Longevity risk debate: How pension schemes cope with an ageing population

Delegate biographies: Learn more about the panelists

PK, Watson Wyatt Let’s start with what we have observed in the past. Why has life expectancy been increasing?

RW, Paternoster A decrease in circulatory diseases – heart disease and stroke – have contributed to about two-thirds of the total improvement. The most significant reason for this is that far fewer people now smoke. But there are also a number of other factors: average population blood pressure and cholesterol have fallen, treatments for people who actually have heart disease are completely different to what they were 10 or 20 years ago, with bypass surgeries and angioplasties. More recently, we’ve also seen improvements in other causes of death. Improvements in survival rates for cancer have helped to drive mortality improvements.

PK, Watson Wyatt And in terms of the so-called cohort effect in the UK – the faster mortality improvements for people born between around 1925 and 1945. What are the drivers in relation to that?

RW, Paternoster I think most people would argue that the changes in smoking have been the most significant factor behind the cohort trends. The generation reaching adulthood after World War II were more aware of the negative effects of smoking, whereas the previous generation were the ones that were given free cigarettes as members of the armed forces.

PK, Watson Wyatt And that throws up an interesting question, of course. Because, although as anyone who’s ever tried to give up smoking will know, you may need to give up more than once before you finally succeed, your health and therefore how long you live can only ever benefit from giving up smoking once. This raises the question of whether these improvements will continue?

RW, Paternoster Yes, absolutely, the drivers of future mortality improvement are likely to be quite different to the drivers of past improvement.

AR, Credit Suisse It may be a bit too early to measure the effect of smoking bans in public places. That could be a driver of future improvements in years to come.

JL, Legal & General I think smoking is one of the many reasons. There are many other potential reasons. ranging from changes in technology, behaviour, lifestyle, working conditions, healthcare and medical advances. I was told that in the 1950s only 2% of the population had central heating, but today it’s 98%. When people get cold they increase their chance of getting high blood pressure. This could contribute to strokes or heart attacks. And there are other things like changes of nutritional uptake over time. For example, with the availability of refrigerators, people don’t have to rely on salted meat. There’s a professor who keeps telling me that it’s not about smoking. He’s an expert in Japanese population. He said that the Japanese still smoke like chimneys but they have experienced a large reduction in mortality. They have got cohort effect (just like the UK) and are one of the longest-living peoples in the world. There are many other factors that have contributed to the increase in life expectancy but they have not been studied extensively.

DR, Watson Wyatt Although the "golden cohort" generation has benefited significantly compared with previous generations, the following generation shows continuing mortality improvements. It implies that each development, whether relating to risk factors or treatment, is consolidated, and changes the bar of people’s expectations. The current emphasis is on cancer treatment and how that can be improved. Among the main drivers for cardiovascular mortality improvements were relatively simple treatments, more so than complex operations. It is not clear whether treatments for other causes of death will also be widely available or whether the associated cost will restrict their availability. If the latter, this could lead to a widening gap of expectations between the general population and those in pension schemes or with insurance policies.

PK, Watson Wyatt  When we talk about things like obesity, the question is: will that actually impact on the fundamentally closed group of people who are in defined benefit pension schemes?

AR, Credit Suisse Indeed. Studies show that the highest incidences of obesity are in the lowest socio-economic groups and the highest membership of final salary pension schemes is in the higher socio-economic groups.

PK, Watson Wyatt Joseph, you mentioned some of the things that are happening internationally in terms of mortality.

JL, Legal & General I’ve looked at the changes in death according to some of the major causes of death (circulatory diseases, cancers and "other" diseases) of seven developed countries. I observe that there has been a large fall in deaths relating to circulatory diseases over the past 20 years in all the countries under study – UK, Japan, US, Germany, Sweden, France and Canada. There has not been much fall in cancer deaths. So I think the phenomenon that we see is something that can be seen worldwide.

