Non-guaranteed benefits: Performance-linkage and discretionary benefits IAA survey on non-guaranteed benefits (draft) Table of contents Introduction

Transcription

1 Table of contents Introduction Background (the economics and purpose of non-guaranteed benefits) Features of non-guaranteed benefits in contracts provided by insurers A. Guaranteed and uncertain benefits B. Performance-linked benefits C. Participation benefits Current accounting approaches worldwide A. Accounting for irrevocably allocated dividends B. Accounting for future dividends C. Accounting for revocably allocated dividends (terminal dividends) Considerations for phase 2 application Introduction Non-guaranteed features in both insurance and investment contracts offered by insurers have been difficult to understand in terms of the extent of their guarantees and the fact that they are an integral part of those contracts, as well as their use as risk management and competitive tools. They vary tremendously by national market, in part due the effect of state regulation. The approaches taken can be complex, particularly when the principles followed are based on regulatory practice rather than on written law. For example, the German Supreme Court decided that German regulations regarding participating business are so complex and difficult, that they cannot be explained to average policyholders in detail, granting exemption from the requirement that policy conditions be transparent, even though transparency is usually considered to be an important public policy objective. Accounting for participating business, especially if contracts restrict the insurer s share in surplus, has proven even more difficult and varied worldwide. Robert Posnak in Ernst & Ernst s GAAP for Stock Life Companies (1974) indicated, As will be seen, restrictions on the earnings of participating business present the accountant with some of the most challenging problems in life insurance accounting. Applying an accounting approach (e.g., USGAAP, embedded values) to participation features in countries for which applicable guidance was not designed has created significant headaches for and discussions within many insurers. In several instances, insurance entities have been closed down or sold by investors who were misled by inappropriate or inadequate information about the effects of participation features resulting, in part from improper application of accounting guidance for participation features (e.g., application of US GAAP and embedded values to German participating business might not have correctly reflected required distribution restrictions, limiting owners profitability to less than that that expected). The relatively straightforward wording of IFRS 4 regarding Discretionary Participation Features has also led to considerable discussion, including significant consultancy efforts, concerning its application in countries such as Australia, Belgium, France, Germany, the Netherlands and the UK. Since US GAAP standards were not written with non-us products and practices in mind, it should not be surprising that its application in other countries has been difficult. As a result, principles-based guidance may be most appropriate, with application based on the underlying economic features of the participating business rather than the details of a particular type of contract. The first experience with IFRS 4 has demonstrated that this has not been fully achieved by IFRS 4, as it does not consider the underlying economics of participating features in all cases. 1

2 A key problem is the wide variety of non-guaranteed features, be they fully or significantly constrained (by contractual or legal restrictions) or of a fully discretionary nature. Even if fully discretionary, other considerations such as competitive pressures, historical precedent (industry, company, product or market) or policyholder reasonable expectations resulting from sales practices can lead to the conclusion that such discretion may be seriously limited and depends more on experience to which policy performance is linked or related to and market movements, rather than individual company discretion. To ignore reasonable expectations of such features can lead to misleading timing of stockholder or entity profits if the underlying economics are ignored. Direct and indirect performance-linkage are often the principle mechanism that allows insurers to provide insurance coverage at guaranteed terms over long periods (sometimes measured in terms of decades) in a reasonable and competitive manner without undue risk of bankruptcy. Such features often form the economic basis for a great number of products offered by insurers and in many insurance markets they are the most significant contract feature. Several forms of longterm life, annuity and health insurance contracts would not be possible without non-guaranteed or participation features. In many insurance markets, policyholders demand such contracts. Smoothing of results has in the past often resulted in enhanced fairness, reliability and predictability of outcome, and is viewed by some as being the main advantage of current participation features. This demonstrates that proper treatment of non-guaranteed and participation features will be key to the long-term success of the IFRS Phase 2 accounting guidance in many countries. In particular, since insurer s profits generated from participating business are mainly affected by the type and application of participation, their characteristics should be appropriately recognized. 2

