The great bulk of life reinsurance is transacted on the basis of proportional reinsurance as described above. However, in recent years increasing interest has developed in an approach that relates the reinsurer�s liability to the mortality experience on all or a specified portion of the primary company�s business, rather than to individual or specific policies of insurance. Widely used in property-casualty insurance, this approach is referred to as nonproportional reinsurance, since the proportion in which the primary company and the reinsurer will share losses is not determinable in advance. This type of reinsurance coverage is available from both American and European reinsurers in three forms: stop-loss reinsurance, catastrophe reinsurance, and spread-loss reinsurance.

Stop-loss reinsurance is highly developed in casualty insurance but it is still in a growth stage in life insurance, serving primarily as a supplement to conventional reinsurance rather than as a substitute for it. Thus plans follow no fixed pattern. In essence, however, stop-loss reinsurance arrangements undertake to indemnify the primary company if its mortality losses in the aggregate, or on specified segments of its business, exceed by a stipulated percentage what might be regarded as the normal or expected mortality. The agreements commonly invoke liability on the reinsurer�s part if the primary company�s aggregate mortality exceeds by more than 10 percent the "normal" mortality, which, of course, must be defined explicitly�or implicitly�in the agreement. Normal mortality is usually defined as a specified percentage of the tabular mortality for the categories of business covered by the agreement. Thus if the mortality under the policies subject to a particular stop-loss reinsurance agreement is running around 50 percent of the 1980 CSO Table, the agreement might stipulate that the reinsurer will absorb all losses in excess of 110 percent of the normal level of mortality, defined as 50 percent of the 1980 CSO Table rates. Another way to express the reinsurer�s obligation is to stipulate that the reinsurer will indemnify the primary company for all claims in excess of a specified percentage of tabular mortality, such as 60 percent of the 1980 CSO Table.

Under some agreements the reinsurer indemnifies the primary company for only a specified percentage (for example, 90 percent) of the excess mortality, an arrangement intended to encourage careful underwriting by the primary company. Under most agreements the reinsurer�s liability during any contract period, generally a calendar year, is limited to a stipulated dollar amount. Under any of these arrangements, if the mortality for the contract period is below the level at which the reinsurer�s obligation would attach, the reinsurer makes no payment to the primary company.

This approach to reinsurance lends itself to great flexibility, since the agreements can be written to cover only selected portions of the primary company�s business with varying levels of mortality and with varying duration periods. The premium for stop-loss reinsurance is arrived at by negotiation and involves the use of highly refined actuarial techniques, as well as a large portion of judgment. The basic appeal of this coverage is that it provides protection against adverse mortality experience arising out of an unexpectedly large number of small claims or an unexpected increase in the average size of claims. It is a form of reinsurance on the amounts at risk retained under conventional reinsurance agreements. Since the unit cost of protection under this approach is less than under proportional reinsurance, a company can reduce its total outlay for reinsurance by increasing its retention limits under conventional agreements and reinsuring the retained amounts under stop-loss arrangements. Another advantage of this approach is its relative ease of administration, attributable to the absence of individual policy records.

Adherents to conventional (proportional) reinsurance arrangements see many practical disadvantages to stop-loss reinsurance. They point out that it is short-term, rate-adjustable, cancelable coverage available under conventional arrangements. They call attention to the limit on the reinsurer�s liability, as well as exclusion of the war risk. They emphasize the restrictions on the primary company�s underwriting practices necessarily imposed by the reinsurer. Finally, they question for a number of reasons whether any cost savings will, in fact, be realized in the long run.

At its present stage of development in the United States, stop-loss reinsurance serves primarily as a supplement to conventional reinsurance arrangements rather than as a substitute for them.

Like stop-loss reinsurance, catastrophe reinsurance was first developed for property-casualty insurance lines. As its name implies, it usually provides for payment by the reinsurer of some fixed percentage, ranging from 90 to 100 percent, of the aggregate losses (net of conventional reinsurance) in excess of a stipulated limit in connection with a single accident or catastrophic event, such as an airplane crash, explosion, fire, or hurricane. Catastrophe reinsurance is clearly intended to serve only as a supplement to proportional reinsurance agreements. The level of losses at which the reinsurer�s liability attaches may be expressed in terms of dollar amount or number of lives. The contract usually covers a period of one year and limits the reinsurer�s liability for that period. The coverage is attractive to insurance companies that have a concentration of risks in one location, such as might arise under a group insurance policy. The risk involved is essentially accidental death attributable to a catastrophic occurrence.

While the reinsurer�s liability under a catastrophe type of agreement is high, the probability of loss is low. Hence the premiums for this type of coverage are generally low. Moreover, being based on reasonably adequate data, they are more readily calculated than premiums for other forms of nonproportional reinsurance. The expense element of the premium is minimized through the use of aggregate reporting procedures.

A final type of nonproportional reinsurance is the spread-loss reinsurance. Under this type of agreement, the reinsurer collects an annual premium of a stipulated minimum amount, of which a certain portion (such as 20 percent) is allocated to expenses and profit, with the balance credited to a refund account until the account reaches a specified maximum figure, such as the sum of 3 years� premiums. During any calendar year when the primary company�s aggregate death claims (net of conventional reinsurance payments) exceed a specified limit, the reinsurer pays the claims in excess of the limit but adjusts the premium to reflect the claims experience. The agreement provides that any amounts paid by the reinsurer for a given year, plus 20 percent, must be returned to the reinsurer by the primary company during the next 5 years.

The spread-loss agreement can be terminated by either party, with proper notice, at the end of any contract year, except that the primary company cannot terminate the arrangement under circumstances that would cause a loss to the reinsurer. In other words, the reinsurer must be permitted to recover all payments made to the primary company. It is apparent that the main purpose of this type of reinsurance is to spread the financial effects of an unfavorable mortality experience in any one year over a period of 5 years. About the only risk the reinsurer takes is the continued solvency of the primary company. Consequently the mathematical basis of the premium charge is completely different from the other two forms of nonproportional reinsurance.