The Chairman (Mr J. M. Lowes, F.I.A.): I will invite Jon Hatchett to introduce the paper.

Dr J.P. L Hatchett F.I.A introduced the paper. The full text of his introduction is given at the Edinburgh discussion on 16 January 2012.

Ms C. K. Lee, F.I.A. (opening the discussion): The earlier part of the paper gives a good summary of the different funding methods and perspectives of different stakeholders. It lays down a solid foundation allowing ideas to be developed in the later part. The authors identify six different funding methods ranging from pay-as-you-go to insurance buy-out. Although it may seem obvious, I find it particularly helpful for the authors to define exactly what each of these terms mean, especially when there is currently no real common definition of what self-sufficiency means within the pensions industry despite frequent reference made in funding discussions.

The authors also identify two funding arrangements: the sustainable approach; and the hedging approach.

The sustainable approach is defined as a scheme which is unlikely to require material funding from the sponsor, one which is likely to be well-funded and broadly matched. The hedging approach, on the other hand, was described by the authors as a scheme invested in assets which have very similar sensitivities to economic and demographic factors as a benefit payment. Re-reading the definitions a number of times, it is unclear to me what the obvious difference is in the two funding methods. The difference, I concluded, may lie within the asset strategy. Judging by the naming convention, I assume one is defined as a scheme which is hedged, and the other is not necessarily the case.

With the numerous funding approaches which have at times attracted criticisms by various stakeholders as confusing, is now the time for us to consider adopting a market-consistent liability valuation approach which is reasonably objective and, most importantly, a funding approach which is not influenced by the asset strategy?

In the middle section of the paper, the authors began to set the scene, bringing in the concept of a flight path and discussing dependencies among various risk factors which may result in deviations from the flight path, factors such as sponsor covenants, contribution, investment and benefit decisions. I personally share the view of the authors who identified funding negotiations which focus on a single discount rate crowd out the real important factors, the strategy of the flight path. The authors discuss the various targets that trustees and sponsors may wish to set, suggesting buy-out may be where some schemes are heading albeit with the obvious barrier of cost.

While different targets were discussed, the authors then stopped short of saying what this target should be. A flight path without its destination is simply driving in the dark. I was therefore quite glad to find that the authors did put forward a clear preference towards the later part of the paper called target expected return. It is, however, just that: a preference.

Legislation requires that an employer cannot walk away from their pension obligations unless a buy-out cost is settled. Yet, under funding legislation in the U.K. schemes do not have to fund to buy-out, but can use the magic of covenant to avoid pre-funding the difference. This leads to confusion when trustees are setting the flight path destination. Are we supposed to think the security level should be as secure as a buy-out, or should it be as secure as the employer is going to be solvent? Until policymakers can be clear about this point, a flight path destiny continues to be open to interpretation and the single number phenomenon has simply been shifted and labelled as something else.

In section 4 of the paper, the authors discuss the sources of risk and uncertainty expressed from the views of trustees versus sponsors. In particular, the authors rightly pointed out that under the current regulatory environment, trustees that target a so-called prudent basis are still significantly relying on the strength of the sponsor's covenant. Any risk borne by the scheme, such as to further contributions from the employer, should be assessed to show the likelihood of an unaffordable deficit. While some of the bets we have made along the flight path can yield positive outcomes, the possibility is often made at the expense of the non-pensioner group thereby creating intergenerational issues. Any bets made by a scheme are, of course, relying on the covenants to make good any bad bets. It is, therefore, impossible to speak of a flight path without giving the credit riskiness of the sponsor a mention.

I have noted comments made in the paper about the ability to quantify the employer covenant. Most would agree that it is impossible to guess the creditworthiness of an employer over the life of a pension scheme. I would however like to pose this question to the audience. The size of Defined Benefit (DB) pension schemes relative to the employer has never been bigger than we are seeing right now. Default rates of companies based on observed historical rates have never allowed for the possibility or the limited experience of DB schemes pushing an employer into insolvency, whether directly or indirectly. I therefore ask is it possible that we are systematically overestimating the strength of the sponsor covenant because we have simply not imagined how it can happen? Can we imagine a large-scale default in 25 years’ time by half the current FTSE 350 companies?

The authors also consider the implications of the presence of the Pensions Protection Fund (PPF) and how that affects the behaviour of the trustees and sponsors differently. Trustees and sponsors are not permitted to allow for the existence of the PPF. Some sponsors might consider that taking on risk to maximise the value of the PPF might be a logical motive for some.

The symbiotic relationship between trustees and sponsors means that trustees may have the motivation or, sometimes, sympathy to ensure cash-flows stay with sponsors as far as possible to maximise the life of a scheme before any sponsor insolvency occurs, knowing that the PPF provides a certain level of benefit security for members. This further complicates the decision when setting a flight path destination.

Section 5 discusses the use of discount rates, a topic which often generates a great deal of contention within and outside of the Actuarial Profession. The authors highlighted the benefit of using a single number to capture a great deal of information which indeed, as we know, has its benefits as well as its downfalls.

Out of the various discount rate methods which were outlined in the paper, the authors favoured the one called the target expected return, which requires the liabilities to be valued based on the long-term asset allocation such as pure gilts, market value of assets and explicitly stating the amount of funding which is coming from investment risks. The authors went on to discuss the benefits of financial economics for setting targets.

While there has been a trend in the last decade to move to a mark-to-market basis, for example in the case of accounting, it is not without its criticism. It can lead to not just the obvious volatility issue, but highlight risks which are short-term in nature. The result can lead to over-emphasis on risks which are arguably less of an issue than in reality. The danger is any risks not understood by policymakers will simply lead to higher capital requirements in whatever form to mitigate the risk. I therefore ask: is holding more and more capital the only answer to all risks? If not, how can we avoid falling into this trap if we are to move to a market-consistent style of liability valuation?

