Carter on Reinsurance Volume 1 Witherby Insurance and Legal A Division of Witherby Publishing Group Ltd 4 Dunlop Square, Livingston, Edinburgh, EH54 8SB, Scotland, UK Tel No: +44(0)1506 463 227 - Fax No: +44(0)1506 468 999 Email: info@emailws.com - Web: www.witherbyinsurance.com iii The Biographical Note at the beginning of this book gives a bare bones summary of Bob’s career. It tells only part of the story, some of what Bob accomplished in his long career but not a lot about the man. Those who worked with Bob, either in university life or in his academic and research work in the insurance and reinsurance market, came to know a man who enjoyed his work, brought enthusiasm to a subject, engaged with people with clarity and generosity in giving time (as his students testify) and brought an academic rigour that made authoritative his published works in the business of insurance, reinsurance, and risk management; and all accompanied by a lively sense of humour. The first edition of Reinsurance was published in 1979 and shortly after Bob was commissioned by the CII to be the editor of three new reinsurance course books that were published in 1981; a no mean task when the contributions from a multitude of market practitioners on different reinsurance topics had to be put into the language in an order that made sense to students! During his tenure at the University of Nottingham, Bob developed some of the UK’s first undergraduate and diploma courses in insurance and risk management, and travelled extensively nationally and internationally to promote the academic study of insurance. His undergraduate courses live on - they’re continually updated and enhanced and are still being taught to current business school students. Bob’s collation of insurance statistics laid the foundation for the use of such data by a generation of insurance researchers at Nottingham. Bob retired from University in 1990; his health was not robust and in his teen years he had been hit by polio. After retirement, Bob continued to write prolifically on insurance, reinsurance and risk management, and to act as editor of three well-regarded practitioner handbooks. His books include Economics and Insurance (PH Press), British Insurance Industry: A Statistical Review (Springer), The Handbook of Insurance (Kluwer) and Reinsurance (Reactions). Bob co-authored Reinsurance Essentials with Leslie Lucas (Reactions 2004). Success in Insurance (John Murray), written with Stephen Diacon, was a widely-read introduction to the subject. His last book, published in 2009 and co-authored with Peter Falush, was “The British Insurance Industry since 1900: the Era of Transformation” (Macmillan). Stephen Diacon, now Professor of Insurance and Risk Management at Nottingham University, said “Bob was very disciplined, organised and relentless in his approach to writing, and would write most days. I remember that he set himself word targets (2,000 per day, which is a lot) and would meet his target wherever he was in the world at the time. The effort and commitment that Bob devoted to his writing was truly stupendous. As an example, he was editor-in-chief for three loose-leaf handbooks at the same time! The best known of these was the Handbook of Insurance, often known as “the insurance bible”. It seems incredible, but Bob had to manage (and often rewrite) over one thousand pages of text each year - and he did this for many years.” Tribute to Professor Robert (Bob) L Carter 23 August 1932 - 16 September 2012 iv Nigel Ralph and I were invited by Bob to share with him in writing Reinsurance, Fourth Edition, published in 2000. With the world of international reinsurance going through more change with new and growing challenges, typified by the World Trade Center terrorist attack (9/11), Bob invited us to join him again in the writing of Reinsurance, Fifth Edition. This was Bob’s last work and he was engaged on it up to his death, and had delivered to the publishers all the chapters for final preparation before publishing. We got to know Bob well. Bob was well rounded; he enjoyed all aspects of life; his work, his family, and people. Reinsurance will stand for many years as a tribute to Bob Carter. Leslie Lucas vii Contents Preface to the Fifth Edition ix Volume 1 1. The Role and Development of Reinsurance 1 2. Types of Market 19 3. Principles and Practice of Reinsurance 65 4. Legal Principles Applicable to Reinsurance Contracts 99 5. Forms of Reinsurance 145 6. The Pricing of Non-life Reinsurance Contracts 191 7. Facultative Reinsurance 255 8. Proportional Reinsurance Treaties 273 9. Non-proportional Treaties 319 Volume 2 10. Fixing Retentions 371 11. Property Reinsurance 403 12. Catastrophe Reinsurance 443 13. Casualty, Personal Accident and Other Non-marine (Except Property) Reinsurance 473 14. Marine and Aviation Reinsurance 519 15. Life Reassurance 581 16. Finite Risk Reinsurance and Alternative Risk Transfer 627 17. Technical Accounting 649 18. Financial Accounting and Management 669 19. The Management of Reinsurance Operations 701 20. International Practice and Problems 731 Bibliography 745 Index 759 147 5 Forms of Reinsurance This chapter examines the characteristics, advantages and disadvantages of the main forms of reinsurance. Pricing is discussed in Chapter 6. Chapters 7, 8 and 9 deal with the methods of operation as laid down in the various types of reinsurance contracts. 