3 Background (the economics and purpose of non-guaranteed benefits) The application of direct or indirect performance-linkage represents a form of partial or complete re-transfer of risk back to the policyholder. The extent of re-transference may either be direct (e.g., via a unit-linked contract) or indirect (through a contract with non-guaranteed elements set at the discretion of the insurer which reflects a share of the risk). However, the fundamental feature of re-transference exists for all such contracts. The economic consequence of transferring and re-transferring insurance and financial risk is that the insurer can mitigate all or part of its risks by pooling and re-transferring at least part of the remaining risk of pool adversity or failure (caused, for example, by a highly unlikely excessive amount of claims) to policyholders. In this instance, policyholders gain from pooling and share in the risk of pool failure. Pooling of risks is a primary method of managing the primary insurance risks of an insurer that reduces the effects of random deviation, which can constitute the main economic purpose of insurance. It is often less expensive for all participants to bear the risk of pool adversity or failure proportionally than to cover the risk in the capital markets through stock issuance. This transfer and re-transfer of risk forms the basis of mutuality. The original reason underlying the introduction of such forms of re-transfer emphasized the retransfer of insurance risk, while re-transfer of financial risk later became a major component in many cases. Performance-linkage developed primarily in life insurance because of the need for this coverage over an extended period of time while the health of the insured can only be validated at the time of issue. Since there is a significant risk that an insured s health condition will become impaired during the contract s coverage period, an alternative life or health insurance contract would either not be available to that person or only available at an increased cost. Policyholders are attracted to the long-term guarantee of insurability offered by the insurer in many contracts (i.e., the right to continue such insurance coverage with the insurer without further medical examination of insurability). As a consequence, if the insured s health condition becomes or is expected to become impaired policyholders who purchase extended life or health coverage tend to continue that contract for its term. Many policyholders are only willing to enter into such a contract if the insurer grants valuable guarantees (e.g., guaranteed benefits or a guarantee of a maximum cost for insurance over the entire life-time of the policy). Since that lifetime is often several decades (typical expected lifetime of permanent life insurance contracts can in some instances be more than 20 years), if the entity is to be soundly run financially, the guarantees need to be set on a conservative basis. For instance, premiums for most traditional participating business can be viewed as being overpriced in relation to best estimate or even moderately adverse assumptions, and usually produce unavoidably large amounts of surplus. It may not be appropriate to consider such surplus as profit in the traditional sense, as if the anticipated experience occurs, it is expected that a corresponding dividend (bonus) will be paid to the policyholder through the performance-linkage feature. This can be viewed as a two-step pricing process first the initial setting of the premium, followed by an annual retrospective pricing based on achieved performance. Even five years after the introduction of this type of life insurance in 1762, surplus was partly refunded to policyholders, not simply from benevolence or from entrepreneurial behavior of the insurer, but rather as the result of conservatively priced premiums that reflected margins associated with long term guarantees. Today, such an approach only exists in the insurance business, through the use of relatively complex performance-linked approaches, requiring detailed experience and surplus analysis and allocation techniques. Other financial industries, such as banks, have not used such expensive 3

4 approaches, since features such as life-long insurability guarantees are not provided. In contrast, practically all other industries use a one-step pricing process. Only the part of the insurance industry that provides long-term business uses a two-step pricing process where the initial premium or transaction price is treated as preliminary, with a subsequent true-up to actual experience and expected future conditions. This two step-process has evolved for a number of reasons, but primarily in reaction to the need for spreading risk and responding to competition, in some cases providing a pooling mechanism to spread risks among policyholders and in others to provide a mechanism to effectively retransfer certain elements of risk back to the policyholder. These participation features have been used by most types of insurers, including those offering short and long duration insurance contracts, investment contracts and those providing life, nonlife and health insurance coverage. Non-life insurers have included premium adjustment clauses, both on a retroactive basis to be able to return excess premiums if not needed to cover losses and on a prospective basis applied to current premiums to reflect historical loss experience, and in some cases even as additional assessments to cover pool failure or premium inadequacies. They may be offered for a single large risk with no insurance being transferred, typically covering large numbers of homogenous and independent risks (e.g., an insurer can manage car fleet coverage for the owner of the fleet, with a primary objective being to improve risk management and internal cost control, in essence providing an administrative service rather than insurance coverage). In the case of insurance provided directly to consumers such as car insurance, policyholders can be given the opportunity to share in any surplus generated by their class of insurance, although not in losses. Such generated surplus is sometimes allocated in proportion to premiums, while in other cases it is only used to benefit those without claims. Such features can be distinguished from bonus/malus systems, which do not provide performance-linkage, but are rather forms of experience rating where premiums are adjusted prospectively to reflect current expectations of future claims experience of individual classes of business. In experience rating, each class of past claim experience represents its own pricing class and each such class forms a separate sub-pool. Performance-linkage originally related only to insurance risks (e.g., mortality or morbidity experience). It has often also included a part or all the actual investment returns in excess of those investment returns assumed in pricing. Such excess interest credits may be based on the amount in excess of that guaranteed, as well as on competitive conditions and the price sensitivity of the market. In some European countries, life insurers have focused on being specialized providers of insurance coverage, while in other countries life insurers have evolved to become general financial institutions, focusing mainly on investment products. A significant reason for this divergence arose from differences in the banking systems in those countries. In some, banks were authorized to provide the entire range of banking services to everyone (e.g., loans, short and mid term savings, and giro accounts), while in others they were only allowed to provide a very limited range of banking services. In the latter case, consumers historically did not have access what many are used to as full banking services providing a short and medium term savings vehicles for small amounts through the banking system. Insurers, using active sales forces that needed to sell insurance products, took advantage of that niche to provide savings services to satisfy those clients needs. In countries with intensive and wide-ranging bank-consumer relationships (e.g., payments for daily needs are often made by giro 4

