The Chair (Mr J. S. R. Ritchie, F.F.A.): Our author, Mr Chris O'Brien, did an economics degree at the University of Cambridge. He worked at Royal Insurance, which became RSA, in Liverpool for 27 years. He was Appointed Actuary for Royal's life insurance companies and Chief Actuary of RSA's UK Life Division.
He has been President of the Manchester Actuarial Society. He moved to Nottingham in the year 2000 to become Director of the Centre for Risk and Insurance Studies. He has carried out research projects for the Actuarial Profession on general insurance reserving and on mortality perceptions as well as other academic research. He has been a member of numerous committees of the Profession and also a member of the working party on international accounting standards and life insurance.
Mr C. D. O'Brien, F.I.A. (introducing the paper): The financial services industry has advanced greatly over the past 20 years, as has the Actuarial Profession. An important factor that has enabled the Actuarial Profession to progress and contribute to these changes is its research activity. There is a tremendous amount of actuarial research carried out by both academics at universities and by practitioners. Often, though not always, they are working in working parties of the Profession. This is a considerable achievement. There is nothing comparable in the banking industry. But the research which is carried out will often pose new ideas and will challenge existing practices. My paper sets out challenges to existing practices and will stimulate a debate that may lead to changes to some of these existing practices.
The paper is about proprietary life insurance companies and the interests of with-profits policyholders and shareholders. While those interests may coincide, they do not always do so.
Firstly, considering equity, policyholders enter into a contract and the profits are divided 90:10 in most cases. So why do we need to be concerned about equity?
We do not expect policyholders to understand difficult contracts, for example, when an insurer has to change bonus rates or investment strategy. Discretion is important and vital for with-profits to operate on a sensible basis. However, there is a concern, that discretion may be exercised in a way that disadvantages policyholders, which is a particular concern of the FSA.
A working party of the Profession cited in the paper (Thompson et al., 2002) looked at what we mean by “fair treatment of policyholders”. Fairness is about honouring expectations, assurances and guarantees. So, legally, insurers should meet their obligations under contracts.
Where does the legal approach take us? If a case went to court, then the court would try to interpret the contract. The court would say, “The contract is incomplete. The policy document does not say much” and so the courts may often imply terms into that contract.
There are two different types of implied terms. In one type, implied terms are where the court would specify what the parties would have agreed at the outset if they had decided to document everything. These are really gap fillers: the parties agreed but they did not bother to write everything down. Equitable Life is a case in point.
There is another category of implied terms which are implied terms in law. These are terms that the courts may imply to all contracts of a specific type because of the nature of the contract. So, for example, cases where the courts have said, “Contracts between employers and employees or master/servant contracts implicitly involve the following type of terms.” In those cases the intention of the parties is irrelevant. So whether or not one party said, “We did not agree with that” or “We would not have written that into the contract in the first instance”, is not the point.
What the courts do is look at issues of social policy, fairness and justice, and where they feel it is necessary to have reference to those, to imply terms in law. They take account of the imbalance in the bargaining power of the two parties. So if there are two parties which have quite different bargaining power, then the courts are more likely to imply terms in law.
Implied terms in law can work for with-profits contracts. These are cases where there is an imbalance in the bargaining power, and the courts could intervene. If the courts do intervene, then they would say what is fair, and would make reference to the principles of fairness in the Unfair Terms in Consumer Contracts Regulations.
These terms, which apply to insurance contracts, have four requirements for a term to be unfair. First, that the term is not individually negotiated. Second, that there is a significant imbalance in the parties’ rights and obligations. Both of these conditions apply to with-profits contracts. The third requirement is that the term is ‘contrary to the requirements of good faith’. People may say, “That does not apply because the insurers are acting in good faith when they determine bonus rates”, for example. Lawyers would say that if the other two requirements are met, then that one is automatically met as well. So it does not add anything to the argument. The fourth requirement is that there has to be consumer detriment. If there is consumer detriment arising from the way that discretion is operated, then that is unfair.
If we assume that the first three requirements relating to individual negotiation, imbalance of rights and obligations and good faith are met the outstanding question is whether the way that insurers typically operate with-profits contracts leads to consumer detriment.
The with-profits concept can also work well for policyholders. The profits are divided in the ratio 90:10, so it is in the interests of both policyholders and shareholders to make good profits. This paper therefore concentrates on where the potential issues lie. In this context therefore fairness means that an insurer complies with its obligations under the contract, and also with any terms that would be implied by the courts. It does not seek to impose any terms that would be unfair under the regulations.
To apply this concept of fairness, though, a benchmark is needed to determine whether or not consumer detriment applies to with-profits contracts. Relevant here are three basic principles of with-profits contracts.
