Our panel of experts discuss the de-risking options available for pension schemes, and how they compare against buying annuities from an insurer

Chair: Myles Pink, Co-Head, Business Development Team, Rothesay Life

Mark Ashworth, Chairman of LawDeb Pension Trustees

Guy Freeman, Co-Head, Business Development Team, Rothesay Life

Ian Mills, Partner, Lane Clark & Peacock

Max Ramirez, Senior Member, Pensions Strategy Group, Goldman Sachs International

Mike Samuel, Pension Fund Trustee, Unilver Pension Fund


Chair (Myles Pink): It seems that this discussion is most relevant to those pension schemes between £500 million and £3 billion. We are here to consider how the de-risking tools that are now available to pension schemes to manage risk themselves (through what we might call a ‘DIY approach’ to risk management) can be compared with passing assets and liabilities to an insurer in the form of an annuity. Some of these DIY de-risking tools have been around for a while such as interest rate and inflation swaps. LDI programmes and strategies for de-risking assets into bonds have been used for some time to more closely match asset cash flows with pension scheme obligations. It seems to some that the longevity swap provides the missing piece to the DIY toolkit and that pension schemes in the size range we are talking about can now go it alone and not pay the additional cost of purchasing an annuity.

Max, can we start with how pension schemes should build a framework for a low-risk strategy, particularly with a view to moving towards self-sufficiency?

Ramirez: The majority of the work we do with pension funds is first to help them define what they mean by risk and how to decide how much risk they want to take. Once you know how much risk you want to take, it’s all about finding the asset classes that allow you to earn an efficient return for taking that amount of risk.

Mills: The way we think about allocating assets for our clients starts with what assets they are comfortable owning and what they are trying to achieve. They are usually seeking to generate return for an acceptable amount of risk, and they want risk defined in a particular way. I think it’s most helpful to think about the changes in risk rather than the absolute level. When we are making changes to investments the question we ask is “is the risk moving in the right direction?”

Samuel: I think it’s worth standing back a bit and asking where the trustees are in all of this? We talk a lot about best implementation but if you look where most schemes are, if they are closed to new recruits they are working towards an end-game and may well be steadily de-risking around a benchmark which the actuary has set. This typically suggests that for covering the liabilities for post-65 year olds, you should be in bonds and gilts; for pre-65 year olds, you should have an allocation to growth (risk) assets. So, if the trustees did nothing else, as the scheme matured they would just end up with everything in bonds and gilts because that’s where the actuary would lead them. So the real issue for trustees is do they want to take a position away from that journey or flight path. Any decision to move away would be governed by all sorts of factors, like what’s your perception of the employer covenant and its ability to support the scheme.

Mills: I think that’s right, but I think you have just highlighted one of the flaws. The way I see it, is that through that traditional actuarial approach, effectively what you are saying is: “I am going to back pensioner liabilities with a matching asset strategy. On the other hand, I am going to back my deferred member who is one year off retirement with riskier, unmatched assets.” But if you think about the obligation to the pensioner, it’s got payments next year, the year after that and every year until death. We back all these payments with bonds – even the ones due in 30 years’ time. On the other hand, the deferred pensioner who retires next year is fully backed with risky assets. I’ve never understood the economic rationale why some actuaries should treat cashflows due to deferreds and cashflows due to pensioners so differently. Instead, I think what matters is the risk and return characteristics of the whole asset portfolio and whether that is suited to the overall liabilities and the trustees’ objectives.

Chair: In that more holistic view, is there a role for an annuity or would you continue to dynamically manage everybody’s assets through to the run-off of the whole pension scheme?

Mills: That depends on the objectives of the trustees and what their risk appetite is. If you have got a scheme that is in deficit and very immature, then there probably isn’t a place for an annuity, but if you have got a more mature scheme that is better funded and the trustees don’t have to take as much risk, then maybe there is.

Ashworth: I agree with that. There’s not really a category difference between an annuity and other forms of investment. An annuity is a particular type of bond written by a particular type of counter-party, and it can be highly appropriate. As Ian says, it might well be inappropriate for a range of reasons including current unaffordability but I look upon it as a particular type of bond with characteristics that are tailored to the scheme’s particular liabilities.