5 transfer in the bank branch rather than by checks), banks had regular access to consumers to provide products for most investment needs, while insurers became specialized providers of insurance coverage and long term (old-age protection) saving needs, requiring a long term performance focus rather than short term performance-oriented fund vehicles. In some countries tax rules support certain life insurer products, in some cases of a pure savings nature. Banks have in some cases developed cooperative arrangements with insurers to utilize the intensive client contact at the bank counter for selling such life insurer savings products, especially in countries such as France. Currently there are several countries where life insurers provide products (e.g., capitalization contracts) with little if any insurance coverage (e.g., UK, France, Belgium). In these countries, participation features provide policyholders a share in the insurer s investment performance in the insurer s capacity of an asset manager. In such countries, the original purpose of performance-linkage (i.e., the re-transfer of the risk of the failure of a pool of insurance risks) disappeared, in which case the term life insurance has come to refer to a financial service rather than insurance of mortality risks. Thus, in some countries insurance products are distinguished from banking products only in the duration of the contract, while in others it is distinguished by the existence of significant insurance risk. Local contract law has played a significant role in determining the characteristics of a country s participating business. While in most Continental European countries contract law and regulation have been based on the Napoleonic Code, in the UK and other Anglo-Saxon countries the common law is its basis. In most Continental European countries, contract provisions describe the parties rights and obligations in detail, including performance-linkage features, and government bodies regulate the insurance industry. The insurer s discretion is relatively minimal. In contrast, in the UK and certain other Anglo-Saxon countries, the prudent man principle applies where the insurer is assumed to be responsible for maintaining their policyholders interests and national regulators have primarily relied on a system of self-regulation. Hence, in such countries, insurers have had significant ability to exercise discretion in applying policyholders rights, while in most Continental European countries performance-linkage features are tightly determined contractually. The participation systems of many countries were not designed to ensure participation in changes in the market values of assets. As a result, the treatment of unrealized gains or losses has been a significant issue in bonus allocations. This type of participation is currently done in the UK, Australia, Canada and Ireland. Sharing unrealized gains as they arise can significantly affect the participation system. These systems can incorporate features such as the use of discretion regarding both the amount and timing of the distribution of policyholders share in surplus, the use of revocable bonus allocations such as terminal bonuses, and the deferral of the distribution of recognized realized and unrealized gains by the use of special balance sheet items such as a fund for future appropriation. Some view the participation of policyholders in unrealized gains to be premature at best. Others view it as an appropriate means of preserving intergenerational equity. The intense focus on realized and unrealized gains in those countries has contributed to the impression that this is one of the main issues in participation systems, overshadowing the basic framework of the participating process. Participation systems in most countries have been designed in a manner that unrealized gains do not affect required distributions. Policyholders primarily share in total realized surplus, not just from investments, thus making participation in unrealized gains moot. Many contracts have such an extended lifetime that unrealized gains are often realized during the lifetime of the policies and therefore the investment returns will 5