The final section of this paper, in my view the most important section, summarises the whole point of this discussion paper and answers the “so what?” question. A different way to how most pension actuaries have considered disclosure so far, the authors proposed an alternative method, although not so radical, of communicating risks and uncertainties.

The authors’ suggestions of disclosing probabilities of outcomes from the member's point of view, such as how likely members will get their full benefits, and from the employer's point of view, such as how likely will deficit contributions be required, are welcome suggestions. In a world where accountants are good at telling you how much you have yesterday, and insolvency practitioners are good at confirming that you are bust, it is only right that actuaries have a role to play somewhere in between, assessing how likely you are to go bust before it actually happens.

An excellent, yet simplistic, example the authors put forward towards the end of the paper is about quoting the likelihood of outcomes. The concept of quoting probabilistic outcomes is nothing new, of course. By showing probabilities in two different views, the likelihood of getting full benefit from the member's view versus the likelihood of employer having to top up the scheme, shows just how different are the perspectives. While actuaries and investment professionals do talk about investment risk by means of probability of success (for example, 50% chance of getting investment returns of 7%) which is often shown in the eyes of the sponsor or trustee.

But telling members that they will have only 50% chance of getting their full benefit is a rather different story. To stretch this argument further, can you imagine, in the next annual summary funding statement if we were required to disclose that pensioner members have an 80% chance of getting their full benefits, while non-pensioner members have a 30% chance of getting their full benefits. Can you imagine disclosing that? And yet that is exactly what it means when a scheme is taking risks, using the creditworthiness of a sponsor as backing.

As good as the current disclosure requirements of pension schemes are, members are simply not aware that is the case. Interestingly, the concept of a holistic balance sheet, as proposed in the recent European Insurance and Occupational Pensions Authority (EIOPA) consultation on the call for advice response for the so called Solvency II on pensions, was explicit that, in considering the concept of the holistic balance sheet, it is viewed from the eyes of the members, which clearly yields a very different perspective to that of the sponsors.

Affordability and security are two quite contradictory terms. Both appeared in the paper numerous times. Can a benefit promise be truly affordable while perfectly secure at the same time?

Mr P. N. Thornton, F.I.A.: I have certainly become increasingly concerned in the last year or two about the way in which the actuarial advice to pension scheme trustees has been sitting in something of a silo, and rather hooked up on deterministic valuation bases. I appreciate that this is driven by regulatory requirements, but as the authors quite validly point out, it is not the best way of approaching things any more.

The comments I want to make, however, are around the sponsor covenant, in particular the allowance for insolvency risk. I rather think that there is too much complacency among trustees about sponsor covenants, particularly when they think it is currently strong. Of course when they no longer think it is currently strong, it is too late to get much done about it.

It would be healthy if trustees thought more deeply about the possibility that their sponsor might not be around in 20 years’ time.

I believe that we need more sophisticated allowance for insolvency risks, and that this should indeed impact on how trustees view flight paths, actuarial assumptions and investment strategies. I am currently working in the covenant assessment area rather than providing actuarial advice and I should like to see this side of advisory work come together with the actuarial side in a rather better way. I look forward to working with different actuarial firms on this topic.

It is not just the, possibly extreme, case of insolvency that matters. Some research which we have been doing in my firm into the fate of the companies which were, 25 years ago, the FTSE 30, shows that out of those 30 companies over 25 years, only four of them had a good covenant experience. One of them defaulted – only one – but eight of them at least were financially stressed. (These numbers do not add up to 30, by the way, because more than one experience happened in some cases.) 19 of them got taken over and 10 of those ended up with overseas parents. 10 companies experienced demerger or major downsizing. So even in the biggest, most famous companies in the land a great deal happened to their pensions covenant over a 25 year period, which is not an unrealistic timescale in terms of pension schemes. There are certainly no grounds at all for complacency. Accordingly, I think there is more work to be done on taking proper account of the risk of the sponsor covenant.

Mr I. J. Kenna, A.I.A.: I was disturbed to read the reference to credit default swaps in section 2.3.1 of the paper. In his article in November 2011, Social Sciences in China, Professor Xu Duoqi wrote as follows: “However, research estimates indicate that total US financial product contracts before the outbreak of the crisis amounted to 530 trillion US dollars in which credit default swaps exceeded 60 trillion while the collateral security or material financial assets on which they were based were worth only 2.7 trillion. Their leverage rate thus exceeded 200.” I should not expect the U.K. to be any better than the U.S. in this respect.

Section 3.2.4 contains a timely reference to the effects of scheme disinvestment. When a scheme's cash expenditure exceeds a scheme's cash income, it is necessary for the scheme to sell assets. When the total cash expenditure of long-term savings institutions as a whole exceeds their total cash income, selling assets becomes mandatory. Official figures show that the point of total expenditure exceeding total income overall has nearly been reached. With the closure or partial closure of defined benefit schemes, this tendency will accelerate.

I did not detect any reference to longevity improvement in the paper. I thought it might prove of interest to say something about longevity all the same. Has there been any real longevity improvement? Is more improvement likely to occur?

The profession needs three pieces of fairly easily obtainable information:

• - Firstly, the total cash expenditure of U.K. long-term savings institutions in a given period;

• - Secondly, the total cash income of U.K. long-term savings institutions in the same period; and

• - Thirdly, the current age of the oldest person in the developed world.

I suggest that this information be published regularly in The Actuary.

Mr A. M. Slater, F.I.A.: There is no focus on income in the paper. It is all about asset values. Focusing on asset values is well known. It is called speculation. If you are dealing with flight paths for pension schemes, you have to be focusing on income as well.