However, there are also different trends in various countries. In Canada, for example, the improvement there has slowed down because, as circulatory deaths reduce, they exhausted that cause. In Japan, the death rates are still falling. That leads to the question: Is there a limit to life expectancy? I think there is a limit, because there are some insects that live for a few days, mice live for three years, dogs 12, 15 years. There’s a limit to this improvement, in life expectancy, unless some major intervention in our species occurs.

RW, Paternoster If there is a limit, it’s quite a lot higher than the average life expectancy for a defined benefit scheme member at the moment, because a few people do survive to 110 or 115. An age of 122 is the record.

JL, Legal & General  All our cells get replaced and there’s a large turnover over time, so there’s not many parts of us that are left from the day we were born. And this process of turnover, every time there’s turnover there’s a chance of making errors and mistakes, and this error accumulates into the future. There comes a time when this rate of accumulation of error exceeds the ability of the body to repair. That is ageing. No matter how much money we have, no matter how much we give up smoking or what medicine we pump into ourselves, there’s this process that we have to fight. So there is a limit.

DR, Watson Wyatt We should not forget that the most important part of us is almost immortal. Our DNA uses cells in our body to pass itself down from generation to generation almost unchanged. The innate ability for cells to live longer by repairing any damage is not promoted beyond sexual maturity. Once your cells start to decay, you start to be more susceptible to disease, and this increases the likelihood of your dying. If that repair capacity could be upgraded, the ageing process could be slowed, with significant benefits to mortality. The question is whether this can in any way be controlled properly.

AR, Credit Suisse There is the reservatrol suggestion: that this substance found in wines from Sicily and southwestern France, can extend life quite considerably. Unfortunately you would need to drink about 750 bottles a day to be able to do that, but you can condense it and take it in tablets.

DR, Watson Wyatt We are moving from a situation where long-standing treatments are applied on a strong evidence base to a wider population to the possibility of treatments being tailored to the precise medical problem. That leads to problems for the pharmaceutical industry in funding and developing these treatments and for society in being able to pay for such treatments. The relative shifts in the impact of different diseases will depend on where society is prepared to spend to achieve improvements, and that will be driven by how much publicity particular diseases will receive and whether developments can be turned into clinical practice.

The problem

PK, Watson Wyatt Only one in every 200 people born in 1908 lived to see their 100th birthday. By contrast, the Office for National Statistics projects that boys born in 2008 have a 24% chance of becoming centenarians, while girls have a 30% chance. Improvements in life expectancy are good news for the people who live longer but a concern for those who underwrite longevity risk. What should pension schemes actually do about it?

GS, KPMG Well, the first step is to quantify the problem, most often by comparing the scheme members with the broader actuarial studies. Traditionally, schemes have done this every three years as they come through to their actuarial valuations, but we’re increasingly seeing corporates particularly taking an active interest in mortality assumptions, conducting and sanctioning their own investigation work.

NP, Alliance Boots Yes, there is a sea change here. Historically the three-year actuarial valuation has been the trigger for looking at mortality, but that’s changing. For example, we are now thinking about this much more actively and although we are a relatively big pension scheme, with £3.5 billion-worth of assets, this activity will feed down into the smaller schemes. The reason for the change is simple: the huge cost of providing these benefits. Anything we can do to analyse the drivers and perhaps find ways to reduce the cost is now a board issue. So we are seeing a big change in the way that large corporates look at mortality.

PK, Watson Wyatt Companies can look at these risks holistically. Think of a large supermarket selling the proverbial basket of goods: it has a natural inflation hedge to take into account when looking at pensions. Neil: Boots sells products which people will use more of if they live longer. Does that mean that improved life expectancy will increase the money you take at the till as well as your pension costs?

NP, Alliance Boots Yes, there is a business hedge, but we’ve got to look at trends as well; both trends in the business and trends in mortality, and how they interact to get to the right solutions. If you look back to the early part of this decade when Boots hedged a lot of its inflation and interest rate risk, that was almost unheard of at the time, but looking at these risks now is standard practice. I think mortality risk will be looked at as part of standard practice in future.