6 ultimately be shared amongst the affected policyholders. This significant difference in participation systems should be considered in understanding their economic character. It is quite difficult, once established, for one system to switch to the other. For example, changing the underlying accounting approach used for distribution to one considering unrealized gains would necessarily require a second accounting approach to be applied or special consideration for the basis on which participation is determined. Although it can be difficult to determine whether the types of investments used by companies or consumer expectations in a particular market came first, the nature of an insurer s investment portfolio has tended to determine the approach taken to participating business. For example, in the UK, Ireland and Australia, stocks have played a big role. As a result, a significant objective of with-profit funds in those countries has been to smooth the effect of the volatility of the fair value of investments. This concern does not exist in other countries in which realized gains only affect the reported value when market value is less than cost, with funds primarily involving fixed interest bonds and property. The automatic smoothing effect of reflecting realized gains in some countries (e.g., statutory accounting in the Netherlands), is increased by deferring gains from the sale of fixed-interest bonds, which forms part of the participation system. In some countries (e.g., UK, Ireland and Australia), the results can be smoothed through the use of a welldiversified investment portfolio, discretion in the amount and timing of bonus allocations and through the use of the fund for future appropriation and terminal bonuses. In such countries, policyholders bear a larger degree of risk and volatility in maturity values than in other countries, particularly in the case of contracts of a relatively short duration. The perceived value to the policyholder of future insurability can be considerable, although evidence from experience can vary by country, product and company. For example, lapse rates for German life insurance contracts (all of which contain considerable insurance risk) are not noticeably correlated with market interest rates, the insurer s credit standing or the insurer s bonus rates, as policyholders concern with their future insurability is significant. A similar situation exists for with-profits life business in the UK. Many policyholders of contracts designed primarily for investment purposes do not attribute significant value to their guaranteed insurability. These policyholders are often willing to change insurance carriers if better prices present themselves, particularly if evidence of insurability is not a concern, is not required, if bonuses paid to existing policyholders are less than those on new business, or the market is sensitive to competitive pressures that affect bonus rates. This in part is due to the types of distribution channels used (e.g., the UK has a significant number of independent brokers while tied agents are common in Germany). In countries such as Germany where individuals do not commonly purchase stocks, most insurance policyholders expect that over the long run their insurance contract will outperform other investments available through the financial markets. As a result of significant differences in regulation among countries, performance-linked clauses can either be specifically regulated by the state, while in many other countries it is assumed that competition will force insurers to provide adequate returns. In the latter case, bonuses are significantly, if not entirely, at the discretion of the insurer, even if policyholders have the right to share in surplus. However, in many jurisdictions, such extended rights of a party to a contract to unilaterally influence the price-benefit relationship are not compatible with contract law, requiring strict contractual regulations for the entire price-benefit relationship. In the case of universal life contracts, policyholders do not have any right to share in surplus, since crediting any interest to policyholders is at the sole discretion of the insurer, subject to minimum guarantees, and is mainly driven by competitive considerations. At the other extreme, unit linked 6

7 policies are designed to allow policyholders direct access to investment funds, in which case the insurer acts as an intermediary, acquiring the units and selling them directly to policyholders. The popularity of unit-linked life insurance or annuity contracts varies by country, depending on the availability of similar products from other financial intermediaries and their attractiveness compared to traditional life insurance contracts in terms of transparency, understandability and compensation to distributors. Participation is significantly influenced by the key profit triggers of a life office. Considering the large margins in premiums of participating contracts and the relatively small share that stockholders have in surplus, the design of the participation systems as well as the accounting system, strongly determine the emergence of profit. Often, the participation system result in different sources of surplus (e.g., surplus from investment activities, mortality, administration) having different consequences. Further, rules often exist not to carry forward losses. Hence the period an expense or earnings item occurs will affect its impact on profit. Further, in some cases losses from one source may not be offset by surplus from another or may only participate in profits from certain sources and not others. Complex participation rules might cause complex effects and interdependencies between effects. Since the performance of an entity can be influenced by management s decisions, nearly any performance-linkage feature is subject to managements influence. Cases in which significant discretion can be applied, as in the case of US universal life contracts, are not considered performance-linked, since there is no obligation that benefits be correlated with performance. In some countries (e.g., Belgium, New Zealand), contracts at least contain some expectation that additional benefits are provided with some relation to the performance of the insurer. However, these are payable at the discretion of the insurer and may not reflect actual performance achieved. In other countries, policyholders have specific rights to receive a certain share of surplus. In some cases, contracts may note that expectations should be limited and that only the minimum is guaranteed and that they might receive a further share of surplus at the insurer s discretion (e.g., UK). In rare cases insurers even publish which share of surplus has actually been provided to policyholders; hence the percentage of any share in excess of the contractual minimum becomes a marketing feature. In most cases the amount paid in excess of a contractual minimum is voluntary and principally reflects marketing considerations, since policyholders receive only information about the absolute amount paid at maturity and not the information regarding the percentage of surplus that is represented by the bonuses in excess of the contractual minimum share in surplus. Although in many cases discretion is available, with benefit payments primarily subject to competitive forces and the price sensitivity of the market in which the company is a participant, discretion often is significantly constrained by country, industry and company practice. In many cases a company has a stated strategic policy regarding the extent to which such experience is shared or the competitive position it aims to be in. In addition, in some cases the very existence in sales illustrations creates expectations on the part of the consumer who decides to purchase the contract that if conditions relevant to contract performance remain similar (or in some cases even in spite of changes), the benefits provided will be at least similar to those illustrated, thus possibly forming a constructive obligation of the insurer. In many countries, the contractual share of surplus is specified in the contract as a right with no expectation that anything more will be paid. However, in some countries that contractual share is relatively small, the margins are very large or the competition is very strong, causing insurers to voluntarily pay out more than required just for competitive reasons. Only in a few countries (e.g., Italy) do insurers provide exactly the contractual share of surplus to policyholders, since 7

