Intertemporal vs. Intergenerational Risk Sharing: Effects of Return Smoothing in Life Insurance

In traditional (i.e., not unit-linked) life insurance, it is common practice that many policyholders participate in the same pool of assets. An individual policyholder’s return does not directly depend on the assets that were bought from her premiums, but rather on the return of the total asset pool. Moreover, certain mechanisms are typically used to reduce the volatility of this return. E.g., in Germany, "buffers” exist on both sides of the balance sheet that can be used to reduce the volatility of returns of traditional life insurance contracts. In typical UK-with-profit products, often a return is credited to policyholders that depends on some average performance of the corresponding assets over the last years.It is worth noting that these mechanisms can vary heavily from country to country. Moreover, whilst there exists regulation (that may limit the size of certain buffers or regulates that money which is taken from such buffers must be credited to policyholders’ accounts) often some management discretion remains.As a consequence, there exists a very large variety of different return smoothing mechanisms some of which are rather complex since they may depend on complex accounting rules as well as management rules.

We argue in this paper, that the concrete design of a smoothing mechanism can have a significant effect on the resulting product. We illustrate our argument by analyzing two illustrative return smoothing mechanisms. Our main result is that a mechanism which is primarily based on crediting average historic asset returns results in a significant reduction of pathwise volatility of a contract but has hardly any impact on the volatility of the terminal benefit. Hence such mechanisms primarily lead to an intertemporal smoothing of a contract’s annual return but results in very little intergenerational risk sharing (i.e., risk sharing between policyholders who enter or leave the system at different points in time). In contrast, a mechanism that uses buffers which are built up "in good years” and are used to increase returns "in bad years” results in a much higher degree of intergenerational risk sharing. This effect is even larger if buffers are not only used to increase returns in “bad years” but rather to increase the value of contracts that have a “poor value” at or close to maturity.This result has important consequences: First, simple generic mechanisms that are often used in academic papers may not adequately cover the effects resulting from concrete return smoothing mechanisms applied in practice. Second, if mechanisms prevailing in certain countries are modelled in great detail, the results may not be applicable to other countries where other return smoothing regimes prevail.