Chair: Does everybody agree that if you are fully funded on a buyout basis, almost irrespective of the demographic profile of the pension scheme, an annuity is always the right answer?

Mills: If you are a closed scheme then I would agree, but it might be different if you are an open scheme and new members are coming in.

Ramirez: So if your question is: if it was possible (i.e. you had enough money) would you buy annuities, I think we would all agree and say yes. Whether that is possible or not, that’s a different question.

Samuel: All schemes are going to end up there eventually because each one is not going to be around paying its last pensioner: they will either have too much money or not enough. So you have to move into an insured pool arrangement at some stage and for those schemes that are closed to new recruits, the trustees know now when all of their members will become pensioners.

Ashworth: There has been quite a sea change. But there has been a realisation that we are all heading in a particular direction, whether we are large enough to manage it ourselves and effectively become an insurance type undertaking or whether we contract with an insurer to get there.

Freeman: Certainly how we get there is a very interesting topic. As trustees, do you see perhaps annuities being part of the regular asset allocation debate that’s had every few years? You described it as being like a bond and I completely agree with you. The emergence of the longevity swap market enables trustees to price the difference between a regular bond and an annuity in that longevity hedging can now be added to conventional LDI. So you can factor annuities into asset allocation analysis nowadays but I’m not sure that’s really happening yet. It seems to be something that could happen in the future.

Ashworth: Yes, I agree. It’s happening slowly and we have seen a number of schemes annuitise their pensioner liabilities through buy-ins and who now see that as part of their liability matching portfolio.

Samuel: It’s all about the funding position. If you are well-funded and are confident about investing in what is essentially an illiquid investment then you can annuitise your pensioners. You don’t really want to do that until you are confident that you can manage the remaining liabilities because if things change and you are still relying on significant improvements in your funding level to support deferreds then you have limited your flexibility.

Mills: That could be one of the attractions of a DIY approach versus the buy-in approach: you can more easily reverse at least the asset side even if it’s hard to reverse the longevity hedging.

Chair: But that comes with a potential downside as well because you have to then reinvest and you need to think what the reinvestment rate might be. When we compare the DIY approach with annuities, even though the DIY products are getting more and more sophisticated, there are still gaps around inflation caps and floors, operational and regulatory risks that can be passed to an insurer through purchasing an annuity.

Samuel: You also need to consider that most schemes are cash-negative and are much more exposed to price volatility in their risk assets. You may become a forced seller at low prices to pay pensions and you are never going to recover that loss, so for mature schemes which are cash-negative, price volatility is a real risk.

Ramirez: So would you be willing to buy an annuity for the pensioners which are the ones causing the cash-flow negativity and use the remaining assets to capture the expected return?

Samuel: Yes, but if you have crystallised the pensioner liabilities and you are under-funded in respect of your deferreds then you are heavily reliant on high growth assets and employer contributions. If that goes wrong in the early years then you have got an enormous hill to climb and that’s when you might take a view that you need to be less well-matched on the pensioner side and use some of that capacity to improve your overall funding.

Freeman: Do trustees start looking forward in terms of what their cash flow needs are and what they might actually need if, in the worst case scenario, they need more liquidity? I think there’s a lot of untapped illiquidity benefit that pension funds have; pensioner cash flows are fairly predictable apart from perhaps their linkage to inflation. But with deferreds you don’t know when they are going to retire; you don’t know if they are going to take a lump sum; or whether they are going to transfer out.

Chair: But of course, it’s not just member-related liquidity that needs to be managed; if pension schemes are heavy users of derivatives they may need a lot of cash and gilts to manage those positions. This is particularly important as the CSAs are moving towards that type of collateral and therefore a pension scheme that has a lot of gilts and cash may need them for collateral purposes as much as for the members.

Mills: I think that’s a really good point actually. A lot of our clients have backed their pensioner liabilities with bonds. Apart from the actuarial reasons for doing so that we talked about earlier, they also use those bonds as collateral to support hedges.