8 competition does not force them to pay more and the remaining amount is needed to provide needed capital. In these countries, the only effect that management can have on participation benefits is to influence the timing of surplus recognition. This influence is economically close to zero where policyholders share directly in specified assets held by the insurer and receive the entire surplus from those assets. Those contracts are similar to unit-linked contracts, but the difference is that the contract refers explicitly to surplus of the insurer from assets held by the insurer rather than to an externally determinable value of units of a unit fund. In some countries (e.g., Netherlands, Israel) insurance contracts have been developed which are linked explicitly to an internal fund (i.e., they are essentially performance-linked, since they refer to the insurer s net earnings specifically from investments rather than to an external financial index), but on the other hand they have the transparency and objectivity of unit-linked contracts, since the entire performance of linked assets is provided entirely to policyholders. Similar to unit-linked contracts, such approaches typically do not include minimum guarantees for benefits. The insurer is merely an asset manager rather than accepting risk, pooling it and retransferring the risk of pool failure. As a consequence, from the perspective of company solvency, the pricing for participating contracts typically requires less consideration of the competitive relationship of premiums and guarantees, since the contract is ultimately re-priced via the participation features and the policyholder receives what was achieved by the insurer. Competitive and marketing considerations may prevail in the setting of premiums and design of products. Insurers typically require low equity to cover participating contracts due to the security provided by the large margins in the premiums. Where investment guarantees exist, these tend to be addressed through use of an appropriate investment policy. Thus, in many jurisdictions where all, or virtually all, contracts are participating, stock companies operate in a manner very similar to mutual companies. The actual profit margin in premiums is extremely small. Often regulation ensures that policyholders actually receive what they contributed to surplus and insurers merely get a certain return on equity provided as their contribution to surplus. As a consequence, it is a very important accounting issue to describe that small profit margin adequately. Considering the volatility of the balance sheet and income statement, the remaining profit is very small, since the participation system allows insurers to pass almost all such volatility on to policyholders. Since it is very difficult to adequately describe the extent to which the insurer contributed to surplus, in most cases that contribution is expressed as specific amount, normally a percentage set by regulators. 8

9 Features of non-guaranteed benefits in contracts provided by insurers It is assumed that all benefits are provided as defined based on an obligation or due to an economic reason external to the contract (e.g., competition). Purely voluntary payments may effectively reflect a certain intention of the insurer but the economic character of the benefit is that of a voluntarily provided benefit. However, it can be assumed that insurers typically do not provide any amount to anyone without a good economic reason. In reality, it is often assumed by policyholders that if conditions remain similar to the conditions at the time the contract is issued that the amounts illustrated at the time of application will continue. The term benefit is used in the most general meaning (i.e., includes reductions of premiums otherwise due, provision of services, providing payments earlier than otherwise due (interest benefit) or deferral of payments to be received (loan benefit)) and takes into account the time value of money. Also included are negative benefits (i.e., cases where benefits otherwise due can be reduced or premiums increased). A. Guaranteed and uncertain benefits Guaranteed benefits are those where for each amount paid by the policyholder it is certain (except for the risk of insurer default) which amount will be repaid at any possible repayment date expressed in a fixed currency amount. That means that for each possible duration there is explicitly or implicitly a fixed effective interest rate or insurance charge. Event-based guaranteed benefits are those whose benefits are similar to guaranteed benefits, but a specific guaranteed benefit is determined at the payment date for each possible alternative occurrence or timing of occurrence. For example, in the case of death within a specific period a guaranteed benefit is provided, or in the case of surrender, a different guaranteed benefit is provided. In other cases the timing of death also influences the amount (e.g., the death benefit may be linked to inflation). Guaranteed benefits are a trivial special case of event-based guaranteed benefits. Insurance benefits are those for which the value of the benefits (i.e., amount of benefit or timing of the benefit in the case of a fixed amount) is based upon the type and possibly the severity of insured event. The insured event determines the value of the benefit to be provided, usually reimbursement of the damage caused by the event. If the event is an insured event, event-based guaranteed benefits are simple special cases of insurance benefits. Insured events are those events, specific to the insurance contract, that have potentially adverse effects, have an influence on the cash flows paid by the insurer in performing the contract, can be observed objectively, and are to a large extent outside of the control of the policyholder. Performance-linked benefits are those for which the value of the benefits is correlated to the net earnings of the insurer before providing performance-linked benefits. Experience rating benefits are those for which individual policyholders receive premium reductions in future if past experience indicates that the individual risk is better than average and lower premiums are required in the future. Such future premium reductions do not intend to return any of the profit achieved from the favorable past experience to the policyholder. The same applies to increases in premiums as a consequence of adverse past experience. 9

