcare and financial services come up so often in demographics-led equity portfolios is probably because these funds go, understandably, where the margins are. They reflect how to make profit from the older generation, focusing on revenue from the wealthier people, rather than the older generation as a whole - let alone the particular profile of any pension scheme. This presents a problem for any investor thinking they might offset some of their liabilities' longevity risk. But it's also somewhat strange to see these sector biases: fund managers' disagreements about how to play specific sectors suggests that the ageing demographic risk premium in equities probably isn't a sector risk.

We can see this if we consider how the increasing dependency ratio will sweep its way through the economy. Some argue that a reduced supply of workers will result in redistribution of returns back away from capital and towards labour. In this environment, such economic growth and return on equity as there is will come from those enterprises best able to boost productivity by, for example, specialising in high-technology manufacturing or services. This will become a particular point of advantage now that the productivity gains available from exporting less scaleable labour (via offshoring or immigration) appear to have run their course - "Chinese labour is three times as expensive as it was 10 years ago," as Falle at LVAM observes.

If higher productivity (figure 4) is one solution to the increasing dependency ratio, longer working lives is another. To some extent this will evolve naturally - as we live longer we will simply be less able to afford early retirement, and we may be tempted to stay on if it is true that labour costs are on the way up. But there will also be a concrete advantage for companies able to use more mature workers; more older people than young may result in a cheaper price for each comparable unit of experience, for example. Schneider at RCM notes that BMW has an experimental plant in Germany dedicated to figuring out how best to configure car manufacturing for an ageing workforce - which gives first-mover advantage to BMW, of course (in a sector whose revenues are not associated with the over-65s), but also opportunity for a range of firms to develop technologies for that plant. As Maisonneuve at Schroders puts it: "There are big advantages to maintaining these older workers, so do we need better lighting, higher chairs?"

Counter-intuitively, then, the hi-tech firm that can maximise productivity from it grey-haired workforce while selling high-margin products to young consumers might be a better bet than the provider of careworkers from the Philippines to nursing homes. The nurses demand ever-higher pay, but there is a limit to the number of patients they can see in one day. Optimising for the ageing demographic risk premium, in other words, should be achieved not only within sectors but also across sectors, regardless of whether or not revenues ultimately come from that demographic. Indeed, this becomes obvious as one understands that the ageing demographic can be the source of revenue, of capital, of labour, or of all three, for any business in any industry.

For this reason, the pension scheme that wants to maintain sector weights close to its benchmark's might prefer some kind of ageing demographic overlay rather than one of the available demographics-led thematic strategies. Despite their managers' claims of diversification they tend to take predictable sector bets. Lombard Odier's Golden Age fund is not atypical, with 45% in healthcare, 35% in pharmaceuticals, biotech and life sciences and 20% in financials and consumer stocks. Schroders' Demographic Opportunities portfolio has big overweights in healthcare and consumer stocks, and its biggest underweight is IT, an attractive sector in its own right at the moment, but also a key potential beneficiary of the productivity drive.

Long time horizon - long ageing risk?

But even if we think we can isolate the ageing demographic risk from sector risk, if we try to get it via equities we certainly expose ourselves to all sorts of other risks - chiefly equity risk, of course. "Let's say you set off on this strategy in 1999," says Antony Barker, managing director with JLT's Pension Capital Strategies. "You get the 2001 census, new life tables, smoking banned in English pubs and Parisian cafés - but you make a loss in your equity portfolio regardless of the revenues it's tilted towards."

As BlackRock's head of strategic advice, Robert Hayes, observes, we know what equity or property beta is, but defining ageing and making it a part of a principal components analysis is a long way off. "The one area where our analysts did find some correlation [to ageing population growth] was marinas - the berth holders tend to be pensioners," he says. "But we could easily find that, say, taxation going up to fund growing pension liabilities, reveals that those marinas were just massively exposed to top-end discretionary spending and perform terribly."

Indeed, this is why, despite passionately believing in the long mortality/short longevity position, Cavalleri at Securis buys Swiss Re's life books rather than its equity. Still, ultimately that equity must be a claim on Swiss Re's revenues - why else would anyone buy them? It might take a long time for the long-mortality risk premium to come through in the price, but it must happen eventually. That should be no more controversial than to claim that the equity risk premium will be paid, eventually. If our starting assumption is correct and the mortality risk premium is under-valued, we should be paid that premium over and above any (positive or negative) equity risk premium we receive.

But the key word is ‘eventually': over what time horizon will this premium come to us? This is undeniably a growth opportunity, and like all growth the risk is that it appears in the price all at once thanks to an onrush of capital. This will determine the size of the ageing demographic risk premium at any point in time - and, for any kind of offsetting of liability-related risk the more its delivery matches the pace of change in unexpected longevity improvements, the better. In this context, this particular form of growth investment might, indeed, have some appealing characteristics.

"When you start to see all sorts of company reports from the Street's analysts talking about the impact of demographics then you know that a large part is going to be in the price - but we are not there yet," notes Maisonneuve. "And in any case, it's a path: you don't get to 2050 with nine billion people on the planet overnight."

This is a point made by a number of thematic equity managers. This growth should "come through in stages rather than in bubbles", suggests Frances Hudson, global thematic strategist at Standard Life Investments, "because demographics change very slowly". Schneider at RCM notes that analysts outside those covering life insurance are not focused on this theme because they simply cannot take such a long-term view. "For that reason I hope to see fairly regular, constant upside surprises for these companies, and don't expect to get this growth in great periodic bubbles," he says. Batrell at Lombard Odier says that the Golden Age fund's companies do not show higher valuation multiples than comparable sector peers. "We don't pay any particular premium for this potential growth," he insists. "We are confident in the potential earnings growth of our portfolio because we like the top-line drivers - and we expect relative re-rating and de-rating of companies around this theme. If the market re-rates tomorrow we will still benefit from the superior growth of our companies, but we would like to see this play out over time."

This takes us back to our starting assumption once again. If life expectancy continues to be revised upwards, the longevity risk premium will continue to be low and the mortality risk premium high. If this is true in the pure life risk market, it will be true in the equity market too, even when its analysts begin seriously to price in demographic risks. How could it be any other way, if the market's only sources on longevity assumptions are the experts who continually undershoot? The long mortality/short longevity spread might be arbitraged away, but only until the next tables appear to open it up again: this grand periodical re-rating of life risk could only stopped by the market taking a wholesale decision to bet hard against the actuarial profession.

Of course, in the short term equity cash flows will not look like longevity-hedge cash flows. But over the long term the risks ought to converge, and there may be reason to believe re-rating of stocks should come as upside surprises that exhibit some correlation with periodic revisions to longevity projections - assuming that the demographic risk premium can be identified and optimised. That is a big assumption, but one worth investigating, certainly as a potential source of superior returns - and possibly as a source of risk reduction.