The first of these is that surplus is assets minus liabilities for the whole fund, not just a particular part of the fund. The second principle is that we accept that there can be intergenerational transfers of surplus. So one generation does not necessarily receive the whole of the surplus arising from the asset shares and stock market returns – it may be carried forward to another generation. The third principle is that at least 90% of distributed surplus is allocated to policyholders. That 90% has been criticised as both arbitrary and excessive, although it may not be excessive in current financial conditions.
The paper looks at some specific situations which, in the event, have been resolved. One example is that there have been firms that have issued with-profits products in recent years with, say, an allocation of less than 10% (say, 6%) of surplus to shareholders. In some of those cases the shareholders still took 10% of the distributed surplus by taking 4% from the inherited estate.
However, policyholders have a contingent interest in the whole of the fund, which includes the inherited estate. The above practice erodes the inherited estate by allocating some of it to the shareholders, a feature which the FSA finds unacceptable. It could also be deemed unacceptable in that the inherited estate is part of the surplus of the fund and may reflect intergenerational transfers from past policyholders, and we distribute the surplus 90:10. So the inherited estate should not be used in this way to allocate money to the shareholders.
There are some outstanding issues about fairness. Some relate to management actions. For instance, exercising discretion in the with-profits fund is important; but unfairness could arise from the way that discretion is exercised. In an interesting court case involving Paragon Finance and Nash, Nash complained that the interest rate he was being charged on his variable rate mortgage increased unfairly. Paragon Finance said that its costs had increased and the court accepted that an increase in charges caused by an increase in the cost of providing the mortgage was valid.
The FSA, in their report cited in the paper (Financial Services Authority (2005) Fairness of terms in consumer contracts: Statement of Good Practice), looked at cases where firms were able to review their premiums under reviewable rate policies and accepted that there may be instances where firms could validly increase their charges, provided that the increase in charge was proportionate and that the ability to review its charges was not used as an excuse suddenly to increase profit margins.
The conclusion is that management actions can be fine: they are often important and, indeed, vital. However, the firm should not use its discretion to favour shareholders over policyholders or to increase shareholder value over policyholder value. This is therefore an implied term of a with-profits contract.
There is also an issue about management services companies. The issue is that a management services company can be established and make a charge to the with-profits fund that includes a profit margin. Are customers thereby disadvantaged compared with the with-profits fund operated on a standard basis whereby asset shares are debited with the cost of providing the service without a profit margin? The circumstances in which charging a profit margin is acceptable are debatable but the view of the author is that if you establish a management services company then there should be an expectation of no consumer detriment.
Another issue that has arisen is about taxation of life companies and what is sometimes called “shareholders’ tax”, although it should not necessarily be construed that shareholders should pay this tax. There are issues where companies pay tax on their pension profits and additional tax on their life insurance business because the shareholder tax rate exceeds the policyholder tax rate. The question is: who should pay this tax? Should it be shareholders or could it be charged validly to asset shares or to the inherited estate?
My view is that it should be charged to the shareholders because tax is paid at the rate applicable to companies, not to individual policyholders, and the tax arises because there are shareholders and they are receiving their 10% share of surplus. The magazine Which? took Counsel's opinion on this and Counsel argued that the practice of charging the tax to the inherited estate was inconsistent with the 10% principle.
FSA rules, however, say that it is acceptable to pay the tax from the inherited estate if it is past practice and in the Principles and Practices of Financial Management (PPFM). This may be questioned because the fact that it is in the PPFM and may be past practice does not make it a term of the contract. The FSA has acknowledged that this is an issue which it will need to re-examine in the light of Solvency II.
However, independent of Solvency II, some firms have in the past deducted tax on pension profits from their asset shares. In 2004 the FSA introduced a new rule whereby shareholder tax cannot be deducted from asset shares. However, we know that there are some companies that charged the tax to asset shares prior to that. Consider a with-profits pension policy that is going to mature next year, and where the company deducted this tax from asset shares up to 2004. The company will almost certainly have marketed it on the basis that the policyholder will receive a tax-free return, yet the tax has been charged to asset shares. A policyholder, due to receive the policy proceeds next year would expect to receive a return that does not have any tax deductions, so the asset share should not reflect deductions for pensions profit tax.
The last section of the paper deals with reattributions of the inherited estate. Consider a firm with, say, assets of 1000, liabilities of 900 and an inherited estate of 100. In the reattribution the firm indicates to current policyholders that they are unlikely to receive a distribution from the inherited estate, so they are offered a ‘policyholder incentive payment’ under which they give up their rights to future distributions from the inherited estate. Current policyholders would want to know what they might expect from the inherited estate. If there was excess surplus in the fund, the policyholder would expect to receive 90% of it. It would be difficult for the policyholder to work out how much they would receive, because whether there is excess surplus or not depends on what the firm says about how much new business it intends to write and how prudent it wants to be.
Policyholders would expect 90% of the surplus if the fund were to close, but they cannot make a judgement about how likely it is that the fund would close, and it is in the firm's interest to say, “We are not going to close, so do not expect anything.”