10 Prospective premium adjustment benefits are those for which insurers are obliged (or entitled) to adjust future premiums of long-duration contracts on a portfolio basis if past experience indicates that the premiums otherwise due are not adequate in the future. Item-linked benefits are those whose value of the benefits is determined as a share of changes in value of a specified item between the payment date of the premium and the due date of the benefit. Although it is not required that the insurer controls the item, it is possible that it does control the item and that the benefit is effectively equivalent to a performance-linked benefit. The difference is that the benefit is not explicitly based on the performance of the insurer but it is at the decision of the insurer whether it controls the matching items. A typical example is a unitlinked contract where the item is a unit of a unit fund. Whether the insurer is obliged to control or own the items or not may constitute a significant difference for policyholders, since the default risk can be significantly larger without matching. Although it provides the insurer greater flexibility, it does not benefit the policyholder. On the other hand, since the contract refers to the item itself rather than to its economic use by the insurer gained from holding the item, insurer is not covered against the loss of the item (e.g., by theft). That is the significant difference between performance-linkage and item-linkage. The first refers to the economic use the insurer made from items held, and the second to the item directly. Index-linked benefits are those where the value of the benefits is determined as a function of an index. All benefits described before are special cases of index-linked benefits. A type of indexlinked benefit not previously mentioned are called external-index-linked benefits. In particular, any benefit provided for competitive reasons can be considered to be an external-index-linked benefit. Providing benefits for competitive reasons implies that they are provided to a similar extent as benefits provided by competitors. Such benefits could be considered to be more strongly linked to an external index rather than to the net earnings of the insurer, although clearly the available resources of the insurer and considerations about the expected future returns from the business to be acquired limit the value of the benefit. The latter limit means that the benefit is influenced by the economic value of new business to be acquired (i.e., providing the benefit is a form of acquisition cost for new business and could be considered in a similar manner). Also, most benefits provided by means of constructive obligations will be external-index-linked benefits, since internal indices are typically not observable for policyholders. Hence, constructive obligations do not often arise from internal indices. Policyholders will not necessarily expect that the insurer will provide a certain percentage of its own surplus, but will more typically expect market consistent benefits and may disregard the net earnings of the insurer as long as they enable the insurer to provide such market consistent benefits. Regulatory-imposed benefits are those which are not one of the above types, but are paid on regulatory requirement. In some cases, regulators force insurers to provide additional benefits to policyholders for political reasons without that obligation arising from contracts. Considering that insurance contracts can last decades, regulatory authorities have been known to intervene in the benefits payable, despite the contractual relationship between policyholders and insurers, to establish and cease obligations, to require changes of contract clauses or to require or prohibit payments. Voluntary benefits are those which are not one of the above, but while paid without any external obligation, are self-imposed. Clearly, any person can decide to provide his or her own resources to others. Hence, insurers can over-perform their obligations. However, typically any gift by a profit-oriented entity (ownership benefits of mutuals can be regarded as a special case of performance-linked benefits) will usually be made based on current or expected future economic considerations. The most important reason for voluntary benefits are competitive reasons, if it 10

