J.I.A. Vol. LXVIII, p. 86 (11th line from bottom of page)

For “in the nth year of assurance”

read “in the (n + 1)th year of assurance.”

P. 74, line 13 for $100,000,000 read $500,000,000.

P. 104, line 2 for “James” read “John”.

It is probably true to say that during the past two decades the most perplexing and important of the problems, with which those responsible for the management of Life Assurance Companies have been called upon to deal, has arisen from the investment side of the business rather than from the commercial and actuarial sides. On the whole, the investment policy followed has so far emerged successfully from the searching tests imposed upon it by the events of recent years but the future remains obscure and there are many lessons still to be learned.

Life Assurance Companies, by the very nature of their business, are compelled to interest themselves in the problems of long-term investment, and it is worth while, therefore, to enquire whether the ordeal through which they have passed has revealed any defects in the present machinery of long-term lending and investing or in the basic principles which have heretofore governed investment policy.

I should like to say first that I am extremely sensible of and greatly appreciate the honour that the Institute has done me by inviting me to submit this paper. In writing it I have felt throughout the difficulty that arises from the very different approach which a lawyer makes to legal decisions from that which an actuary may be expected to make. Cases interest me because of the niceties of construction they involve or of their subtle distinctions from other cases, or again because they mark the gradual development of the judicial mind in conformity with the public opinion of the preceding generation. To actuaries the interest of cases must primarily lie in their bearing on the practical problems of life assurance work. In discussing cases, therefore, I have felt that the reaction of readers will be sometimes that I have been labouring a decision that was obvious from the start, at others that the point at issue is obsolete because no company has had such a condition in its policies for the last ten years or for some equally good reason. Knowing next to nothing of life assurance practice, I have been unable to avoid this defect, and can only ask that it be excused.

It is remarkable how little attention has been given to this subject. In the past thirty years, only two papers on expenses have been submitted to this Institute, one written by C. H. Maltby over twenty years ago (J.I.A. Vol. XLIX) and the other by J. Rietchel (J.I.A. Vol. XLIV) five years earlier.

We probably know less about expenses than about any other factor in our daily work. No doubt the chief reason for this apparent neglect is the fact that in the past we have had comfortable surplus margins in our premium rates. To-day we are calculating rates on much narrower margins as to interest and mortality, and a study of the expense problem, therefore, appears opportune.

Section 28 of the Friendly Societies Act 1896 provides for the assets and liabilities of every registered friendly society to be valued at least once in every five years but gives the Chief Registrar of Friendly Societies the power to grant exemption from valuation in suitable cases. Until 1933 this power to grant exemption from valuation was normally exercised in the case of a dividing friendly society or of the dividing section of a society with more than one section. The Chief Registrar, in his Report for the year 1933, indicated his intention of giving individual consideration, as each certificate of exemption expired, to the alternative of requiring the society to be valued. Since then many dividing societies have been refused exemption from valuation and what was formerly little more than a theoretical problem, namely, whether or not the liabilities of a dividing friendly society are susceptible to actuarial valuation, has become a practical problem to which has been added the difficulty of giving suitable advice and of explaining a technical subject in a manner which will assist the members to understand the true financial condition of the society.

The investigations discussed in this paper have been made for the purpose of examining the errors which may occur in a mortality table based on the experience of insurance offices when the Census Method of construction is used for a period which has been subject to extreme fluctuations. Hitherto, although the theory, with practical but empirical illustrations, has received considerable attention, no actual experience has been investigated. The present investigation is possible because, owing to the abnormality of the 1931-35 material for annuitants, further information was furnished by the offices.

Most actuaries would agree that the determination of the rate of interest to be employed in the valuation of a pension fund is one of the most difficult of the problems which have to be faced. It is usually possible, either from the actual experience of the staff in question or from a knowledge of the experience of other staffs which are similar in character, to form a fairly accurate estimate of the rates of mortality likely to be experienced before the pension age is reached and, with perhaps rather less certainty, the rates of withdrawal from service. So far as the rates of mortality after retirement are concerned, there is available a considerable volume of data regarding the experience of annuitants and of the pensioners of a number of large funds, and so long as a reasonable allowance is made for future improvement no great difficulty is likely to be experienced on this account. It is true that the experience of some of the larger pension funds has shown that the mortality of pensioners tends to fluctuate over a period of years, but on the whole such a standard can be adopted as will be likely to be adequate for many years.