In our example we might conclude that the 100 inherited estate was expected to be distributed, say, 27 to current policyholders and perhaps 36 to future policyholders, with shareholders receiving 10% of these two items. This would leave about 30 that was a residual, which was not expected to be distributed at all. Therefore, the firm can offer current policyholders an incentive payment of 35, which given the improvement over 27 they may accept. In this case 65 goes to the shareholders.
This is not consistent with the three with-profits principles. The inherited estate is surplus in which the policyholders have an interest. It reflects intergenerational transfers of surplus in which policyholders have a 90% interest (not 35%).
It is true that if reattribution goes ahead then the policyholders gain in terms of liquidity and certainty. So 90% interest in the future distributions does mean slightly less than 90% of the cash inherited estate that the policyholders’ payment implies. So a reattribution can unlock value because it is better for the firm to bear the uncertainty and illiquidity than it is for the policyholders to bear them. The difficulty is in achieving a reattribution where there are high costs and a tax disadvantage and all the parties’ expectations cannot be met. Regulation is needed because there is an imbalance between the parties concerned.
To conclude, the paper has offered a suggestion as to what fairness means expressed in terms of legal principles that may make it easier to apply in practice. It has also proposed three with-profits principles that can provide the commercial context for applying this definition of fairness.
It is hoped that the FSA will be interested in the principles and the legal issues so that it can regulate consistently with what I have suggested.
Mr R. D. Muckart, F.F.A. (opening the discussion): This paper is not without its critics and makes some controversial statements. However that is exactly why it is worthy of discussion. Too often as a profession we are reticent about speaking our minds, concerned that we will do ourselves down by arguing in public and by doing so reduce our professional image. Lawyers thrive on adversarial contests and develop the law through case history. We should be prepared to do the same, especially in an area where there is likely to be no single right answer.
I am going to focus from a non-technical viewpoint on the four Rs: reasonableness, regulation, reattribution, and risk.
Firstly, I question whether the 90:10 split is reasonable. Although this has been the basis of sharing profits between with-profits policyholders and shareholders for a very long time, should this be so in all cases? As an investment specialist, I would look at this from the shareholder's perspective on the basis of a return on capital employed. The strength of the assets relative to the liabilities will vary from fund to fund, and so the amount of capital required will vary. When companies were mutual they had the inherited estate as a buffer that allowed them to follow riskier long-term investment strategies in the knowledge that the odd bad year could be absorbed and the assets in the estate replaced in the good years. For a proprietary company today this would be an unacceptable risk and consequently would result in less risky investment strategies being pursued. So the opportunity to profit is reduced leading to lower bonus being paid. The policyholder might then reasonably claim that the 90:10 distribution should be moved in his favour to offset that “loss”.
Let me turn now to the issue of regulation and the views of the FSA. The FSA's approach is disappointing in that it has become prescriptive and short-term focused, blunting a number of the tools that were at our disposal. We should not hide behind the various documents that now have to be given to policyholders. We should be much more open in explaining how with-profits products work. This is especially important where the policy has been in force for some time and was probably issued prior to the latest paperwork requirements.
The third R – reattribution – should be left to the specialists, although a general comment may be appropriate. Were we looking at an industrial company or a fund management business, any company taking over would have full use of the assets and would be likely to use them in a way that benefits all parties and certainly looks after customers. The rules around with-profits funds, the use of inherited estates and how group companies might interact, seem to be against the free market or at least a constraint on it.
Finally, we come to R for risk. Assets and liabilities should be managed in an appropriate manner to reflect the nature of the product and the policyholder's attitude to risk. It is not acceptable for a third party to move the goal posts in the middle of the contract. This happened to many with-profits policyholders, particularly endowment policyholders, in the first half of the last decade. All of a sudden sums assured and attached bonuses (the guaranteed payments at maturity) had to be covered by appropriate investments which tended to be fixed interest in nature. The ability to prosper from equity returns was gone, and the markets had lurched downward only to recover slowly.
I found some of the arguments in section 5.3, treating management service companies, circuitous. A board of directors would always be expected to look at making best use of available resources. However, the nature of contracts and policies means that there is usually a duration mismatch. Consequently, one must balance risk, expense saving and opportunity cost. The asymmetry of knowledge means that the company knows more than policyholders but it can do much to reassure them that they will not be worse off. Using a managed service company can be advantageous, perhaps more so if the fund owns a share of that company so that it can gain from the long-term benefits that the group believes will accrue from it.
In section 5.4 the subjects of tax changes and new business come into focus. I found it strange in paragraphs 5.4.15 and 5.4.16 that the FSA would err on the side of helping shareholders rather than protecting policyholders if the basis of taxation changes. If policyholders had been subject to a new charge, would the FSA have levelled the playing field in a similar way? Or perhaps I have misunderstood this.