That’s one of the real reasons why we have to manage pension schemes holistically and focus on what we are trying to achieve overall, rather than doing one thing for non-pensioners and something else for pensioners.

Freeman: The management of pension funds is becoming more akin to the way insurers manage their risk and their liquidity. One interesting difference between a pension fund and an insurer is that an insurer at least has the ability to borrow money to fund those collateral calls, whereas a pension fund doesn’t have the access to do that unless they can perhaps sell or lend some of their assets.

Samuel: Yes, you could carve out a part of your equity portfolio to be run as a synthetic portfolio, selling physical equities to release cash and use futures contracts to maintain market exposure. You don’t get the full benefits of asset management but it is the sort of arrangement which pension schemes can use to fund their requirements on hedging programmes.

Ashworth: I agree with Mike, there are ways of indirectly borrowing but again you need to be sure you have thought through the implications if things don’t go as you hope and expect.

Chair: Rothesay Life manages itself as a low risk model through which most annuity risks are hedged out to create stability. The question is: can pension schemes become more like that given the products that are available to them? They could actually be more efficient in this under pensions regulation as they don’t have to set aside capital for the potential of very improbable outcomes. This compares with an insurer that does have to hold this capital and charge for it.

Freeman: Rothesay Life manages its assets and liabilities on a low-risk basis as you said. That is because we believe it is better for us to run a low level of risk and generate perhaps a smaller amount of return than holding more capital and taking more risk to reward the extra capital under a different strategy. That means that we are typically hedging out all interest rate and inflation risks and don’t take unsecured credit risks. We are trying to generate the best return that we can from the illiquidity advantage that we have as an insurer being the end depository of assets and liabilities.

That’s something that pension funds could mimic. A significant difference is that the pension fund may one day have to wind up and pass the assets and liabilities on to another party (probably an insurer). Requiring that flexibility to annuitise can be a block on maximising the return that a pension scheme can get from being an illiquid investor. Or put another way, trying to mimic an insurer could hamper a pension scheme’s flexibility prematurely.

Ashworth: I agree with you. I think that’s particularly acute with longevity swap transactions. Even with unwind and exit clauses in such transactions, it’s always going to be inherently difficult and potentially quite expensive to unwind and exit those contracts. There must be a risk that the contract won’t be acceptable to an ultimate insurer. The world may have changed in a way that perhaps at the moment we can’t envisage but makes them somehow inappropriate. So, thinking about the exit strategy when you enter into any form of long-term transaction is crucial.

Longevity swaps
Chair: So should more longevity swaps come with the option to pay deferred premiums and turn them into annuities with the same provider?

Ashworth: I think that’s attractive, but the question is what you are paying for that and whether you are comfortable limiting your choice in that way. You want to have a competitive choice at the time you come to buy-in or buyout so one just needs to think through whether there is a price to that option.

Freeman: The suitability of holding longevity swaps in pension schemes depends on the view of best estimate demographic assumptions, the cost of transferring the risk to the market and how significant the trustees perceive that risk to be.

Mills: The question is: is the cost that you are paying representing a good price for the risk reduction you are achieving? The big challenge with longevity swaps is quantifying how much risk is being removed and assessing it within the same framework as all the other risks that we are running. Longevity risk is quite different from most other risks. Interest rate and inflation risks move around on a day-by-day basis. Longevity risk doesn’t move quickly but when it does move, it tends to move persistently.

Samuel: People feel they don’t really have any good benchmarks for gauging movements in longevity risk, other than clearly it has been trending upwards. Actuaries make assumptions but nobody really knows. There are so many other risks to consider in the short term that trustees want to get rid of: volatility and so forth. So they park longevity for the time-being because it’s not going to surprise them. There is a tendency to push that risk back a bit, possibly wrongly, but nonetheless I think people worry about these short-term risks much more.

Ashworth: What we have actually suffered from more is repeatedly having to revise unrealistic assumptions and that process of revision has made us think that the risk is greater than it actually is because it hasn’t actually presented unexpected shocks. Perhaps there’s a remote chance something could happen tomorrow that suddenly makes us all nearly immortal! In that case, we are probably all in trouble whether or not we’ve got insurance contracts. I accept there are those sorts of risks but I think we have fooled ourselves about the extent of such risks and therefore perhaps we are over-inclined to want to insure against them.