5.1 Quota Share Reinsurance The simplest form of traditional treaty reinsurance is the quota share contract under which the reinsurer agrees to reinsure a fixed proportion of every risk accepted by the ceding company and so share proportionately in all premiums and losses. Financial reinsurances may also take the form of quota share (see Chapter 16). In return, the reinsurer receives the same proportion of all direct premiums (net of return premiums), less the agreed reinsurance commission. The treaty will specify, inter alia, the class(es) of insurance covered; the geographical limits and any other restrictions, such as any specific types of risk or perils excluded from the treaty and whether a monetary limit applies to any one risk or loss accumulation. The treaty will provide automatically that the ceding company will cede and the reinsurer will accept the agreed share of every risk underwritten that falls within the contract. 5.1.1 The Advantages The main advantage for a primary insurer of quota share reinsurance is its simplicity and low cost of operation. Once the treaty has been arranged, very little administration and accounting are required. Provided all of the risks accepted by the ceding company fall within the terms of the treaty, no special effort needs to be devoted to the reinsurance of any individual risk (unless additional reinsurance protection is required) or to the apportionment of premiums and claims. This enables experienced staff to devote their time to other aspects of the business. Likewise quota share treaties require little administration by reinsurers. The ceding company also benefits from being able to accept larger risks than otherwise would be possible, knowing that reinsurance is available automatically to contain potential losses within an acceptable limit. Moreover it can write a larger account than could be supported by its capital base. By receiving a share of every risk written by the ceding company, the reinsurer will obtain a more balanced portfolio of business than could be obtained by reinsuring the same account by any other form of reinsurance. Such participation in every risk creates an identity of interest between ceding company and reinsurer. It often induces the latter to provide ceding companies with technical training facilities and other services, which are of particular value to new companies, particularly in developing countries. The favourable view that reinsurers take of quota share business leads to higher rates of reinsurance commission being paid than on surplus treaties covering comparable original portfolios. On a very profitable account, a ceding company may be able to obtain: • A reinsurance commission at a rate sufficiently high to cover its acquisition costs and contribute to administrative expenses, and • a profit commission, which effectively would provide it with a risk-free profit. 148 Also the retention of premium and possibly loss reserves, and the payment of reinsurance premiums after the receipt of the original premiums, may leave a ceding company with a source of investment income above any interest payable to the reinsurer. That will help to finance the expansion of the company’s own business. As a company builds up its gross premium income and reserves, so it will be able to increase its retention limits and reduce the share ceded, if it so desires. 5.1.2 The Disadvantages Quota share reinsurance does suffer from significant disadvantages. In particular, the ceding company cannot select the risks it cedes, so the reinsurer will take a share of the premium for risks that lie well within the cedant’s own financial capacity. The ceding company will gain from being able to accept larger risks than it otherwise could, but because it reinsures the same proportion of every risk underwritten, the relative variability of expected losses on the retained portfolio will be the same as on the total portfolio. Therefore, judged on its impact on the risk borne by the ceding company, quota share reinsurance fails to reduce: 1. The incurred loss ratio on the retained account, and 2. The possible relative variation in actual retained losses incurred during any one year from the expected retained losses (ie it will not reduce the coefficient of variation of retained losses). On the other hand, by reinsuring a fixed proportion of all insurances written, a company will reduce its probability of ruin because it will achieve a smaller absolute variation in retained losses (i.e. the standard deviation) relative to its capital and free reserves. Quota share reinsurance, however, does not adequately protect an insurer against the risks of either accumulations of losses from one event, or an overall increase in the frequency and/or severity of losses in any one year. Although the reinsurer is liable to pay its agreed share of all of the individual losses, quota share reinsurance does not provide a cap to the ceding company’s aggregate retained losses from either one event or in any one year. 5.1.3 Uses of Quota Share Reinsurance Because of its disadvantages, insurers do not tend to rely solely on quota share reinsurance for protecting any particular class of business. The main exceptions are new companies and companies commencing a new class of business or new area of operation, who require the reinsurance protection afforded by a quota share treaty. Although initially the portfolio of business may be small and limited in its spread, and the ceding company inexperienced in handling such business, the reinsurer can be certain of receiving a share of every risk, with no possibility of anti-selection by the cedant. Depending on the nature of the business to be transacted, the reinsured may need to supplement the quota share reinsurance with a non-proportional cover to provide protection against an accumulation of risks. Quota share reinsurance may also form part of a reinsurance programme for other reasons, such as: 1. Where the ceding company wishes to arrange reciprocal exchanges of business. 2. For classes of business where defining a single risk is difficult, eg crop hail insurance. 3. For reducing a ceding company’s exposure under policies covering natural perils. 4. For classes of insurance where, although there may be policy limits, the incidence and size of losses are uncertain, eg liability insurance. 149 5. At times of financial difficulty, or perhaps following the implementation of more stringent solvency regulations, to provide an easy and effective means whereby a company can quickly reduce its retained premium income to enable it to continue to meet a prescribed ratio of net premium income to capital and free reserves. 6. As part of a group-underwriting practice under which all members of a group, instead of participating directly in an insurance written by one member, share it by means of quota share reinsurance up to agreed limits before interesting outside reinsurers. 5.2 Surplus Reinsurance 5.2.1 The Advantages Surplus reinsurance is the most commonly employed form of proportional treaty reinsurance. Like quota share, it is a form of proportional reinsurance by which the reinsurer accepts a certain share of a risk, receiving an equivalent proportion of the gross premium (less reinsurance commission) and paying the same portion of all claims. The basic differences between the two are: • Under surplus treaties, the reinsured only reinsures that portion of any risk that exceeds its own retention limit, and • quota share reinsurance can be used for any class of insurance whereas surplus treaties can only operate for property and those other classes of insurance where the insurer’s potential maximum liability is expressed as a sum insured or policy limit. Normally under a surplus treaty, the ceding company will adopt varying retention limits directly related to the degree of risk associated with different types of exposure units. For example, a property underwriter may fix a higher retention for, say, engineering factories of fire-resistant construction than for timber-built woodworking shops. That is because both the probability of a fire loss occurring will be lower in the former type of premises, and the extent of any loss relative to value at risk will probably be smaller too. In other words, in fixing its retention limits an insurer will consider the same factors as are taken into account in fixing premium rates. Some companies draw up their tables of retentions based on the premium rates applied to the original insurances; this practice is examined in Chapter 10. Thus, surplus treaties overcome one main disadvantage of quota share treaties. The ceding company can keep for its own account the entire premium for risks underwritten that fall within the scope of its own retention, differentiating for some portfolios between classes of risk. Also, by ceding amounts above its retention limits, but retaining for itself 100% of risks falling below those limits, the ceding company reduces the range of possible losses on its retained portfolio, thereby reducing the potential relative variability of its aggregate claims costs. Reinarz1 says on this point that a surplus treaty enables ‘the law of large numbers to operate with maximum efficiency’, and adds: ‘ The surplus treaty imparts a limit of size homogeneity to the primary insurer through the working of