I am not sure why anyone would deliberately write new business in the expectation of making a loss, although I could put it in the context of the supermarket and a loss leader to encourage people in the door before selling them other products at good margins from other parts of the business. But surely those governing the business would be wary of such a practice.
Section 6.2 looks at the source of inherited estates and possible implications for fairness. As an outsider, the injection of capital into an inherited estate is difficult to envisage. Furthermore, it is capital employed and as such needs a return. It covers guarantees or strengthens balance sheets and so should receive a “premium” for the service rather than a share of profits.
The whole subject of the inherited estate seems to be causing much debate, but should have been sorted out at the time the with-profits fund transferred to the proprietary company. In giving up mutual status, the policyholders should have received appropriate recompense for this change. Having started work at a time when inherited estates passed from generation to generation and reflected the handing on of sound financial backing, I can understand why the current generation of policyholders should not have it distributed to them. However, if a fund is closed to new business and therefore in run-off, who should benefit from the estate? There is no single fair answer, but then there often never is in a takeover.
It seems to me that equity between policyholders and shareholders is an art rather than a science. If a company can exceed policyholders’ expectations, then its shareholders deserve to reap the rewards. Regulation, risk and reattribution are being over-thought, and we should revert to reasonableness as the basis for solving problems.
Mr A. M. Eastwood, F.F.A.: The existence of implied terms was new to me when the House of Lords opined on the Equitable Life case. It is undesirable for there to be so much that is subject to interpretation and therefore uncertain for the company. There may be a big incentive as a result to have longer contracts; perhaps that is one of the reasons why with-profits business is waning.
One of the author's three basic with-profits principles is that the surplus is calculated as the excess of assets over liabilities of the with-profits fund as a whole. It is suggested that this has been so from the beginnings of with-profits business. I suspect that most of the arguments in support of this point revolve around the inherited estate and its sources which are lost in the mists of time.
The excess of assets over liabilities depends on how it is measured. It would have been determined historically in the context of something close to the current regulations; namely, that the insurance company would have had the right to place a value on the assets, such as book value that is lower than the market value of the assets, and it would have used a prudent valuation of the guaranteed liabilities. It is not certain that this necessarily affects demutualised companies, where there is a scheme that sets everything out clearly. However, if the genesis of the inherited estate is lost, it is possible that assets were added to the fund many decades ago, and that the transfer was recorded using the then book value of the assets.
The 90:10 principle discussed in the paper is probably more relevant to companies that have gone through demutualisation or to more modern with-profits business. A lot of business is not written on a 90:10 basis these days. 100:0 might be better than 90:10; but it does lead to a somewhat more lopsided set of interests – a point that is probably more relevant to some of the rules that the FSA is proposing than to the conclusions in the paper.
We should bear in mind the original customer proposition. What the customer bought, possibly many decades ago, was a right to certain benefits and the right to share in the profits of the fund or a company. That is not the same as an asset share. The asset share is a useful tool that was devised as a fair way to share out the profits that were derived. However, policies were not sold, certainly up until the mid-1980s, on the basis of an asset share as we know it today.
Turning to the inherited estate and the implications of contingent rights: any rights of customers to the inherited estate are contingent, and we need to be careful in this territory. If a consumer goods retailer, for example, commits as a sales incentive that all purchases of a product are entered into free prize draw, then every purchaser of that product after the first one is adversely affecting the contingent interest of the previous purchasers in the prize draw. It is not any detriment in a contingent interest that is objectionable or unfair; but, stated in the paper in paragraph 4.2.3, it is the manipulation of business to allow 100% of the estate to come within the 10% part of the 90:10 gate that is unfair.
It should be remembered that the contingent interest in an inherited estate is contingent. It only becomes an entitlement if it is not required for its original raison d’être, to provide capital to support the writing of with-profits business.
Reading through CP 11/05, one might be drawn into concluding that if there is a significant inherited estate then one should stop writing new business because by writing the new business the interests of the existing customers are being diluted. We need to keep the contingent interest in perspective.
The issue of management service companies is a difficult area. The proposed FSA rule goes too far in effectively banning profits being made by management services companies on charges levied on with-profits funds. It will depend on the circumstances of the fund concerned, but the proposed rule will give the shareholder insufficient incentive to improve the efficiency of administration of with-profits policies. This is particularly the case for 100:0 business. Certainly, if the fund is protected from increases in unit cost by the terms of a court-approved scheme, and so is not bearing the expense risk, the proposal goes too far. Policyholders would be better off with 50% of a potential expense saving than with 100% of nothing.
FSA rules should not prevent better solutions to problems, especially for court-approved versions, or prevent companies from operating effectively.
Section 5.5 discusses the appropriate discount rate to use when determining the shareholders’ share of the cost of reversionary bonus. As suggested this is probably not a major point these days. However, in a demutualisation, an amount may be paid into the new with-profits fund representing the then value of shareholder interest in future payments anticipated in a with-profits fund. The amount paid reflects the new structure, and will typically be determined on embedded value principles. Fairness here means that the starting point should be consistency with that methodology rather than necessarily moving to the risk-free rate that is proposed elsewhere in the paper.