I think the attraction of an annuity is that it deals with all the risks together, being investment, interest rate, inflation, operational and change of law; all of those things are more important. It’s great if I can cover off longevity as well as part of the package, but longevity isn’t particularly correlated with the sponsor covenant unless your sponsor is an insurance company. But, investment risk is typically correlated with the sponsor risk and those are the ones that can ‘kill’. I think you should focus primarily on the risks that can ‘kill’ and then worry about the ones that can be a bit painful.

Ramirez: Within the annuity, longevity seems to be a smaller part of the overall risk, but I can deal with most other major risks through a DIY approach. So is an annuity attractive because it deals with everything all in one? Isn’t it the fact that at certain points in time you can get better value than through DIY and the fact that all risks are taken away? So for schemes that have a lower governance budget it’s a much easier way to implement than DIY?

Ashworth: It is a perfect package. It leaves you with one thing to worry about which is the insurer, but one has the benefit of the insurance regulatory regime and infrastructure which gives a high degree of confidence around that.

Samuel: On the other hand, although you can monitor the employer covenant you can’t change it and the half-life of a pension scheme tends to be much longer than the half-life of the average corporate; let’s face it. So if you roll forward 20 years you will find a lot of funds, hopefully well-funded but not at self-sufficiency, who’ve lost their employer covenant. At that point they will have to be run like with-profits funds, whereby they pay those benefits they can afford, rather than honouring absolute guarantees. Pension schemes in the Netherlands are required to do that already; they have to cut back benefits if their funding level drops below a certain level. So, in the long term, realistically, there will be a lot of schemes having to be run like that.

Chair: Can we just take a few minutes to discuss what pension schemes are doing and where their priorities are? We’ve talked about some of the things that are low down in the priority order, but what’s right at the top at the moment?

Samuel: I think it’s improving the funding of the scheme - a combination of contributions and return on assets. Most of them are well below where they would want to be and certainly not self-sufficient.

Ashworth: The difficult discussions arise where trustees feel that they would like to de-risk and believe that the company’s ability to pay contributions is consistent with that de-risking but the employer, for understandable reasons, is reluctant to see that happen.

Mills: It may sound strange to some people, but there are actually some companies still out there who are concerned that if they put money in to bring the pension fund up to full funding that the pension scheme might move into surplus. We know that interest rates are currently very low by long-term standards; we don’t know whether they are going to go up or down, but if they do go up and you have just funded your pension scheme, then you might actually find yourself in surplus and then the sponsor will be thinking: “I’ve got a surplus but can’t get the money back out. The scheme is now closed so I can’t set it against future accrual either. It’s effectively gone.”

There’s also a security question to be asked. If you buy an annuity for the whole of your pension scheme, you are giving all of your assets to an insurer but if that insurer goes bust you have lost money. Owning the assets yourself may be the most secure way of running your pension scheme. Clearly there are lots of complications involved in running it yourself but you could deal with them and there is an argument that certainly needs to be considered. It might well be cheaper, not least because you don’t have to capitalise for the risks.

Ashworth: I agree. I just think you need to be very clear about the remaining risks and responsibilities that you’ve got and the cost implications of operating the scheme. Providing you have done that, it may well be a long-term option that some people would go for but you’ve really got to understand residual risks and the ongoing costs of that before you decide on taking the DIY approach.

Samuel: It might be right for a period. Eventually the numbers of members you have will get down to a level where risk then increases again and you can’t do anything about it. If you’ve only got two people left in the pension scheme, how much money do you need? So eventually you will have to go to an insured solution. The only question is: do you want to run it over a long period of time before you do that and when is scheme risk really starting to increase? Because the lives of individual members are very widely spread and you are losing the benefits of scale, eventually you will have to join an insured arrangement.

Solvency II
Mills: There’s also the really big question over Solvency II. If they did implement that legislation in the near term, would it mean that significant amounts of capital would need to be re-directed from corporate UK into pension schemes to bring them up to some minimum level?