PK, Watson Wyatt  How up to date are the mortality assumptions used in recent pension scheme valuations? Are there still more cost increases to come through?

AR, Credit SuisseA recent Watson Wyatt survey showed that, typically, over the most recent period of three years in between actuarial valuations, schemes had allowed for life expectancy to increase by 1.5 to two years, which is, say, adding 5% on to the liabilities purely as a result of longevity. So the trustees are certainly reacting. However, since no one really knows what the correct longevity assumptions are, it’s hard to say definitively that valuations are spot on.

RW, Paternoster One of the first things we do when we get a request for a buy-out quotation is to look at the mortality assumptions being used, and compare those with what we would think would be a prudent assumption for that scheme, given the characteristics of the individual pensioners within it. We’ve certainly seen a very big change over the past 12 or 18 months. Eighteen months ago, on average, we found that pension schemes would typically assume a value of liabilities about 5% less than we thought was the best estimate based on our view of mortality improvements. It was very rare to see a scheme that was using more prudent assumptions, either for accounting or for valuation purposes, than our best estimate. Now we’re seeing schemes being more prudent than that. But there still is a huge variation.

GS, KPMG Yes, because of the three-yearly valuation cycle, surveys will inevitably show an out-of-date position. As each new development comes out, it takes time to be fed through that cycle.

DR, Watson Wyatt A number of pension trustees would argue that there are fundamental differences in the characteristics of their particular pension scheme that affect their current mortality experience and might affect their future mortality experience. My point is that there are potentially many different reasonable and supportable views. We should consider a range of scenarios that are then presented to pension trustees or sponsoring companies to illustrate the likely range of life expectancy.

Capital markets solutions

PK, Watson Wyatt  Clearly the most interesting things for the people with the problem are potential solutions. And there is a wide spectrum from the full buy-out – transferring the entire pension scheme to an insurance company, removing all of the risk, inflation, investment and mortality – to products that look just to move that mortality risk while retaining the investment and other risks within the pension scheme. Andrew, give us some of the details, and the pros and cons of those possible solutions?

AR, Credit Suisse Sure. Let’s take the capital markets solutions. These fall into two groups – funded and unfunded. Funded means the pension scheme passes over assets on day one, unfunded means it would pay over the duration of the contract instead.

The main unfunded solution is a longevity swap written as a derivative contract. The principle is that for a portfolio of pensioners it’s a standard cashflow swap, an exchange of payments, so each month the scheme would pay to the provider a payment, and each month the provider would make a payment to the scheme. The fixed leg – that is, what the scheme pays to the swap provider – is set at the outset, so the scheme knows its schedule of payments and would effectively be immunized from increases in longevity. The floating leg, which the provider pays to the scheme, would be constructed to mimic the cashflows that the scheme is obliged to make.

Now, there are features that differ between providers. The first is the term of the swap. Some providers will provide a short-term swap, say 10 years, others will provide longer, 30, 40, 50, in a few cases unlimited term. Another feature is what you get back from the provider. Do you get back payments that are scheme-specific – individual life, tailored to your scheme’s population, so that you get a hedge according to your scheme? Or do you get one back, an index swap, as they’re called, based on the population as a whole?

We’ve had traction with longevity swaps for schemes which for one reason or another think insurance is too expensive at the moment, and/or they want to keep the assets themselves in order to benefit from any upside and to manage the other risks separately. A scheme can embed an inflation swap in a longevity swap, for example, or choose to retain inflation or hedge it out elsewhere.

And the final feature that differs is how the swap is secured. Swaps are collateralized, so if it turned out that there were net payments being made to the client, then collateral would be posted by the provider. If something did happen to the provider, then that collateral would be there to provide security. Commonly it would be two-way collateral, so the provider would be protected similarly. That contrasts with insurance-based solutions, which would typically be covered under the Financial Services Compensation Scheme.

Insurance solutions

PK, Watson Wyatt Yes, let’s look at insurance – the full buy-out solution, the complete transfer of risk to insurers.