The same argument would, in principle, extend to other aspects of with-profits operation, including the allocation of shareholder tax, although it seems improbable that there is any court-approved scheme that says that the with-profits fund should pick up the tab for shareholder tax.
Paragraph 6.7, on estate attribution, is probably the most contentious and interesting part of the debate. The interests in a with-profits fund should be seen in a series of stages. First are the core rights and expectations that the policyholder should have by virtue of the contract between him and the insurance company as he might have understood it if properly advised at the time that he entered into the contract. This should not be compromised in any way. The inherited estate is capital that is there to support the writing of with-profits business or the continued operation of with-profits business. Existing policyholders will typically have benefited from its existence and they should reasonably expect subsequent generations to benefit before their own contingent rights lead to a windfall. Using some of the inherited estate should certainly come before any moving of goalposts, which represent core rights and expectations.
In a declining fund the expectation would be for the inherited estate to be invested in hedges to a greater extent than in a continuing open fund. It is more likely to be in the interests of a declining fund to accept a greater cost than it is for a fund that is open and can continue with intergenerational transfers. There will, nevertheless, come a time when a fund is contracting to such an extent that the vast majority of risks that can be hedged cost-effectively have been hedged and there is a decision as to what to do with the rest of the fund. The policyholders’ residual right to the rest arises because of the existence of ring-fencing protections, arising from FSA regulation or from court-approved schemes. In that context, an estate attribution may be necessary in the absence of a demutualisation or scheme of transfer. In other circumstances the residual inherited estate should be spent on policyholders in accordance with the author's with-profits principles. The tricky task is trying to do so as fairly as possible and in a way that does not leave it all to the last person out.
Quite how the courts would interpret the related issues is uncertain. It seems likely that the usual test of reasonableness would apply. There will be a number of reasonable courses of action. In deciding on what is reasonable clearly there should not be any customer detriment and it is likely that legal advice will be necessary.
The Chair: I wish to ask a few questions and make a number of points.
It is mentioned in paragraph 1.3 of the paper how many life insurance firms have demutualised so that proprietary firms now dominate the market. But was it not also the case that mutual offices could have equity problems? Equity problems arising, for example, between with-profits and non-profits policyholders, between different generations of with-profits policyholders, or between policyholders and the management team and employees of the company? In the last example – between policyholders, management and employees – there is the issue of the proper allowance for the cost of a defined benefit pension scheme for the staff. Many of these fairness and equity issues can arise as much in a mutual office as in a proprietary one.
The paper refers to the document, “The Principles and Practices of Financial Management” (PPFM) from 2004. Given the PPFM's complexity, even for an actuary, it seems unlikely that with-profits policyholders have read and understood them.
Paragraph 2.4.5, covers the declining number of insurers offering with-profits business and the market becoming increasingly concentrated with many funds now closed to new business. So the further you go into run-off, the greater the issues of equity seem to become from a policyholder's point of view for the declining number of with-profits policyholders that remain. But if the office itself is continuing to thrive, then perhaps these same issues become less important from a shareholder's perspective because of the other business that is going on.
I was interested that the first explicit mention of Equitable Life was on paragraph 2.4.10, although I do take the point that it was not a proprietary office so it did not have shareholders. In terms of fairness, estate and bonuses over the last decade, Equitable Life is the spectre of what can happen if we do not get this correct.
Paragraph 5.1.5 focuses on how legitimate increases in the cost of providing a financial service may be a valid reason for a firm exercising its discretion to pass on charges, provided the increases are proportionate. However, a company may delay in increasing charges until a large increase is necessary, and the large increase can then appear harsh to policyholders and unacceptable to the FSA.
Mr C. B. Russell, F.F.A.: I would like to focus on the question of whether part of the tax at the shareholder rate can be fairly charged to policyholders, as I have wrestled with this for many years. The author is certain that the tax should not be charged to the policyholder but I believe other conclusions are possible.
Paragraph 5.4.2 reminds us that before 1989 there was only one tax rate applicable to basic life assurance business, so the issue did not arise. This was “pegged relief” at 7s 6d in pre-decimal currency, which was the policyholder rate, although the term was not then used. The higher income tax rates, and then corporation tax, applied to the amount that was not regarded as the policyholders’ in tax terms.
I agree with the author that it seems absurd to say that you cannot charge certain tax partly to policyholders, but you can charge it to the estate. The fund is the fund and regarding the estate as something separate can be useful but not in this instance. If tax is charged to the estate, something is lost to policyholders.