I fully support the promotion of stochastic asset-liability techniques to both understand and communicate the risks inherent in policy decisions. If you are brave enough to go to a stochastic valuation, then that is a good thing, too.

The authors take the view, perhaps implicitly through familiarity, of what I call a scheme-centric perspective on pensions. However, a pension stems from a labour contract, the exchange of labour for deferred pay so an employer-centric perspective is fundamental. It has to exist. A pension scheme can be sufficient as the delivery mechanism for pensions, but it is not necessary.

The authors get close in several places, such as section 2.1 when they discuss alternative approaches to funding and financing, but are hampered by a scheme-centric perspective. The fundamental question is how to provide security to the deferred element of deferred pay. The build up of a portfolio of assets is a security mechanism and itself can be enhanced by placing the assets in a trust separate from the employer. However, there are other security mechanisms used around the world such as indemnity insurance against sponsor insolvency. We are talking so much about the problems of the employer's covenant, let us tackle that risk head on. If funding is the chosen security mechanism, then its quantum and investment strategy is second order: there is a taxonomy with a hierarchy and we need to get it right and we need to get it explicit.

In section 4.2 the authors say, “a discount rate might impact on the present value placed on a set of obligations, but it is not a source of uncertainty in the obligations themselves.” Absolutely right!

In section 6.4.1 the paper states: “it might be convenient shorthand to discount these payments to time zero to get a present value, but care should be taken to avoid losing information that might be valuable”. This is my second takeaway message; discounting destroys information.

Mr J. G. Spain, F.I.A.: The authors draw attention to the tendency to focus upon a single value at a single point, to which Mr Slater has just referred. However hallowed, it does not help. It should be stopped.

I was really pleased to see the time horizon mentioned but it was not mentioned, in my view, as far up the ladder as it should have been. It is all important to ask “how long do we have? “Mr Employer, do we have two years? Do we have five years? Do we have 10 years? You cannot guarantee anything.” We cannot go long and short at the same time. If we want to go long in reward seeking securities, we would like to know if we have a reasonable chance. I want at least 10 years to make that worthwhile.

The opportunities, essentially for extracting reward from other people's fears, which is quite a good thing for trustees to do, have basically been ignored.

However, I think that there is a severe imbalance in the paper. A word count shows the word “risk” coming up 250 times, which is not bad. The word “reward”, which is the one which I was looking for, comes up three times, and two of those were within the phrase “risk-reward relationship”. We also have the statement “risk will happen, reward may happen”. Why must we always race down to the bottom and assume everything is going to be worse?

I think I am probably one of the few people who holds the opinion that taking everything at market value does not work for a long-term institution because short-term market movements are very, very poor long-term indicators. It does not work. If you have a longish horizon, 10-plus years, so long as you do have it, then you have the trustees having investment opportunities which they will not have otherwise. In fact you could say that market values are, in a sense, a different centre of the world.

Way back, before market values became the unit used in actuarial valuations, we were living in a different world. We were living in a world where employers were not forced to take on benefits which they had never agreed to take on in the first place: guaranteed revaluation; guaranteed pension increases; preservation of benefits itself; security of the kind that we now see imposed upon employers. They would have walked away long before June 2003 if they had known that the bar was going to come down very suddenly.

Why ignore the evidence that market value does not work for a long-term entity if it is a long-term entity? Is this a flight path that we see before us? No, I do not think so. If the trustees can stop us doing that, and the authors can stop us doing that, then that will be a real service.

The Chairman: We have had a number of contributions on theoretical aspects. No one has commented that, although in an ideal world we might want to disclose in this way and tell the full truth, the full gory detail, do we think trustees are comfortable disclosing a deficit against a long-term asset strategy-based funding basis when in fact they are, if you like, in line on the flight path with their current investment strategy expected to produce the goods with no further contributions?

Mr M. G. White, F.I.A.: Directly in response to that question, I think it would be difficult for the profession to hold and justify to the world the “don't disclose in case you frighten them” view, given the discussion so far.

Mr Slater: I suppose the question should be not whether the trustees are happy to disclose (and really it should be trustees and sponsors together) but whether the Actuarial Profession is happy to advise people to do so.

Mr D. I. W. Reynolds, F.I.A.: When we are talking about stakeholders, I think that the most important stakeholder, is the member.

The first instinct I suspect to that will be to say that the member must be certain to get their benefits.

So if you talk to members and say, “Do you want a defined contribution scheme which is going to be at risk when you take your pension, from what has happened to investments, what has happened to inflation, what has happened to mortality? You may know you will get a certain sum, but you have no idea what income you will get. Or would you rather have a defined benefit of which you may only be 80% certain of receiving?”

I suspect that if you talk to a lot of members they would say, “I will go for the 80% certainty of a defined benefit rather than take risks that I do not understand. The actuaries cannot explain to me what the risk is of what income I will get when I get to retirement, not least because they probably cannot tell me what retirement age I am going to have.”

So, if we look at matters from the viewpoint of a member of a pension scheme, we will see that what we should be trying to do is to find how to keep defined benefit schemes in existence, and how do we make them more fundable in the future.

Dr Hatchett said trigger points are very useful in obtaining agreements between stakeholders. So you will agree a point at which the trustees will de-risk the scheme and sell the equities. This will be a bit like the way that the FSA has approached equity risk for insurance companies and said, “Fine. The stock market is at 5,600/5,700 today. But what are you going to do if it gets down to 4,600?” And what if the companies say, “We have a trigger point. At 4,600 we will sell all our equities.” That seems to me absolutely ludicrous. What I want us to see is the method whereby if the market falls to 4,600, you are in a position where you can buy. So your trigger point should be “It is at 5,600, if it gets to 6,000, I will sell, and then I will put in the position that if it goes to 5,000 I will buy.”