11 can be assumed that benefits to existing policyholders may either motivate existing policyholders to maintain or increase the business relationship or, if the voluntary benefit is known to public might attract new customers. Typically, information about the total benefits provided by insurers to policyholders, especially in life insurers, is available to the public and often is used as a basis for an insurance purchase decision. Hence, increasing total benefits by voluntary amounts can improve the competitiveness of insurers. If the insurer has the permanent intention to compete in the market, the benefit is not a voluntary benefit but an external-index-linked benefit. Comment: Financial instruments typically provide guaranteed benefits, item-linked or externalindex-linked benefits. In the case of stocks, benefits are typically performance-linked benefits. The characteristics of item-linked benefits allow a kind of hedging mechanism. If the entity decides to acquire items on receipt of payments, the economic situation is nearly equivalent to performance-linked benefits. If the entity does not acquire the items, the situation is equivalent to external-index-linked benefits. The naming convention describes the economic cause for the benefit provided, not whether the benefit was caused by a current obligation or by an intent of the insurer (voluntary), which although it does not reflect a current obligation is payable as a result of a business decision. The split between current obligations and business decisions of the insurer is another dimension of classifying benefits. Typically, many benefits are based on current obligations, while voluntary benefits and external-index-linked benefits based in part on competitive considerations. B. Performance-linked benefits A performance-linked benefit is determined based on the net earnings or experience of the insurer before any performance-linked benefits are paid and performance-linked benefits are a share (i.e., in total not more than 100%) of those net earnings or experience. The linkage ranges from a general correlation between benefits and net earnings (i.e., a general intention to return some share of surplus to policyholders in the long run) to a fixed percentage, from a minor or infrequent share in surplus or experience to 100% of surplus. Performance-linked benefits mean that the actual pricing process is deferred at the outset and completed later considering the actual cost of services provided. Retrospective pricing benefits are those benefits whose value is determined as a share of the net earnings from the individual contract (i.e., a retrospective pricing 1 of payments made or services provided under the contract in an individual assessment for the contract). These benefits include benefits from contracts priced retrospectively at cost, but also contracts where the contract is explicitly invested in specified investments held by the insurer and the policyholders receive part of or the entire investment return of that investment. Participation benefits are those benefits whose value is determined as a share of the net earnings or experience of a group of typically similar contracts before any participation benefits are provided. These cover both shares of the individual policy in the entire or partial surplus of the insurer based on the contribution principle and cases where the entire surplus is split between groups of contracts and those parts of surplus are then allocated to individual policyholders within the particular group. These include cases where the premiums of a group of policies are invested and each policyholder shares in the returns of the invested assets, but assets are not allocated to individual contracts. 1 Retrospective pricing is distinguished from prospective pricing in which past experience is used to provide a more adequate pricing of future contract periods, without the intention to recover losses occurred in the past. 11

12 Ownership benefits are those benefits whose value is determined as a share of the net earnings of the entire entity before any ownership benefit is provided. These include ownership rights of stockholders of a stock company, but also membership rights of members of a mutual entity. Such ownership benefits are different from participation benefits. Ownership benefits do not reflect shares in surplus according to the contribution principle for individual contracts or group of contracts (i.e., they do not reflect the relative contribution of each individual contract or a group of contracts to the surplus) but typically refer to capital provided under the contract covering the general risks of the insurer. Hence, ownership benefits are usually provided based on measures such as premiums paid, sum assured, and amount of liability, as a relative share of the entire surplus of the insurer, if those amounts do not reflect the contribution of the individual policyholder. Clearly, if nearly the entire surplus of the insurer were generated as a result of investment income, allocating the surplus to policyholders based on the amount of a liability could be considered to be a participation benefit. In the case of non-life insurance, where premiums and risk born by all policyholders is typically comparable, a percentage of premium would also usually be considered to be a participation benefit. C. Participation benefits Participation benefits return the risk of adverse experience or failure of mitigating processes in the insurer back to policyholders as a group, rather than on an individual basis. The most well known example today is when policyholders funds are invested commonly in a portfolio of well-diversified investments and policyholders share the outcome, while no individual policyholder could maintain such a well-diversified portfolio at reasonable cost. However, the original purpose of participation benefits is the participation in the risk process, meaning both the refund of unneeded security margins in premiums (which were set at a level to provide long-term guarantees) and the return of any failure of pooling to policyholders. If the premium is set in sufficient size to cover virtually any potential loss, the failure of pooling compared with income available to cover claims can only be a less large surplus. 1. Source of participation benefits a) Participation required by law or state regulator; b) Participation agreed by contract; and c) Participation as a consequence of a constructive obligation 2. Location of participation rules Determination of participation benefits can be specified in: a) Law or state regulations; b) Contracts; c) Past behavior of the insurer creating a constructive obligation to continue to provide participation benefits in a specific manner; or d) Nowhere (i.e., although there is formally an obligation to provide participation benefits, the consequences of that obligation are not described anywhere). 3. Process of determining participation benefits The process may contain all or some (some steps may be combined) of the following steps: 12