I would answer the opener who asked whether, if the policyholders’ rate of tax went up, shareholders would suffer a share that they would, and they did. When tax credits no longer became payable on equity dividends, profits must have reduced as a result, and shareholders undoubtedly would have suffered 10% of the additional tax cost. So a case can be made that policies were entered into in a situation where there had historically been dual tax rates and yet net of tax surplus was shared on the 10% rule.
Those of us who know life office tax computations might regard the number that comes out at the end as somewhat random. It does not seem obvious that this random outcome should not be suffered in the ratio of 90:10.
Mr R. K. Sloan, F.F.A.: I was interested by the Chairman's comments because he widened the discussion from the more academic aspect of equity between policyholders and shareholders to the question of equity between generations of policyholder and different classes of policyholder.
I am a believer in with-profits. It provided a great service to members of the public over many years by giving them a stake in real assets in which they otherwise would not have invested. That was partly, of course, because of the opacity of the contract, and they perhaps were not fully aware of what they were entering into. Lessons can be drawn from that in terms of, for example, the new NEST plan and what default funds it should offer. The with-profits concept should still have a place and a role to play.
This discussion could be widened to include ways in which actuaries could improve the effectiveness of with-profits contracts, albeit that there are few of these still remaining.
Let me focus on the equity between different generations of policyholder and how this is achieved. We see a variety of different ways in which with-profits bonuses are distributed. Some methods are archaic in declaring a rate of bonus on the prospective sum assured under the contract, which bears no obvious relationship to where the policy has come from captured in the build up of its asset share. That was an unhelpful approach and, of course, it was replaced by the newer type of contract, unitised with-profits.
Interestingly, or ironically, one company that tried to address that issue was Equitable Life, which declared a rate of total return each year, but unfortunately got it badly wrong. The way they declared their bonuses was manifestly unfair between even the close generations of policyholders. People who joined recently were being treated in a somewhat arbitrary and inconsistent fashion. If ways could be found of simplifying the way bonuses are declared – perhaps not even calling them “bonuses” but “returns” – and communicating them better to policyholders, then that would be beneficial.
Consider the method of declaring bonuses on an endowment policy, which is often used to repay a loan or mortgage at the end of a set period. It is to some extent academic what the rate of bonus is on the way through the policy term, other than that policyholders may like to have some indication as to where the policy is going. But there could be better ways of expressing this. For example, for some 20 years now we have had with-profits annuities where the rate of bonus is a current issue, because it directly affects what policyholders receive in annuity payments in each year that the policy runs.
I have my whole pension in a with-profits annuity. What my contract does is to declare a temporary terminal bonus, as it were, in addition to modest regular bonuses. The temporary bonus lasts for only a year and is then reviewed in the light of prevailing financial conditions.
There have been some interesting ways in which new bonus methods have been developed to make the declared returns on with-profits policies more user-friendly.
On a small point, the word “shareholders” could also be widened to include “potential shareholders”, as in when dealing with a mutual that is contemplating demutualisation. I was involved with Standard Life when it was in this situation at the beginning of this millennium. I was in favour of demutualising because I felt that policyholders’ money was being put into risky situations, such as expanding into overseas markets. Interestingly, Standard Life had originally been a proprietary company and had demutualised only in the 1920s. Scottish Life, likewise, changed from being proprietary to mutual fairly recently, so life companies have a history of both mutualising and demutualising.
In conclusion, if some of the lessons drawn from the analysis carried out by the author on the relationship between policyholders and shareholders could be transferred to the wider issue of general equity and fairness in order to improve the attractiveness of with-profits contracts, then that would be of great service to the public.
Mr D. G. Robinson, F.F.A.: I particularly welcome the fact that the paper comes from the world of academia, as the link between the Profession and universities is increasingly important. I welcome the lack of formulae in the paper. It makes a refreshing change to have a paper that focuses on issues of principle and does not require a deep understanding of mathematics.
I am going to use this discussion as an opportunity to talk about the treatment of tax in the calculation of asset shares for with-profits contracts, and specifically the impact of writing protection insurance business, such as term insurance and critical illness insurance.
Life and annuity business is taxed on the so-called “I minus E” basis. To illustrate the issue, consider an example of a company that has investment income of 100 and expenses of 50 for its with-profits life business. It will pay tax on 50 (100 less 50) at a rate of, say, 20%, so the asset share will be credited with net investment income of 90 (i.e. 100 less 10).
Suppose now that the company writes some protection business, and incurs extra expenses of, say, 20. By its nature, protection business does not generate much by way of investment income so this can be disregarded. If we now look at the company's revised tax position, income remains 100, the expenses have now increased to 70 and tax reduces to 6, so the net investment income increases from 90 to 94. One might reasonably imagine that the impact on the asset share would be for the increased net investment income of 94 to be credited to the asset share. But that is not what tends to happen in reality. Instead, market practice seems to be for the tax relief on the protection expenses not to be attributed to the investment income, but instead to be fed back into the pricing of the protection policies so as to reduce the premium rates (that is, pricing these products on net E rather than gross E).