Let us obtain some sort of self-protecting situation rather than setting a target to sell when markets have fallen.

If we also introduce solvency regimes for pension funds and the pension funds having to divest at the same time, what do we obtain? We actually have a mechanism for driving the market further and further down.

Let us try to make sure that we have trigger points, but make sure they are actually, in the long run, beneficial and not a short run view.

The whole issue of market price is that it is a subjective price between a series of people and it has become worse with hedge funds and the ability of people to short the market. The one thing that the market value is not is objective.

Dividend income and interest income are very significant parts of what is going to drive what value assets will have in the future.

Can we turn the debate so that we are looking at issues always through the view of the members?

The Chairman: Judging from the reaction to my previous comment, it feels that everybody in the room thinks that we should be disclosing to members a deficit when we are on track. How does everybody feel that fits with the regulatory regime where if you have a deficit you must get rid of it as quickly as possible even if that deficit is there on a planned basis rather than because you are behind where you want to be?

Mr B. H. Davies, F.I.A.: I am going to express some doubts about the widespread myth that if we had not had a full protection for early leavers, if we had not had a requirement for solvency supervision, that if the other changes which have taken place in pension schemes had not happened, then all would be fine. I do not think that there is any evidence to believe that.

The big change which has taken place is in the attitude of employers towards providing decent pensions. The legislative framework has moved on – there is no doubt about that. I have seen the change over the last 30 years or more, and the implication that employers would cheerfully be providing defined benefit schemes if we had not had all these changes in the legislative framework, to my mind, is a myth. It is an argument that I want to pursue and needs a lot more work, but, at this stage, I am putting down a marker.

The second point that I wanted to make was about market values. I have always found the problem with market-based valuations is there is not a market in pension scheme liabilities. So how can you have a market basis when you do not have a market from which you obtain the price? There is a market in insured liabilities for pension schemes, but we know through the discussion that is taking place on the solvency standard that there is a distinction between pension liabilities and insured liabilities.

I have made the point previously and it was said, in response, that you have to price pension liabilities on the occasion of company mergers and acquisitions. My experience is that the price put on pension liabilities in those circumstances is that it is a figure fixed at the end of the process and is not arrived at independently.

Since there is no independent market in pension liabilities how can you have a market-based valuation?

Mr White: A very brief response on the change of attitude of employers. In contrast to the situation as we perceived it back in the late 80's, the expected real return from assets is now far lower. It's not just tax changes. The fundamental problem is low real interest rates and consequently high asset values. So the cost of providing a pension of X is much higher than it was.

The Chairman: Nobody wants to come back at me on the comment on the regulatory position?

Do we think the profession can define what we would like to use as a term for self-sufficiency? Do we think that is a term we can choose to define ourselves and impose that or do we think we have lost that one to others outside the profession?

Mr D. C. E. Wilson, F.I.A.: I did not feel that there was much of a definition in this paper of self-sufficiency. I understand that there is a requirement here to say that it must be more than the total assets that you think you need to meet the liabilities because you need some sort of level of certainty around that.

However, there was nothing about what probability level or what you would do with the excess if you ended up with an excess after the event. I would appreciate more of an explanation about actually what is meant by self-sufficiency here.

Dr Hatchett: The words I used in my introduction were: having a pretty good chance of paying the benefits in full without financial or any other recourse to a third party.

What do I mean by “pretty good chance”? To me a self-sufficient scheme ought to need more money with a very low level of probability. We did not say what that level of security is, and that is really for trustees to decide. I think if a member is told that a scheme is self-sufficient you would not want to go back to them five or 10 years later and say it is not any more because A, B, C and D happened. Or, at the very least, you would only want there to be a small chance of that happening. To be clear, I am not saying schemes should fund to a self-sufficient level, but if they do describe themselves as being self-sufficient then they ought to be self-sufficient.

This may feel all a bit “hand wavy”. But I think, inevitably, it is subjective. Actually trying to be too precise about something which is a tail event and is subjective is not helpful. I don't think it is necessarily for the Profession, and it is certainly not for me, to say, for instance, that self-sufficiency should mean a 95% chance of delivering benefits in full over the lifetime of the scheme. It depends on your risk appetite in making that statement. But I think a one in 20 chance of failure seems comparable to me for the common understanding I would have of the word.

Mr M. C. T. Hurd, F.I.A.: We stayed clear of giving a very hard definition of self-sufficiency because we thought that this would detract from the main aims of the paper, which were the importance of setting clear long-term targets and how pension schemes would reach those targets.

If a pension scheme is self-sufficient from the sponsor, then it should have minimal reliance on the sponsor. This is distinct from self-sufficiency being used to mean posing a lower risk than at the current time.

Mr Wilson: Having defined self-sufficiency like that, are you advocating that there may be a reason for pension schemes to fund to that level or are you asking them to come up with a different name for what they are doing?

Mr Hurd: We were not advocating either way. We were highlighting the issue and encouraging stakeholders to the pension scheme to agree between themselves what the appropriate target might be for the pension scheme.

Mr Thornton: This comment is not about self-sufficiency, but another aspect of the paper which I think is quite interesting but has not had much airtime so far, which is the sponsor's perspective and the shareholder's perspective.

I think the market is not particularly rational. As we have all these different measures of what a deficit might be, and we have financial reporting figures which are substantially different from technical provisions which may be quite different from what the actuary suggests could be appropriate on an on-going basis, and certainly based on different assumptions from what is used in setting the asset strategy, the picture I take is that market analysts do not pay much attention to the actual financial reporting numbers in the financial statements. Instead they look at the disclosures and work out for themselves the affordability of the contributions that have been committed.

Perhaps that is not quite as definitive as one might think, but what then happens is that if there is an aura about a particular company that it has a pension problem, it becomes far more attractive to that company to do something about the pension problem because that can improve its credit rating, or its share price, or its access to capital, or increase its strategic opportunities.