13 a) Identification of distributable net earnings that are the basis for participation benefits or participation surplus (i.e., the assessment of net earnings of the insurer in accordance with the specifications for participation benefits). Such net earnings specifically reflect management measures and hence participation surplus is generally subject in some extent to management decisions, except in those rare cases where the entire net earnings of the insurer is administered by independent third parties (e.g., in case of external funds owned by the entity which are the contractual basis of participation benefits). b) Determination of a part of that participation surplus to be used for participation benefits of policyholders and another part that belongs ultimately to the insurer. Amounts allocated for participation benefits might in some countries or contracts later be utilized to the benefit of the insurer in very specific and usually state regulated cases (e.g., to avoid a loss which might endanger the claims payment ability of the insurer). c) Determination of a participation benefits for individual contracts by allocating specific amounts to individual contracts. In most countries, step a) requires that step b) be immediately performed for the entire participation surplus. That is due to the fact that the split between the policyholders share and the insurer s share is not subject to the discretion of the insurer but is based on a fixed percentage. In a few countries (e.g., UK, Republic of Ireland, New Zealand, and some contracts in Portugal) the identified participation surplus can be parked for some time before the decision in step b) is made. These contracts grant insurers discretion about the amount of policyholders share in surplus and the execution of that discretion can be deferred. In some countries the parked amount has been reflected as an intermediary item in the balance sheet ( fund for future appropriations ) and in other countries (e.g., Portugal) the amount is reported as a liability. The insurer normally has a significant degree of discretion regarding the timing of step b). However, earnings from investing those amounts are often subject to participation rules as well, and the potential benefit of the insurer from that discretion is small. However, there is uncertainty about the insurer s share for a significant time. In those countries where step b) is applied immediately after step a), the part belonging to the policyholders (often including future policyholders) is usually reported as a special liability ( liability for future participation benefits ). The character of that liability depends significantly on the restrictions on making use of those amounts, e.g., in case of emergency. In some countries (e.g., Italy) step I immediately requires step b), which is combined with step c) (i.e., the amount to be used for participation benefits is determined by the allocation of participation benefits to individual policyholders and any remaining amount belongs ultimately to the insurer) directly at identification of the participation surplus at year-end. Hence, there is neither a fund for future appropriation nor a liability for future participation benefits, but any surplus is directly transformed to equity or policy benefit liabilities. In rare cases, step a), b) and c) are made at the same time by policyholders sharing directly in the fair value of assets (e.g., Israel). Such contracts are equivalent to unit-linked contracts in which the insurer is contractually obliged to hold the units, although the contract does not refer to the market value of specific assets, but rather to the net earnings of the insurer resulting from assets held by the insurer. Hence, some risk (e.g., the risk that the assets are stolen) is not borne by the insurer but by the policyholders, in contrast to pure unit-linked contracts. 13

14 Step a) is of significant accounting relevance, since the net earnings assessed in determining participation surplus are normally not those currently reported in IFRS financial statements. For example, US GAAP provides detailed guidance on how to consider timing differences between the US GAAP financial statement and the assessment forming the basis of participation surplus (e.g., US statutory accounting). 4. Basis of determining participation surplus The amounts legally available as participation surplus for participation benefits are based on participation accounting. Some examples of approaches to participation accounting are: a) The same accounting approach (the term accounting approach assumes that ultimately all cash flows of the insurer are recognized) as is used for determining other appropriations of net earnings (e.g., taxation or distribution of profit to stockholders). This is typically referred to as statutory, regulatory, solvency or prudent accounting. Policyholders and other stakeholders are treated equitably, since even if the insurer has full discretion about the amount to be allocated, that discretion can only be executed within the range of amounts available for distribution in general. b) A special accounting approach in which net earnings are recognized in other periods than in approach a). Policyholders and other stakeholders are treated ultimately equitably but the insurer may influence the timing to some extent utilizing the timing differences, smoothing results for policyholders or coping better with shortcomings of the participation process (e.g., a restriction on off-setting losses of one period with surpluses of another). This approach also includes cases where the ultimate surplus achieved over the entire contract duration is the basis of participation benefits. c) A special accounting approach in which net earnings recognized are ultimately different from those recognized in approach b). In other words, the permanent difference might grant policyholders advantages or disadvantages in comparison to other stakeholders. Insurers can take advantage of this or might even cause damage to other stakeholders by favoring policyholders. Depending on the extent, it can be questioned whether policyholders are actually sharing in net earnings or whether the benefit is merely an index-linked benefit which is not linked to actual performance. d) No specified accounting approach (i.e., the amounts available for distribution to policyholders are determined on a case by case decision of the insurer). In such cases the policyholders sharing in net earnings will be difficult to prove in practical terms, the reference to net earnings is merely a formality. For example, in Belgium policyholders share in net earnings according to contract, but it is not specified how those net earnings are assessed. The determination of amounts to be distributed to policyholders is entirely at the discretion of the insurer. 5. Potential differences between participation surplus and IFRS net earnings As outlined earlier, the basis of allocation of participation benefits to policyholders is the participation accounting. This means that the allocations recognized legally reflect the performance of the insurer as measured for the current and past periods by participation accounting. That accumulated past performance as measured in participation accounting, the 14