In these circumstances, I question whether or not these with-profits policyholders are being treated fairly. Arguably, they are not, since their asset shares ought to reflect the actual net of tax investment income and this is clearly not the case – there is a shortfall of 4. On the other hand, it could be argued that they are being treated fairly as they are in the identical position to what they would have been in had the extra protection insurance business not been written. At the same time, the company is taking advantage of this investment income to enable it to price its products with expenses net of tax relief and so make extra profits for its shareholders. In these circumstances, should the with-profits policyholders not also benefit in some way?
The issue is whether or not current market practice is fair. I can see both sides of the argument.
Mr D. Morrison, F.F.A. (closing): As a with-profits actuary, I am conscious of the difficult territory that actuaries can get into when having to assess equity in managing with-profits business. “Would this be a fair way to exercise discretion?” is how we often frame our question. This paper uses the term “discretion” quite sparingly, which is probably a good thing. If an office has an obligation to be fair, and it surely does have, then I do not see that it can have discretion to act unfairly, in which case the question should not be whether discretion is being exercised fairly, but whether or not the firm really has the discretion to do what is proposed. In other words, we should be trying to define the limits of discretion. I have read the paper as an attempt to do that.
Some of the issues that arise in managing with-profits business seem to call for judgements that are hardly actuarial in nature and they stray into territory where a lawyer would be more comfortable. Mr Eastwood alluded to this too, and for the need for legal advice.
A with-profits actuary is required by regulations to express his opinion on whether or not policyholders’ interests have been properly taken into account. As an actuary with experience of life assurance, I ought to be well-placed to understand the technical issues and, it is to be hoped, explain them to a board. But, am I, by virtue of being an actuary, better qualified than anybody else to make the ultimate judgement about whether or not what is proposed is fair?
I find it particularly valuable to have a paper that tries to define a clear framework that takes appropriate account of law, of regulations and of actuarial considerations for assessing equity between with-profits policyholders and shareholders in a proprietary office.
The author has noted that there is surprisingly little on this in the existing literature. The clarity of the presentation of the arguments in the paper makes this accessible to all, not just to actuaries who may be experts on with-profits. I hope other actuaries who are not with-profits experts will join this debate.
Without a framework, such as is proposed here, we can find ourselves taking unduly subjective and sometimes quite emotional approaches to assessing what is or is not fair. Trying to explain to a policyholder why it is fair that he is receiving less than his asset share is not unlike some issues arising in my position as the parent of two children. If I have 10 sweets to share between them and I give 6 to my son and 4 to my daughter, it is beyond any shadow of actuarial doubt that there will be an accusation of unfairness. I certainly would not dare to resort to the argument that, because I have given 6 to my son, I have only 4 left and therefore it is fair that my daughter receives only 4.
I suppose the application of the paper to the problem would be limited to denying me any more than one of the 10 sweets for myself. Sharing the sweets between the children may be closer to the concerns of Mr Ritchie and Mr Sloan, who are keen to widen this debate to issues of equity between different groups and generations of policyholders in a with-profits office.
The paper notes that one of the reasons for attention being given increasingly to equity between policyholders and shareholders is that many demutualisations have occurred over the years. Something the paper does not explore, or not in great detail, is how a past demutualisation may bear on the assessment of what is fair now. This was alluded to by Mr Richard Muckart and by Mr Adrian Eastwood in their contributions. It may well be a material consideration.
Consider the extent to which it is reasonable for equity backing ratios or smoothing to take account of the financial strength of a with-profits fund. If the with-profits policies were written when the firm was a mutual, then the investment strategy and scope for smoothing would originally have been dependent on financial strength, because the office probably had no source of capital outside the with-profits fund. There seems to be a stronger argument that it is fair for that to continue to be the case in the proprietary office. Indeed, some demutualisation schemes provide explicitly for such continuity of practice.
Also, a demutualisation scheme is designed to maximise value for the members, subject to maintaining their expectations as policyholders. If before demutualisation they had no specific entitlement in respect of an estate, then there will have been freedom to specify whatever use of the remaining estate after demutualisation maximises the total demutualisation value. That might, perhaps theoretically, include allowing the estate to be used to pay shareholder tax. Mr Eastwood made the point that he does not think any schemes do allow that. But if it were the case then the directors would have quite a strong argument against the suggestion made in the paper that they should revisit such an arrangement.
Aside from the specific circumstances, the paper made a compellingly simple case for shareholders paying for shareholder tax. I did not get the impression from other contributions that there was any great appetite for arguing against it.
An issue that was emphasised by the author is the assessment of management actions that tilt the balance of value from policyholders to shareholders, which the paper argues are unfair. We need to be careful here about the particular comparisons that we make.
If we consider a reduction in the equity backing ratio that is made in response to financial stress, then it is apparent that immediately after that reduction the shareholder value is greater than it was immediately before the reduction. That does not mean that the action was necessarily unfair, and the paper is not trying to argue that it is. What we need to consider is whether the reduction is consistent with the management strategy of the fund as understood when the policyholders invested in it, since it is the long-term strategy that sets the balance between shareholder and policyholder value. Determining that in practice is challenging. Investment strategies in with-profits funds are a sensitive topic, as alluded to by Mr Muckart in his opening remarks, and it is difficult to establish a fully objective basis for assessing whether or not they remain fair.
We cannot have a discussion of with-profits nowadays without some mention of the published Principles and Practices of Financial Management (PPFM). I agree with the sentiment in the paper that an unfair action does not become fair just because it has been described in the PPFM. The great value of the PPFM is in helping to prevent the arbitrariness or capriciousness in managing with-profits that, the paper rightly argues, can be unfair. But they are not documents that are much read by policyholders, or indeed by anybody else.
I was depressed by the author's comment in introducing his paper that we do not expect policyholders to understand with-profits. I agree with Mr Muckart who, in his opening remarks, emphasised the need to continue to strive for transparency and for explanation.
The paper's discussion of the proper involvement in strategic investments of a with-profits fund is clear. It would be quite difficult to disagree in principle with the conclusions. I wonder how confident you could ever be that a strategic investment was on fully commercial terms and at least as beneficial as any alternative investment. Commercial terms can usually only be determined if they are available in the market; and if they are available in the market, why does the company choose to use the with-profits fund's capital?
We need to be similarly careful in the approach to management service companies. I have some sympathy with what the FSA is seeking to achieve in its proposed changes here, perhaps slightly more sympathetic than Mr Eastwood was in his comments. The main reason for a with-profits fund being able to benefit from contracting out services to an external company arises from economies of scale within the service company. These may be less likely within the insurer's own group. I wonder how likely it is that, if the risks and rewards of the servicing arrangement are attractive to shareholders of the group, they are genuinely unattractive to the with-profits fund.
Finally, on reattributions, the point has been made that this is the most contentious area. Reattributions are complicated and few and far between. I work for an office which has had a relatively recent demutualisation, so the entitlements of the estate are clearly defined. I was tempted to skip lightly over this section of the paper. If I had done so I would have missed a thoughtful and a clearly presented analysis, although I agree with the observation that Mr Eastwood made that you do need to be careful about the origins of an estate in coming to conclusions.
I can confine myself to one slightly critical comment in this section of the paper because I found it otherwise to be well argued. In establishing a basis for assessing equity in with-profits business, we need to decide what part FSA rules play. Either equity depends on some more fundamental considerations, in which case we need to assess on those considerations whether or not the rules are fair; or, alternatively, we start by accepting the rules and consider only the fairness of the choices that are permitted by them.
The author wanted to have his cake and to eat it too. Paragraph 6.5.9 relies on COBS rules on distribution of excess surplus to assert that policyholders have a contingent claim on the estate. Elsewhere he seeks to judge the fairness of the rules. Indeed, in paragraph 6.11.5 he states that the paper judges the fairness of reattribution without taking FSA rules as given.
Also on the subject of FSA rules, both Mr Eastwood and Mr Sloan did not balk at questioning our current slavish devotion to asset shares, which is something which I am sure the FSA would regard as near heretical.
In conclusion, it is enormously helpful to have a paper that makes such clear arguments and reaches unequivocal conclusions on a topic on which there is as wide a range of opinion as any that comes before us as a Profession. I would add my congratulations to Mr O'Brien for this paper
Mr O'Brien (responding): On the points about management services companies, maybe the FSA rule has gone too far. The key thing is: does a management service company help policyholders or not? It might, though in some cases it is unlikely.
On tax, I appreciate there are some earlier tax rules that are relevant and could be material. I do not pretend to answer some of the tricky issues on another class of business. There are court-approved schemes for transfers of business between proprietary companies that permit the inherited estate to continue to bear shareholders’ tax, which is unfortunate.
On management actions, I take the issue about what happens the day after you have undertaken management action, and more work is needed to apply the principle. Can we be happy about all management actions? What is the relative impact on shareholders and policyholders? Some more work is needed on these.
On strategic investments, greater transparency would be appropriate. Firms, if they are happy about strategic investments benefiting policyholders, should tell policyholders what they are. Is that a problem?
On reattributions, I appreciate the comment from Mr Morrison. In one place I rely upon the FSA rule saying that policyholders have a contingent interest. I could derive the same conclusion by referring to my three basic with-profits principles – that might have been a more robust way to derive the conclusion.
A number of speakers referred to equity between different groups of policyholders. That is a fair comment and maybe it will lead to another paper on how one might apply legal principles to determine such equity.
The Chair: We have had a good discussion, and I would like you to join me in thanking the author, the opener and closer, and the contributors from the floor.