Some of the companies which have proceeded with buy-outs have done so because it has made sense to do so from a much broader perspective than actuarial, value for money, or even return on capital. If the market is not paying attention to a pension problem in a company, it is probably quite efficient for that company to run a deficit and accept a reasonable proportion of the assets in equities, as long as it feels its covenant is strong enough to take the risks involved.

This is certainly an area where there is not a rational picture in the market, yet it is important to the flight path.

If we were to go down the path of trustees being far more explicit about agreeing a flight path with the sponsor and then disclosing that to all concerned, I think that might produce more rationality over time.

Mr Spain: I think both the sponsor and shareholder are the same. What they should both be looking for is the opportunity perhaps to get some money back. Why should they take downside when they can never get any upside? Paying off a deficit over a short period is almost bound to make things worse.

The current regulatory regime is certainly designed to try to make deficits be paid off over a shorter period rather than a longer period. However, it is just about imaginable that a current regulatory regime might change one way or the other. In fact, the Actuarial Profession ought to be thinking, perhaps, about what would be a good thing to do, rather than how do we fit in with the current regulatory regime? I am not hearing that.

Mr Thornton: I do not think sponsors and shareholders are identical. Sponsors are the entity and the management, and that management is working to maximise shareholder value. The shareholders have different interests and they can buy or sell their shares. They can move in or out of ownership of companies. I do not think that it is right to regard sponsors and shareholders as identical. Their interests may, it is to be hoped, be aligned but they are certainly not always aligned.

Mr J. P. Ryan, F.I.A.: I think that it is worth exploring the covenant a little bit further from the member's perspective, and indeed different members’ perspectives, because enormous variation depends on what age you are, and how close you are to retirement, or indeed whether or not you are a pensioner. I have been involved in situations where companies have become in severe financial trouble with a pension fund. The pension fund was not necessarily in a strong position but clearly had cash because it contained assets, even though there were not enough. The working members were all in favour of investing some of those assets in the company to keep it going because that meant that they kept their jobs.

When you are looking at covenants, you have to be very careful where you look and your point of view. A lot of the paper, which is very valuable, is looking at it from the company's point of view, and it is worth remembering that. If you are looking at things from the member's point of view, then you have to look at the duration and where it comes back through. In many cases the member, certainly the younger member, is going to be much more concerned about job security and their future employment prospects than they are about the ongoing solvency of the fund because they have more at stake in the former.

The same issue also comes through when you are coming up to retirement. People approaching retirement are going to be much more concerned with “Am I going to get most of my pension or nothing at all?” If you are a young employee you really want good prospects, and so on, coming up with something else. When you are close to retirement, you have not an alternative to do anything else. I think that that is something that is worthwhile thinking about.

There have been some rather negative comments on market-based prices. I have some sympathy with that in that a lot of the time market-based prices can be distorted. Certainly going back to the days when everything was done at book value, an enormous number of companies hid behind a book value to avoid giving information that analysts could look at, and that hid all sorts of things.

Do not let us just denigrate market based prices anywhere because we are likely to find that we end up with no disclosure at all, and not enough information for third parties and members to have a look at it.

Mr H. N. H. Peard, F.I.A.: I think maybe what is going on is historically work in the pensions arena relied on what would now be called the sponsor covenant. That was strong. There was an attitude that people would be looked after.

The environment has changed very significantly over the years. I think many of the questions that are now actually in need of being addressed can only be addressed using market-consistent or, at the very least, stochastic-type analyses.

So, for example, if I am a member of a pension scheme and I am not yet in retirement and that pension scheme is not fully funded, my preference, as that member, is that you invest in risky assets and that you are not matched. That increases the probability that I will actually get paid my benefits.

If the scheme is very fully funded, or if I am close to retirement or in retirement, I would probably want to see a totally different fully matched strategy.

What this is pointing to is that there are not quite conflicts but certainly non-alignments of the interests of different parties within a pension fund, also between the pension fund and the company. For example, if I do not work out a deficit on a market consistent basis, and I do not value the sponsor covenant on a market-consistent basis, then somebody is going to get arbitraged, and it is probably going to be the pension fund.

I am going to defend the term about objectivity of market consistency – it's pretty objective. It is something for traded assets. You can look at a value at a time or a date. Yes, there are problems in applying it to some of the pension liabilities, but you can do a proxy mark-to- market. You can deconstruct a lot of the liabilities in the pension fund and come to a good basis.

Does the market-consistent value answer all the questions that you need to answer? My answer is, no. For a long-term view for questions like, “How much do I expect I will have to pay in a real world?”, then I think you have to use a real world valuation using risk premia. But what is the right risk premium? I think you have to show a range of results which show how wrong you might be.

I think the risks are different for different members of a pension fund. One of the things that I had thought up is that the introduction of Solvency II and the possible application to IORPs is close to disastrous in the pension scheme context.

However, it does have a good benefit in that it will ask people to look at the risks and think more about what is actually going on. I think it is essential that we re-phrase the questions and think about what are the questions that we really are trying to answer and what is the right methodology for answering each question.

Dr. Y. Avrahampour, London School of Economics & Political Science (guest): In my view it seems that the discussion illustrates that we stand at an intellectual impasse regarding our thinking about the relationship between pension fund management and governance.

In order to resolve this type of impasse one has to look at the foundations of the different theories that are applied to understand pension fund management and governance.

One approach is to treat the sponsor's shareholders and the members as principals of the pension fund which have conflicting objectives. The shareholders of the sponsor are principals who want less money in the fund. The members are principals of the pension fund who want more money in the pension fund. These conflicting objectives can be reconciled only under situations of ambiguity or lack of transparency. It is for this reason that use of market values is associated with the decline of defined benefit pension revision.