Samuel: If you literally take money from the employer covenant and put it in the pension scheme you haven’t changed the left-hand side of the balance sheet; you have just moved resources from one box to another. Incorporating the value of the sponsor covenant is something that the corporate would much prefer than having to put literally everything that is required into the fund for all the reasons Ian expressed earlier about not being able to get it out if there is a surplus.

Mills: Plus, one other big point: if you do allow for the value of the sponsor covenant in your analysis then effectively what you are saying is that if the company is really strong it doesn’t have to put money in. That might be the short term conclusion but you’ll only ask for the money when the covenant is weaker which is almost by definition when it can’t afford it. So you never get it.

Samuel: And that is why trustees are driving towards self-sufficiency because the only realistic assumption they can make is that at some stage before they finish paying the last pensioner that covenant will change for the worse and they want to be in the best possible position at that time. The focus is on getting the funding level up to a point where hopefully, before the employer covenant has disappeared, they have reached self-sufficiency.

Chair: Can we finish by considering deferred members for a moment. We’ve agreed that if one can afford it, it probably makes sense to buy annuities to secure pensioner liabilities. Would you ever buy an annuity for deferreds?

Mills: It’s a difficult question. Certainly if you look at the pricing at the moment, annuitites for deferred members generally look pretty expensive and there are a few reasons for that. The first is that most of the longevity risk actually lies with deferreds; they are the people who are going to live longest and therefore there is more risk that we got the assessment of their life expectancy wrong because there’s a longer period for that to play out. The second reason is the reinvestment risk. There aren’t that many assets that you can hold with sufficient yield, that have a similar maturity to the deferred members they are backing. Therefore, you will need to roll into something else but you don’t know what that asset will be yielding. Insurance companies will tend to put a premium into their price to reflect that reinvestment risk. So annuities for deferreds appear to be expensive.

Then the other point is that deferreds aren’t like pensioners, in that pensioners don’t have the right to transfer out. Deferreds can just write to the pension scheme at any point, get a transfer value, take that transfer value and go somewhere else with the money. You need to make sure you’ve got that possibility covered.

Freeman: That means that the deferred annuities you might buy will need to allow for the liquidity to pay transfer values and that puts a constraint on the investment policy the insurance company can use to back these liabilities relative to liabilities for pensions in payment.

Ashworth: Deferreds tend to be included in either an immediate buyout or a buy-in that will lead quite quickly to a buyout. I agree with everything that Ian has said; I think the expense is a reflection of the risk. Whether the risk is run in the scheme or by an insurer, it’s got to be faced somewhere and there’s a cost to running it. In the pension scheme, one is hoping that the risks that one is running elsewhere are going to pay off and meet the costs of supporting the deferred members. It comes back to one’s appetite and ability to run that risk and frankly, we keep getting back to this, for many pension schemes, it’s just not affordable. That’s not to say that the deferred annuities are bad value, but it’s just out of reach for very many.

Freeman: I think there are different views out there about whether longevity risk is bigger for deferreds or for pensioners. For instance, if you take a group of 90 year old pensioners, there is absolutely tonnes of longevity risk there relative to the expected costs; if they live an extra year the price is significantly higher. Looking at deferreds it’s not that significant an impact on price. You still have longevity risk but it is an exposure to the long-term improvement trend whereas with a group of 90 year olds you’re more exposed to getting the base table right.

Mills: Maybe the way to think about it is the 30 year old might become a 90 year old at some point and at that point he has the same risk as the current 90 year old but he has also got all the risk along the way.

Samuel: The problem is that trustees recognise the long-term nature of what they are doing but they are held to account over much shorter periods by the regulator, by their members and by the employer. So they are locked into short-term time frames. They’d love to be able to run with a much longer-term strategy but they can’t. The regulator comes along and says: “Your funding level has dropped; what are you doing about it?” They have to respond.

Ashworth: We are on a journey. I think we all agree on where we are heading but it’s how we get there, particularly how we deal with challenges of under-funding. I think that’s high on the list of priorities at present.

Share Story: