The "matching" of liabilities has been referred to and needs a little more explanation. Whenever a life insurance policy is sold, a company takes on a liability, that of paying out the sum insured. The likelihood of this happening at any point in time can be determined from the mortality tables by actuarial calculation. Thus a total of gross liabilities for all current policies can be arrived at over the term of these policies. Since these gross liabilities are spread over a number of years, while the value of assets held today does not include the future premiums to be received, a number of adjustments have to be made to compare the two. Future gross liabilities and premiums are each discounted from their expected date to the present at a rate of interest which should be either the rate the company is earning on its fund or that which it expects to earn over the period. The difference between these discounted amounts, known as the amount of actuarial liabilities, must always be less than the total of assets, otherwise the company is insolvent.
It is in fact possible for a company to match its actuarial liabilities exactly by investing in guaranteed investments such as gilt-edged securities to the extent necessary. Then the liabilities arising through claims each year would be exactly matched by the value of the securities held. "Risk" investments such as shares and property would be purchased only with the bonus loading portion of premiums paid by with-profit policyholders. In practice, however, this degree of matching is not necessary, for three reasons. First, the actual amount of claims in any year will diverge to some extent from the actuarial expectancy, which reduces the merit of exactitude. Secondly, claims are not usually in fact met by selling investments but out of current income from the investments of the life fund, so that variations in the capital value of the investments (within limits) need not affect the ability of the company to meet its claims. Thirdly, so long as the company has adequate reserves, it can afford to "mismatch" to some extent with the aim of increasing the overall investment return - if there is a shortfall, the reserves can bear it. In the long term (which most liabilities are) the chances are that investment in risk assets such as shares and property will produce a better return than investment in more guaranteed assets.
The extent to which an office can safely mismatch therefore depends on the type of policies it is selling, the amount of free reserves it holds, and the relative success of its investment management. Successful mismatching will not benefit holders of term assurances and non-profit policies but will add to the benefits of with-profit policyholders (and also increase the company's free reserves). It is, in fact, the life offices with the largest free reserves and numbers of with-profit polices that have devoted most time to the art of successful mismatching, and this is part of the secret of success of a good with-profit office.
A number of other factors also enter into this equation, including the company's tax position, the type and design of the policies it sells and the investment policy it adopts (if it makes unsuccessful investments, this clearly worsens its position).
But the outcome of all these factors is that for any company at any time there is an optimum rate of growth. If it grows too fast, its expense ratio will rise and it will suffer from "new business strain"; if it grows too slowly, its expenses ratio will also rise, though in this case there will be no new business strain.
Large... see: Other Factors