The criteria for objectivity that we might think about in order to resolve this type of intellectual impasse, rather than being accuracy in determining cost, would be productivity. So focus should be on how actuaries enable managers of the pension fund to maximise the productivity of the assets of the pension fund. We should use that as the criteria by which to compare different solutions regarding the degree of governance that is appropriate.

That type of thinking with respect to how you model the relationship between principals and agents in the pension fund will come to very different conclusions as to the type of asset strategy that one should adopt and the framework of governance within which that strategy should be set.

In order to progress further, intellectually there will ultimately be a need to look at the roots of the theories that we are using to understand the present situation. Financial economics has a lot to contribute but it also needs to be combined with other theories from other social science disciplines in order to provide us with the answers to proceed forward.

I concur with those raising questions regarding the role of the Actuarial Profession in this context because I believe that this is a rigorous type of thinking. And actuarial expertise could be used more widely than is currently the case.

Mr P. M. Greenwood, F.I.A.: Nobody has suggested what I think is a sensible long-term solution. When government legislates, as it may do, for the state pension age ultimately to become longevity-related, we need to build up pressure for the same thing to happen to any remaining defined benefit rights, and to permit that with adequate notice for that to be backdated.

I agree that the bells and whistles added to defined benefit schemes were introduced for a reason: inadequacy of provision. For pension provision to be sustainable long- term, it has to be both affordable and adequate.

Unfortunately, the movement in longevity occurred too quickly. We probably did not have much chance of spotting the cohort effect early because those born in the period of good health caused by war conditions and not over-the-top living standards have reached the age where death starts to be recorded at ages 60 and 70. Those born in the 1930s and the 1940s started to die in the years 2000–2010.

You need some form of sensible valve that is going to control costs when you do not know the future. That sort of provision for longevity changes might do it. Some of the provisions that have been called for, like removing the bells and whistles, will not work because they will just lead to further issues emerging later.

Mr T. P. Connor, F.I.A.: There has been mention of how the regulatory regime perhaps does not encourage such an approach and that disclosure of the information which trustees and sponsors have to record or calculate might be a potential hurdle. Perhaps there are some hurdles in the communication that we have around terms of self-sufficiency.

These are all things which can be overcome. If we as a profession think that this is a good regime, perhaps there are some things which we can look at as positives which we could try to influence to help encourage this framework for better decision-making.

So, for instance, the FSA is considering working with some of the financial institutions to draw up “living will” documents for what will happen in certain circumstances. Perhaps as a profession we can encourage the Pensions Regulator to encourage schemes to have their own living will document for how they expect their own circumstances to evolve, which I think is very similar to the flight plan idea.

There are discussions in the profession at the moment about how scheme actuaries interact with companies, and what restrictions there might be for that. I think one of the potential barriers is how do you get an engaged employer and an engaged trustee to work jointly together.

Are there sufficient incentives for all employers to address this area with the necessary work and effort in order to get a sensible outcome?

We have heard of some of the potential lack of incentives as a result of the current level of disclosures. Again, perhaps that is an area which, if there was fuller disclosure of what companies were doing with their plans, which might again encourage a more sensible framework and support development in this area.

What would be good to see is a document setting out a proper description of what trustees and companies are trying to do with their pension plans, where they are heading, what the contribution requirements are likely to be, and get away from what has evolved, as in so many areas, into a compliance-driven document which does not give a lot of information to those not directly involved in the discussions about what is actually happening with the pension scheme.

Mr Kenna: Redington pointed out about 35 years ago that there was far more money coming into the long-term savings institutions than was going out in cash terms. The resultant investment was driving up the price of shares, with a few hiccups of course. Everybody thought that shares were such a good investment because they were going up under this pressure that they bought even more, pushing share prices up again. This resulted in large surpluses in pension funds. We went out to the clients and we said, “You have a choice. You can either have a contribution holiday or you can improve the benefits.”

Clients took us at our word and they took contribution holidays or improved the benefits. Then we arrived at the year 2000 when shares were obviously overpriced. The FTSE 100 went up to just over 7000 and the yield was $$$2\tfrac{1}{2}\% $$$ , with no growth.

The price of shares slid down and we had to go out to the clients that we had been advising to take contribution holidays or improve benefits, and recommend much increased contributions. Sponsors did not like this: it led to a lot of pension schemes closing.

In the past two or three years, the overall situation has changed just the other way, thanks partly, of course, to the closure of pension schemes or the partial closure of pension schemes. More money is being paid out than there is coming in. Eventually, shares or other securities will have to be sold at a time when it probably will not be very convenient to do so.

It is this kind of gigantic see-saw effect over the year which has led to the trouble. It is not government regulation and it is not perversity or meanness on the part of sponsors. Sponsors are entitled to expect that pensions actuaries take account of the long-term overall position of this market and use their skills to devise smooth contribution rates.

Mr M. S. Demwell F.I.A. (closing the discussion): With respect to self-sufficiency, Mr Wilson pointed out the importance of defining this clearly. The phrase has become commonplace in pensions in recent years. Yet, there is no uniform understanding of exactly what it means. Many would regard it as the broadly matched position that the authors describe as “sustainable”.

I am glad that the authors have shone a light in this area. They suggest that self-sufficiency should be regarded as the Solvency II-type approach currently being considered by the EIOPA. Whether that is a better definition of self-sufficiency remains to be seen. However, those who would impose Solvency II on U.K. occupational pensions should note, as the authors fairly point out, that there is quite a wide range of views about what self-sufficiency means, and that should not necessarily be regarded as the only view within the profession.

I thought the authors very helpfully listed six variables upon which any flight path should be based.

I am less convinced by their argument that “lower contributions will in most cases decrease the likelihood of sponsor default if the cash is used effectively within the sponsor's business”. That is a very big “if”. I do not know of any banks that would consider the more they lend to a borrower, the lower the default risk.

However, I do heartily commend the authors’ admirable clarity in stating that “just because yields are viewed as being low now, it does not mean that they cannot go lower and stay lower for a long time”.

Mr Thornton drew our attention to the dilemma that trustees often face: sometimes they can be a bit complacent about employer covenant, but by the time problems become apparent, it is too late.

Mr Slater supported the value of stochastic modelling and called for more progress towards stochastic valuations.

He also called for an employer-centric focus and pointed out that discounting destroys information. In some ways that is true. It also simplifies a lot. We must be careful that in the search for more meaningful information we do not overlook the value of simplicity.

Mr Reynolds pointed out the importance of being member-centric. And for members to understand their benefits, and the risks around them, they have to be described simply.

There was quite a lot of debate about market value. Mr Spain thought it was unhelpful.

There was a suggestion that triggers should be set so that schemes sell at the top of the market and buy at the bottom of the market. I am happy to say that, in my experience, properly set triggers along funding paths do tend to support that kind of action.

Mr Davies said that the biggest change around leading to DB reductions in the past 30 years was the attitude of employers. Mr White suggested that it was because the cost of providing pensions had increased so much. I am not too sure about that. I think it has increased but I think what has really happened is that people have become more realistic about what are actually the true costs.

As others before me have observed, another key driver has been earnings caps which disenfranchised senior executives from pension schemes. Now, for the avoidance of doubt, I am not in any way putting forward trickle down arguments, but I think when senior executives were first encouraged out of pension schemes that was one of the big drivers to DB reduction.

Mr Davies pointed out that there is no market in pension scheme liabilities. That is true, but that is not a reason to throw away market values of assets.

Mr Spain also said that he thought sponsors and shareholders are the same. Mr Thornton disagreed with that. I would disagree as well. If sponsors and shareholders were the same, we would not have had the fun and games around the bonuses paid to senior executives within publicly-owned banks.

Mr Ryan pointed out the different interests of different categories of members. It is true that members further from retirement have more time to make up a lost £1 of pension. But in my view, trustees should be equally protective of everybody's pensions, no matter how far they are away from retirement.

The authors of the paper have suggested a few areas for further research. They have suggested that an interesting area for future study would be an analysis of pension scheme funding that tackled the approximations inherent in using a risk-free building block. That is a good suggestion although “risk-free” makes me slightly anxious. I think that there was some consensus today that there may not be any such thing as risk free.

The authors also observe that “sponsor covenant risk is rarely included in valuing alternative funding promises or recovery plans.” How this could be achieved may be another area worthy of future research. And of course what exactly do we mean by self-sufficiency? It would be good to see that discussion taken further.

Finally, the authors state that there has often been an under-appreciation of risk. They go on to say “that all schemes already have an implied flight path but that many do not recognise it as such”. I agree that pensions risk is still often poorly defined and understood. However, following all the events of the past few years, the climate is certainly favourable for those who would promote better management of pensions financial risk. Indeed, it is of the utmost importance for actuaries to press the point now because memories fade fast and it may not be long before some will be arguing that risks have receded.

The authors have made a valuable contribution to this debate. There is a lot more work to do and it is to be hoped that members of the profession will continue to develop the arguments.

Mr Hurd (replying): I think that it is worth being clear about what it is we were trying to achieve. The first thing that we were trying to do is set out a framework, and the idea was that stakeholders could come together, agree a mutually beneficial target that both sponsors and trustees wished to achieve and then, over a time frame, deliver an investment strategy, both now and in the future, to help them get there. That is what we were proposing, and I think some of the conversations went off at a tangent, although a lot of it was very valuable conversation.

We go one step further and discuss disclosure and put emphasis on risk and reward. What we are trying to say is where we are reliant on reward let us state that, let us be explicit about that, and give it equal prominence along with the risks being run at the time.

Finally, all of this is around creating a framework to enable sponsors and trustees to deliver better decision-making. I think it is our role as a profession to use the tools that we have available to enable our clients to make better decisions. If using stochastic modelling is going to cloud the decision-making process, and lead them to make an incorrect decision, then we, as a profession, as individuals, have to limit the information we present and present it in the right way for the right client. It is our judgement, and that is where the skill of an actuary really comes in.

However, what we should not do is dismiss developments in the science purely because we find it difficult to communicate. Rather, we should work on the communication, not on limiting ourselves.

There was a lot of talk about stakeholders and, in particular, the sponsor's viewpoint. What we are trying to do here is to create a framework whereby sponsors and trustees reach agreement on what is that long-term target. Experience suggests that in the majority of cases the target is actually relatively easy to agree. For some it may be buy-out. For some it may be fully invested in a very close to matching portfolio. But actually, the target is quite easy to agree. What is more difficult to agree is how you are going to get there.

What we are trying to do is have a collaborative arrangement whereby sponsors and trustees can reach agreement on what they do agree upon and focus on the areas that they do not agree on in a more clinical way.

There was a lot of discussion around income, which rather surprised me. What we are actually trying to do is to deliver benefits to members in an appropriate way and meeting long-term objectives. It is almost like I feel we have chosen the right objective. The whole argument around income I think is actually one of the factors that you need to take into account when designing your flight path. I agree it is important and maybe a reference to it in the paper would be appropriate.

When setting your flight path, clearly income has to be relevant. It will dictate your timeframe, the timeframe over which you are willing to get there, and it may well dictate your investment strategy in the short-term and in the long-term. But I would say that it is a factor and not the centre of our universe.