RW, Paternoster Solutions like those Andrew outlined can also be structured within an insurance environment. There’s a market for mortality-only products based on the actual scheme’s mortality and marital status experiences as well, rather than generally being based on population indices. More commonly within the insurance sector is a request for a quotation for a full buyout of all risks: mortality, investment, administration and so on. The scheme is essentially paying a single premium up front and getting rid of all the risks that it runs in relation to the whole scheme. Schemes can also partially de-risk – this is sometimes called a "buy-in".

AR, Credit Suisse And there are other types of product that have not attracted so much attention. For instance, you might be able to get credit insurance or a variant of credit insurance on the employer, or range insurance on longevity. To explain range insurance, consider an example of a portfolio currently with 150 lives. Say if you have between 100 and 120 lives existing in 10 years’ time then you don’t get a pay-out, but if you have between 120 and 130 then you do get a pay-out, and if you have over 130, you get a pay-out but it’s capped at the pay-out you’d get if 130 were alive. That could be coupled, say, with performance of an asset, for example, equity performance. If the equity market underperforms compared with a certain assumption and people are living longer than expected, then you get a pay-out. But if just one of them happens, then you don’t. It might be possible to combine that with some selling of the upside to reduce the cost of the product, so if equities outperform then you don’t get the full benefit of it.

PK, Watson Wyatt  So one problem for pension schemes – aside from determining how much risk they are running – is how to choose from these different products?

Funded versus unfunded

AR, Credit Suisse Clients start from one of two basic points: one is the funded "let’s get rid of the problem", and that’s typically an insured route. The other is "well, ideally perhaps we’d like to do that, but we’re not going to be able to do it for a while, or we want to keep the assets ourselves and run the scheme ourselves", in which case the unfunded solution, the longevity swap, probably with inflation and interest rate hedging as well, is the first thing they look at.

One interesting possible development is funded "liability-driven investment" or LDI. Several asset managers offer bucketed interest and inflation cover, allowing one to buy a specific cashflow in 10, 15, 20 years’ time. Well if you could get some longevity cover in there as well, that would be quite interesting. Now, if it’s done on a pooled basis, there would probably need to be an index swap, so it’s not specific to the scheme, but for bigger schemes with their own bespoke offerings, you could add longevity. So it will be very interesting to see whether asset managers other than the traditional insurers get into the game of providing longevity cover too.

For me a big issue is the choice between funded and unfunded solutions – do the scheme and sponsor value retaining assets and benefiting from the upside, or do they not want to retain investments.

RW, Paternoster Exactly. If there’s a very low appetite for risk in the investment side of things, then I would argue that the scheme might be better off with a full package solution.

GS, KPMG I think that’s true. I think most sponsors will still see investment risk as the biggest risk, and not all have the appetite for the investment risk hedging that others do. So a key driver of product choice is employer philosophy.

PK, Watson Wyatt  Indeed, and we see very different philosophies across the pension schemes that we work with. We work with many pension schemes, even those that are closed to new hires, that still have a very high equity content, while others are much more conservative in their approach. And, because of the way actuarial valuations are carried out and transition costs, etc, asset choice affects the price one would pay to transfer all the risk versus running it as is.

JL, Legal & GeneralThere’s another product I would like to touch on, which is buy-in. This is slightly different from buy-out, in terms of contractual commitment.

For buy-out, the contract is between the insurance company and individuals. Here, the insurer will pay the individuals for life in return for a premium paid by the trustees.

For buy-in, the contract is between the insurer and the trustees. The trustees will pay the insurer a premium in return for annuities for the members. So to the trustees, the buy-in is essentially an investment product that matches all the risks involved in annuities, such as the investment risk, expenses and mortality risk.

Furthermore, there’s a difference in terms of security that the individuals get between these two settings. For buy-out, the individual gets the expected annuity, and on top of that there’s a security implemented by the FSA, which is a 1-in-200 years event. In addition an insurance company which is double-A rated, such as Legal & General, will be required to stress test a 1-in-400 years event risk for the company. So the individual is getting all this protection.