the retention. The retention effectively cuts off all variance above a certain level of liability in each class, thereby definitely limiting the variance to a smaller margin. The reinsurers absorb the wide variances in size.’ 1 Robert C Reinarz, Property and Liability Reinsurance Management (Greenwich, Conn: Mission Publishing Co, 1968), p.29 150 200,000 150,000 Total number of policies with stated sum insured or more 100,000 50,000 Sum insured £20,000 £40,000 £60,000 £80,000 £100,000 £120,000 £140,000 £160,000 £180,000 £200,000 £220,000 £240,0000 Figure 5.1 Original portfolio Retained sum insured 200,000 150,000 100,000 50,000 0 Total number of policies with stated retained sum insured or more £20,000 £40,000 £60,000 £80,000 £100,000 £120,000 £140,000 £160,000 Figure 5.2 Retained portfolio: 50% Quota share The same point is illustrated in Figures 5.1, 5.2 and 5.3. The first diagram shows the distribution of an original portfolio by sums insured. A 50% quota share treaty would reduce the ceding company’s retentions as shown in Figure 5.2; although its maximum liability would be halved, the relative variations remain unchanged. The effect of a surplus reinsurance is shown in Figure 5.3. For the sake of simplicity, it is assumed that the company fixes a flat retention of M instead of having differing retentions according to categories of risk, as would be more usual in practice. As can be seen, the retained portfolio distribution is contained within narrower limits, so giving a lower coefficient of variation. 214 6.3.2.4 Fluctuations in Loss Experience When the company’s loss experience fluctuates substantially from year to year, the reinsurer must decide what confidence can be obtained from an average rate obtained from several years’ data. The first step is to look for the possible reasons for the fluctuations, such as the random occurrence of a few large claims on a small portfolio or the effect of the odd exceptionally large loss on a sizeable, well-spread block of business. Alternatively, the account may be unbalanced or of poor quality risks and suffer in some years from an unusually high incidence of large losses. Whatever the reason, allowance must be made for it in the reinsurance premium, and the reinsurer may obtain some guidance from its own experience or from market data covering the same type of business. 6.3.2.5 Allowing for Changes Affecting Past Loss Experience Whatever the period of the loss data supplied by a company, its reliability as an indicator of claims that are likely to occur during the treaty year under consideration can usually be enhanced by adjusting it for changes that may have occurred in the size and/or mix of the portfolio to be reinsured, in policy conditions and reinsurance arrangements, and in other factors that may affect net losses. Realisation of the need to build adjustments into loss experience based rating systems came with the high inflation rates of the 1970s and 1980s and the legal and other developments that undermined the reliance on pure burning cost as a method of rating. The development of probabilistic models in the 1990s also helped to supersede the burning cost method. 6.3.2.6 Catastrophe Potential The possibility of exposure to an accumulation of losses from one event needs to be considered not only in respect of catastrophe covers but also for the rating of other forms of non-proportional reinsurance treaty. Factors impinging on the catastrophe potential of a proposed treaty include: 1. The classes of insurance involved, including any aspect presenting a potential catastrophe exposure 2. The portfolio to be reinsured. For example, in the case of a property portfolio including cover against natural perils, the company may be heavily committed in areas exposed to natural disasters. It should be ascertained whether the original insurances are subject to deductibles or other limits on the amount of cover provided and, if so, of what size. 3. The form of reinsurance required and details of inuring reinsurances, including any restrictions imposed thereunder on recoveries for loss accumulations. 4. Proposed exclusions, treaty limits or other terms bearing on the reinsurer’s potential exposure. For example, if a reinsured is willing to accept more restricted geographical and/or time limits in the definition of an ‘occurrence’ or to exclude particularly hazardous risks from top layer working covers, the reinsurer’s potential liability will be reduced and so a lower premium may be warranted. The premium will also require adjustment for the number of reinstatements permitted. The purpose of a catastrophe cover is to protect the company against occasional large losses. As reinsurers expect claims to be few and far between, it is normal practice for property and marine catastrophe treaties (and also the higher layers of risk excess covers) to be subject to an automatic reinstatement clause that will usually require the reinsured to pay an additional premium to reinstate the cover in the event of any claim being incurred under the treaty. In some cases, the reinsured elects to pay for the reinstatement in advance, in which case there is no further 215 premium to pay when the cover is reinstated after the loss. Such provisions do not usually apply to motor third party liability treaties where the expected claims cost to the layer is calculated on a statistical basis that takes into account a reasonable probability variance from the average expected claims cost. 6.3.2.7 Administrative Expenses The costs of administering a working excess of loss treaty, or at least the lower layers of such a cover, will in theory be larger than for either a catastrophe cover or a stop loss treaty. If a treaty is placed by a broker, the cost of the brokerage must be brought into account, although some offsetting savings may be achieved by taking full advantage of the services the broker can provide. The level of service expected of the reinsurer (eg help in training the reinsured’s personnel or in handling claims) and the amount of contact requiring travel to see the reinsured, may also be allowed for in the premium loading. Brokers’ commissions are normally based on a fixed percentage of the reinsurance premium. 6.3.2.8 Continuity Continuity of treaty arrangements between a reinsured company and a reinsurer is another factor to consider when fixing a reinsurance premium. It has been said that: ‘Excess of loss reinsurance to a greater extent than other forms of reinsurance ought to be a long-term relationship’. However, since that was written in 1975, the value placed on continuity has diminished. Partly this is because proportional reinsurance is no longer regarded as support for excess of loss, and excess of loss is now the predominant form of reinsurance. More importantly, the results of direct insurers worldwide have deteriorated, both in terms of average profitability and in the variance of results caused by a higher frequency of very large losses and natural and manmade disasters that have affected many classes of business. It might be thought that the deterioration in underwriting results makes both direct companies, and particularly reinsurers, more reluctant to enter into long-term binding treaty agreements. However, the increase of competitive pressures between reinsurers in the late 1990s has led reinsurers to offer long term deals usually for a period of three years and occasionally longer, and reinsured companies have shown themselves more willing to buy into rates that appeared to be at a historically low level. Such arrangements offered the reinsured the knowledge that his reinsurance costs would be known in advance for a three-year period and would therefore be protected from any sudden upturn in rates. However, he may forego the benefit of further falls in rates unless there is a clause that would allow him to cancel and replace after a shorter period. As mentioned already, when the reinsurance market hardens after a period of underwriting losses, reinsurers are less inclined to enter into the multi-year contracts. Multi-annual contracts are generally not favoured by non-life reinsurers. Despite the competitive reinsurance market and the underwriting cycles, insurers and reinsurers acknowledge that continuity is a good thing. Pricing is and will always be a limitation to continuity, but the security rating of a reinsurer is also a prime condition for continuity. An insurer’s own security rating is dependent on the security quality of its reinsurance. If the security rating of a reinsurer is downgraded, particularly below A-, then irrespective of the number of years of a happy relationship, the reinsured may well be forced to change to a reinsurer with a higher security rating. So-called ‘downgrading’ allows the reinsured the option to cancel if the reinsurer’s rating falls below an agreed minimum such as ‘A-’. 216 6.3.3 Methods of Premium Rating The rates for non-proportional reinsurance treaties may be formulated as: 1. A flat premium, where the reinsurer charges a fixed sum of money 2. An adjustable premium, where a fixed premium rate is applied to the reinsured’s subject premium income or exposed aggregates 3. An adjustable premium, where a variable premium rate is applied to the subject premium income. The third method used to be employed extensively for working excess of loss treaties, with the premium rate being determined by the treaty’s burning cost, but it tends to be less common than method (2). 