We have deliberately stayed clear of making any views on whether it is right or wrong to be reliant on the sponsor. What the framework is trying to do is quantify that reliance. Once you have a well-articulated flight path, and you can see from where your future sources of return are coming, whether it is future improvements in interest rates, investment returns, or contributions, you can quite clearly see your reliance on that sponsoring employer. Having that quantified helps the decision making process and makes it clear for all stakeholders concerned.

I think that there is a tendency to accuse us of being very pro financial economics. Our rationale is this. If you have a well-defined target, and let us say your target is to be 100% invested in gilts, the price of doing that today is relevant.

We are not saying that your target should be to be 100% invested in gilts. Your long-term target may be to have an ongoing reliance on equity-type investments in the future, but that is your target. We are not saying that you should be 100% gilts invested. We are saying define your target well.

The market price today of doing what you want to achieve is relevant. Your next step then is to work out how you are going to get there. If you believe that future investment returns are going to be better than assumed, great, make allowance for that your flight path and also state that you are doing that.

Be aware of the risks. If you believe interest rates are going to improve over the long-term, great, make allowance for it and state that in your flight path. We are not saying that every scheme should be mark-to-market. What we are saying is know the value of the portfolio that you wish to hold because surely knowing what it is priced today is going to help you derive your flight path to get there.

We have been very clear not to go down the financial economics route, and similarly we have been very clear that we do not want to go down the traditional route. What we are trying to do is bring in a framework that encompasses both.

The final area worthy of some kind of comment is the macroeconomic perspective. In particular, there was a lot of talk about the labour agreement, the security mechanism around deferred pay, the decline of DB and the relevance of the social sciences, all of which I think are really good points and some of which already are topics of future research and some other areas that I know particularly around Solvency II that Dr Hatchett is also looking at.

I think that there is a challenge here. We have to be very clear about what our objectives are when advising stakeholders. The problem is that the stakeholder wishes to meet its objective in a predefined period of time. That is one objective. That is what this paper is helping them do.

There is a separate issue about a macroeconomic framework that I think really has to be for other stakeholders, namely, governmental departments, the profession as a whole, and I think that is certainly worthy of separate debate.

In conclusion, what we have tried to do is set out a framework to help stakeholders reached their target over a time period that is convenient to them, and also the other stakeholders concerned. We have tried to set out a framework where we can disclose and we welcome the suggestion that that should be stated in the statement of funding principles or some other document, so all parties concerned know exactly what pension funds are trying to achieve, the timeframe over which they expect to achieve it and how they intend to get there.

The Chairman: It just remains for me to express my own thanks and I am sure the thanks of all us to the authors, the opener and the closer and all those who participated in the discussion.

Subsequent to both meetings the authors sent a response as follows:

The authors would like to thank everybody who has taken the time to submit questions and comments on the paper. Those comments have been extremely informative and illuminating. As a result, the authors would like to make four observations to clarify and expand upon points already covered.

Stakeholders

A number of sound observations were made around the importance of stakeholders in the process of agreeing a well defined target. In particular, conflicts of interest were raised, which are extremely relevant to the challenge of agreeing a target.

From the authors’ experience, the issue of conflicts rarely arises during the process of agreeing a target. Usually, both the pension scheme trustees and sponsors can agree where they would like the scheme to be. The challenge usually arises in the determination of how the scheme should get to that target and, in particular, the balance between investment return and contribution.

One of the advantages of having a well defined target is that the debate between scheme trustees and sponsors can focus on the critical issues of investment and contribution strategy rather than on technical distractions, such as the assumptions used to derive the scheme's technical provisions under the current regulatory framework.

Factors when determining flight paths

A number of very valid comments were made regarding important factors to consider when setting flight paths. The authors did not cover all relevant factors to consider during the process, but would like to make it clear that issues such as sponsor covenant, income, benefit payments and other relevant factors should all be considered when setting the acceptable long-term target, the contribution and investment strategy, and the timeframe over which the target is expected to be reached.

Financial economics

The authors were slightly disappointed at the number of comments arguing against the merits of financial economics with the implication that the paper favoured financial economics over other approaches to pension scheme valuation. The authors did not state that financial economics should be used to value pension scheme obligations. Indeed, the authors do not believe that financial economics, or any other single methodology, should be applied blindly to pension scheme valuation given the funding approaches typically adopted in the U.K.

The authors argued that the price of the scheme's target portfolio at the outset of the flight plan is relevant as a starting point for determining a flight plan. From this point, however, the authors argued that any expected future improvements in asset prices or changes in the price of the target portfolio should be allowed for in the flight plan.

If the stakeholders of a given scheme in making decisions about their flight path believe that interest rates, for example, are likely to rise and reduce the value of the target portfolio, then the authors argued that this should be explicitly allowed for in the flight path. Indeed, the authors argued that not allowing for these improvements could give rise to “reckless prudence” and lead to excessive risk taking in the short-term and suboptimal decision making. In doing so, stakeholders should be made aware of the risks of their beliefs not coming to pass, and the consequences for different stakeholders in this case.

Effective decision making

There are always informational trade-offs when helping pension scheme trustees and sponsors make decisions regarding the pension scheme. The authors believe that it is the job of the actuary to strike a balance between providing enough information to make an informed decision, but not too much (or too complex) information that could lead to misunderstanding and poor decisions.

The authors’ proposed framework was designed to help decision making rather than hinder it. For avoidance of doubt, the authors would encourage users of this paper to consider what is appropriate for the pension scheme stakeholders concerned in their own particular circumstances, after considering the level of understanding amongst the stakeholders, and the ability of those stakeholders to make informed decisions.

Following on from this observation, the authors would like to reiterate their thanks for all of the comments made on the paper. The authors hope that the observations made in the paper and subsequent comments, and that the framework proposed will help stakeholders derive strategies for meeting their pension obligations.