While they’re in buy-in, in addition to all the above, the members would have the right to recourse of funds from the employers as well. If the 1-in-400 years event happened to the insurance company, the members can still go back to their employers to get more money. And in the unlikely event that the employer goes insolvent, they have Pension Protection Fund (PPF) to fall back on. So there’s a difference in the level of security in these two different products which have been useful to our clients.

PK, Watson WyattDo these products tick the boxes of what corporates would like to do in terms of risk reduction and risk transfer?

NP, Alliance Boots  Well, it’s an evolving marketplace so corporates are watching and seeing how it develops, rather than jumping in. But certainly risk management in its widest sense is becoming much more prevalent in large corporates, and will continue to be so.

AR, Credit Suisse I think clients are still deciding what their true objectives are. Many trustees, for example, in their scheme-specific funding plans, have a primary target of getting to their technical reserves in five or 10 years, but then also a secondary objective of getting to self-sufficiency, which may be less than buy-out, or buy-out, in 15 or 20 years. And they need the banks or insurers to help with risk management and consultants for the more strategic advice on objectives in getting to that self-sufficiency target.

NP, Alliance Boots  There’s always a trade-off in this situation. You’ve described the objective, Andrew, as being trying to get to self-sufficiency and hence to buy-out in maybe 15 or 20 years’ time, and I think that’s true, and I think the corporate and the trustee would possibly have the same objective. I think what differs between the two camps is the timing, not the cost – it’s about the pace of funding that cost. And so that’s where you get the potential friction.

GS, KPMG Most pension schemes don’t want mortality risk, full stop and although various people are trying to come up with all sorts of different ideas about slicing and dicing things, most corporates just want the total mortality solution.

NP, Alliance Boots  That’s right. From the corporate perspective the ability to be able to take out mortality completely is ideal.

Unwind/flexibility

PK, Watson Wyatt Can these products cope with changing circumstances in the pension scheme?

AR, Credit Suisse Can you unwind the swap even though there isn’t a public, liquid market yet? There are three ways of doing that. If the scheme purchases an annuity, the swap could be novated to the annuity provider, transferring the obligations of the pension scheme to the insurer. We would expect that to be the standard route. Second, we would offer unwinding terms on the swaps going forward, but that would be on our terms. Third, we could embed some options to unwind on guaranteed terms.

RW, Paternoster If you have a mortality swap with an insurance company you can have an option to essentially commute the payments for the swap into a full buy-out. On a daily basis you can work out how much you’d have to pay to switch the swap to a buy-out or transfer to another insurance company.

PK, Watson Wyatt  What if there was a regulatory change that required the scheme to exit the arrangement or change it? For example, the recent civil partnership legislation changed some schemes’ exposure to mortality risk. Could they have reopened a swap agreement if they had one in place?

AR, Credit Suisse That just comes down to what is covered in the swap contract. Generally, if we can price it we can discuss it.

RW, Paternoster On the EMAP transaction we received a premium to take the data risk as part of the overall transaction.

Lack of capacity

PK, Watson Wyatt One issue is the potential size of the problem. Aren’t there more people wanting to get rid of mortality risk than willing to take it on? Is there enough capacity to transfer all this risk?

AR, Credit Suisse Market capacity generally is a big point. People bandy round figures like £50 billion-worth of capacity within the insurance market, and compare that with about £1 trillion in UK private sector occupational pension schemes. No matter what you add from the banks, there’s still a big gap.

PK, Watson Wyatt So, to what extent will that gap be filled?

GS, KPMG That comes down to the question of, outside the fully insured solution, who wants to go long of longevity. Yes, risk-takers, to some extent, but I think most people would agree that the non-insured market will remain just a small niche.