6.3.3.1 Flat Premiums v Adjustable Premiums Usually a flat premium is quoted when there is no evident link between the treaty exposure and the underlying premium income base or where the premium base is difficult to assess. One example would be where cover is given for physical damage of a motor account but the underlying premium base is based on full cover including third party experience, which is excluded from the reinsurance contract. Another example is for a short period cover where it is difficult to assess the premium base of the protected account. The major weakness of flat premiums is their inability to respond automatically to changes in the company’s business or underwriting conditions, which may adversely affect its loss experience. (The inclusion in a treaty of ‘Underwriting Policy’ and ‘Changes in Law’ clauses, may provide some protection for the reinsurer.) Therefore, flat premiums must be reviewed each year, particularly when used for treaties covering new, quite possibly rapidly expanding, companies. 6.3.3.2 Adjustable Premiums Probably the best single measure of changes in a reinsured company’s portfolio, and thus in the reinsurer’s exposure, is the gross net annual premium income (GNPI) (ie gross original premiums less returns and cancellations, and less premiums paid for reinsurances which inure for the benefit of the treaty). An increase in the GNPI will indicate that the company has either: • Insured a larger number of exposure units, or • altered its portfolio mix to include a larger proportion of more hazardous, more highly rated risks, or • increased its net retentions (although under the terms of the treaty this probably could not be done without reference to the reinsurer) • or simply increased its premium rates. Except for the final point, the reinsurer’s loss expectancy would also tend to increase. Thus, by linking the reinsurance premium to the reinsured’s GNPI, the reinsurer automatically obtains some compensation for possible changes in its potential liability. That linkage can be achieved by the reinsurer applying an agreed premium rate to the reinsured’s GNPI for the treaty year. Normally, the company pays a deposit premium at the beginning of each year. Then at the end of the year, the final premium is calculated by applying the agreed premium rate to the company’s GNPI for the year. In addition, a minimum premium is normally applied ensuring that the reinsurer earns a minimum price for the cover given. 217 Where the reinsurer’s liability is for losses occurring on policies in force on the reinsured’s books during the year, the most appropriate measure of exposure may be the reinsured’s earned premium income, which: 1. Brings into account that portion of the premiums paid on policies written during the previous accounting periods that apply to the periods of insurance run during the current year, and 2. deducts that portion of the premiums written during the year that applies to the periods of insurance unexpired at the end of the year and so may give rise to claims in future years. As can be seen from the following example, if an account is expanding, the written premium income for the year will be greater than the earned premium income. If the business is contracting, the position will be reversed. Therefore, if the reinsurance premium is based on the reinsured’s gross net written premium income, it may over or under-compensate the reinsurer for its exposure during the year. Year 1 Year 2 Year 3 Year 4 £000s £000s £000s £000s Gross net written premium income 1,000 1,200 1,500 1,400 + Unearned premium reserve at beginning of year at 40% 340 400 480 600 1,340 1,600 1,980 2,000 - Unearned premium reserve at end of year at 40% 400 480 600 560 Gross net earned premium income 940 1,120 1,380 1,440 Nevertheless, despite the distortion produced by the use of written premiums, in practice written premium income or accounted premium income is normally used for ease of calculation. Also it could be argued that written premium income responds more quickly to increases in the size of the cedant’s portfolio and thus the reinsurer’s exposure. 6.3.4 Rating Risk Excess of Loss Treaties Two basic systems are employed for rating risk excess of loss treaties: 1. Systems based on loss experience of the individual contract, and 2. Systems designed to reflect the reinsurer’s exposure to loss based on the use of generic loss frequency distributions. See Chapter 11 for a fuller discussion of how these methods relate to property business. 6.3.4.1 Rating on Loss Experience - Burning Costs The aim of the burning cost system of premium rating is to adjust regularly and gradually the premiums for excess of loss treaties to changes in the reinsured’s own loss experience. The method at best is suitable only for working excess of loss treaties that each year produce a regular flow of reinsured losses that are settled fairly quickly. In addition to property risk excess of loss, the method is commonly used for working layer motor and liability treaties where the past loss experience shows that a regular pattern of claims is to be expected. Liability claims usually take a long time to settle, which means that the reinsurance premium calculated on a burning cost basis must be adjusted