AR, Credit Suisse Well I can give you our figures on that. Our business grew from life settlements in the US. There we targeted the high-net-worth individuals who have life insurance – 70% of the population does there – and who no longer need it. The settlement we were able to offer them was attractive compared with what an insurance company surrender value might be. We would effectively take on the policy, still in their name, and that makes us long on longevity. We’ve bought hundreds of these policies, up to several million dollars each in face value. We then intermediate – we package up the risks, change them into a format that investors – typically insurance-linked securities funds and hedge funds – want, and sell those that we can. And it has worked. We’ve placed moe than $7 billion-worth of longevity risk with investors in that form. Why do they do it? They see it as an uncorrelated asset class, one that has been not volatile compared with others and has offered attractive returns. And some of these investors are crying out for UK risk now. Now I accept that that might not solve £1 trillion-worth of problems but the challenge now is to do a first longevity-only trade with a pension scheme and then take it from there.

PK, Watson Wyatt  Are there any lessons to learn from the insurance space here? There have been a number of transactions where insurers have reinsured mortality risk via these sorts of derivative contracts. Are these indicative of what is being offered to pension schemes?

RW, Paternoster Yes, there is growth here. I think it’s almost inevitable that the first trades in the mortality-only space were between insurance companies and reinsurers, because both parties in the transaction felt that they knew the risks and rewards intimately. In the future, it will be investors, who have not traditionally been experts in this field, but are attracted by the diversification and the potential returns that they can earn.

One thing that’s certainly changed a lot over the past five years is that the gap has closed between insurers’ and pension schemes’ projections of future mortality improvements. I think we’re also getting to the stage where we’ll be able to convince external investors that the bases that we’re using to project future mortality improvements are reasonable ones on which to take risk.

Market barriers

RW, Paternoster It’s quite difficult to get rid of all the mortality risk for a whole scheme with actives and deferreds, as you don’t know quite what age they’re going to retire, or whether they’re going to transfer out before they reach retirement. That makes the structure of the contract much more complicated than for a current pensioner-only mortality swap.

GS, KPMG Even pensioner-only isn’t necessarily completely straightforward, depending what detail you get down to.

AR, Credit Suisse The emphasis does seem to be on pensioner solutions, whether it’s buy-in, which seems, over this year, to have been the more popular insured solution, or on the derivative side.

NP, Alliance Boots Price is clearly a potential barrier. Both sides have to agree on the risks and the assumptions and so on. I think the other barrier, as I see it, is a reluctance to dip a toe into the market, because it’s evolving. Participants in this market are not noted for their desire to be at the cutting edge. "Brave" is not necessarily a word they want to be using.

PK, Watson Wyatt  But someone had to be the first to do a buy-in and someone had to be the first to do a buy-out. We’ve already seen billions of pounds-worth of transactions in those markets, with much more predicted to come. Couldn’t we see a similar snowball effect with mortality products?

NP, Alliance Boots True but in this case you – the consultants – have to get comfortable with what are very complex new products, and it will take time.

DR, Watson Wyatt  Trustees could consider offering only a slice of the pensioners, rather than the whole scheme. This would lead to a well-diversified asset for the pension scheme, and provide limited exposure over a number of years before deciding whether to seek a complete solution.

NP, Alliance Boots  Absolutely right. I think I mentioned earlier the fact that there may not be the appetite or the ability to buy out your risk entirely in one go, and therefore buying it in slices and a proportion is one way of slicing that cake has got to be a viable option.

JL, Legal & General And the employees still have the employer as the key point of contact.

RW, Paternoster We’re certainly seeing a lot more quotations that are split into tranches: someone will ask for your price for males over age 75 and females under 75 and so on. At different times there may be particular tranches that are more attractive in terms of pricing relative to other ones.

PK, Watson Wyatt  At what sort of size are people doing that? Is it only the billion pound and bigger pension schemes?

RW, Paternoster Not at all. It can be small schemes as well, and the definition of "small" in the buy-out world has changed very radically over the past year – we’re now talking about schemes of £100 million to £200 million.

AR, Credit Suisse My take on the buy-out providers is that their target scheme is increasing in size, so if you’re sitting on a £20 million or £50 million scheme you might not have as much choice as you did a year ago.

GS, KPMG Now that providers can see how much business there is, yes, they are now concentrating on the larger schemes. It’s just market dynamics.

PK, Watson Wyatt I think one of the things we’ve seen is not just the market overall moving in terms of its size, but it’s actually different players focusing on different areas of the market, so we are seeing people who are focusing more on the very small schemes, those focusing on the larger ones, those who focus, because of the particular design of their product, on members not yet retired, and there are those who are much more focused or even to the point of drawing a line and saying "we only want the pensioner members". And in fact some of the added value is understanding who and how to split the pension scheme and who are the right people to take those different tranches of business, accepting that as soon as you split up a book of lives you obviously then need to think again about how those things interact and what are the pros and cons of doing that, as opposed to placing it all with one provider.

JL, Legal & General And there are times it will appear cheaper to do a buy-out. For example, if the stock market is high and the yield is high at the same time. The pension fund would be well funded and the buy-out price would be cheaper, because the price is driven by corporate bonds.

PK, Watson Wyatt  Regulation might drive timing and price. If a mortality trigger accounting standard pushes your accounting and funding position to such a point that it’s much closer to a buy-out level, then would you just bite the bullet?

NP, Alliance Boots We’d have to see. We’ve done some analysis on its effect on our balance sheet, and it’s enormous. It would mean an awful lot of companies would opt for buy-out.

Future possibilities

PK, Watson Wyatt  To round off, can I ask for some predictions of how the market will develop either in the next year or on the next five years?

JL, Legal & General I have seen larger pension funds coming to the markets and using more buy-in/buy-out. I expect potential use of other hedging instruments such as mortality swaps. What I’d like to see is a liquid mortality market as well to back this up, to transfer the risk to the market.

NP, Alliance Boots  I would confidently predict that this issue will become much more a matter of debate in boardrooms and across trustee tables. I also predict that somebody somewhere in the next five years will do something big that will move the market and make everyone else think seriously. I think the other thing that I would say, which isn’t so much a prediction as almost a statement of fact, is that this will all develop against the backdrop of a decrease in final salary provision. I don’t just mean schemes closing to new entrants, I mean there’s going to be closure to future accrual in large corporates.

We’re talking about a very emotive subject and you can pretty much guarantee that whoever the first large company is that makes this step, there’s going to be uproar from employees, from unions, from Joe Public, from the Daily Mail, from government. But it will happen.

RW, Paternoster In 12 months’ time I wouldn’t be at all surprised if we’ve already seen the first significant mortality swap transaction, whether it’s a capital market solution or insurance markets solution. I also imagine we will see the first syndicated buy-out, with a number of insurance companies involved in the same transaction.

GS, KPMG I’ll go for longer-term predictions. In terms of five years-ish, then actually I do very much agree with Joseph about seeing a much more liquid mortality market and mortality indices being developed, and more attention being paid on how to cope with basis risk – the difference between what’s in the index and what’s in your scheme. I also think as far as the buy-out and quasi buy-out market is concerned that actually we’ll see fewer providers in five years’ time than we have today. I think there will be consolidation.

DR, Watson Wyatt The mortality swap providers that will be most successful over the next five years will be those that can be most transparent in the construction of their assumptions and provide evidence of a multi-disciplinary approach that convinces trustees that they have considered the future, rather than relied on past methodologies.

AR, Credit Suisse I see in five years’ time a substantial market in longevity swaps, a much increased insured buy-out market – many billions in both cases. As for non-insured, full transfer of risk, there’ve been a few goes at it and it seems to have fallen out of favour, but I wouldn’t be surprised if somebody finds another way of doing that, maybe selling LDI bonds including longevity, or a non-recourse transfer to a provider, but one that’s acceptable to company trustees and regulator. And finally, I see a bigger role for consultants, or what we know as consultants now, becoming quasi-providers or having a provider role too. They’ve got a lot of expertise to give in helping their clients match up the different solutions.

PK, Watson Wyatt Will there be a market price in five years for mortality do you think?

AR, Credit Suisse It wouldn’t surprise me if there were.

PK, Watson Wyatt  Thank you everyone. That’s all we have time for.