This Paper outlines the underwriting practice of the life insurance companies in Canada and the United States where, because of the importance of mortality results in the successful operation of a life insurance company, underwriting is considered to be a major executive function.

There is a discussion of general underwriting considerations and a more particular discussion of general underwriting practice. This covers the use of an Underwriting Manual, the Numerical Rating System, the use of additional information such as statements from attending physicians, special tests, etc., the question of non-medical insurance and the methods of setting up substandard classes.

Mortality investigations of various kinds are considered from the Underwriter's viewpoint which is to test the results of underwriting both in general and on particular impairments. There is a discussion of some current problems such as insuring female lives, insuring hypertensives, and the use of experimental underwriting.

The final discussion is on the importance of establishing an underwriting goal for the life insurance company, a goal which should be made clear to the agency force. It also stresses the importance of a good agency force in achieving the desired results.

In this Paper various methods of maintaining records and performing the annual costing and valuation of Group Life Assurance and Pension Schemes are discussed with particular reference to the “single premium” type of scheme.

In Part I a simple form of single premium scheme is defined and the problem to be solved is set out. Possible solutions are discussed involving handwritten cards, punched cards and electronic computers.

In Part II some of the added complications which arise in practice are introduced and a proposed system involving use of punched cards and a small-scale electronic computer is given in some detail.

Since the paper “Industrial Assurance Today” by the same author was submitted to the Faculty in January 1953, publication of the annual reports of the Industrial Assurance Commissioner has been resumed after a gap of about 15 years due to the war. These annual reports have contained more detailed figures than those shown in the previous statistical summaries, and it is possible to follow the trends in the business over a period of years.

The following trends are quite clearly visible in these statistics and are illustrated by tables made up of figures extracted from the reports :—

• (i) A fairly rapid rise in sums assured and premiums per new policy, and also a more gradual rise in sums assured and premiums for paying policies in force. The premium income has increased by 33 per cent. between 1950 and 1957.

• (ii) A steady tendency to more endowment assurances and, because such policies are usually by lunar monthly premiums, towards monthly premiums.

• (iii) Some rise in the expense ratios over the whole business, although societies are able to show decreases due to progress in the elimination of book interest and to an increase in the proportion of endowment assurance business which carries lower commission rates.

• (iv) A gradual fall in the lapse rate for the whole business from 24.0 per cent. experienced in 1949 to 16.8 per cent. in 1957. Some offices have distinctly lower rates.

• (v) Rising interest rates which, in conjunction with falling mortality rates, have resulted in increased surpluses and improvements in bonus schemes. The share of surplus allocated to policyholders in the proprietary offices is now about the same as in proprietary ordinary life offices.

Other features referred to in the paper are the contraction over a period in the number of agents resulting in a considerable increase in the average collections per agent per week. The earnings of agents have considerably increased. The gradual disappearance of small collecting societies is noted.

The paper discusses some of the questions arising in a life office from a policy of equity investment. An interpretation of the expected yield of such a portfolio is given with its implications for diversification. It is suggested that the chief criterion for the selection of individual stocks should be the quality of management. A description of this and certain other aids to selection is given together with a note on the effects of gearing.

The question of price level is referred to in its relationship with requirements of growth and income distribution, and a policy to be adopted in phasing purchases of such stocks is proposed. Certain of the effects of these methods on portfolio maintenance are mentioned, and there is a brief discussion of reserve requirements for an immunised fund.

1. The E.D.P. system which has been developed in the author's company a few years ago was based on the use of a small wired computer with a storage capacity of 30 words of 12 decimals each for data to be processed and (intermediary) results and with a practically unlimited capacity as regards the length of a programme because of the possibility of cardprogramming (48 instructions on each card). The system which in the meantime has been fully realised includes : manually punching of all data given in an application form for a new policy or in a similar internally prepared document with regard to an alteration, conversion or exit of an existing policy and automatically (a) producing of (new) policy record punched cards serving for printing a policy (if necessary) and to be filed for future use, (b) producing of punched cards for renewals, (c) idem for payment of agency commission and (d) of cards containing constants for valuation purposes, (e) updating valuation totals and totals of the accounts for surplus analysis as well as statistical totals. All data in the cards i.e. agency commission, Altenburger constants to be calculated from first principles.

2. Since a computer is available with much higher capacity (operation speed ± 64 microsec. for short operations, 500 microsec. for multiplication and division, storage capacity for data and programmes up to more than 30,000 words of 27 bits each, possibility of connecting the computer with a group of punched card machines and with a high speed photoprinter, if needed magnetic tape equipment) a new system has been developed which is based on the following assumptions :

(a) it is possible not only to do all calculations by aid of an electronic computer, but also the part of the daily work, which consists of the interpretation (in daily case work as it is centralised at the head office of the author's company) of the rules contained in a policy and its clauses ;

(b) it is possible to store information contained in or connected with a policy (including historical data) with the aid of a compact coding system on about two and a half punched cards ;

(c) it is possible to use the cards mentioned under (b) for monthly renewals, because of the given speed of the connected reading unit. This leads to the following general structure of data-processing :

(a) interpreting of the content of incoming letters and copying with the aid of a typewriter-tape punch this content on a form in a code adapted to the automatic data handling of the computer, producing simultaneously punched tape, (b) selecting the punched cards of the policy involved from the only existing consolidated file, (c) data-processing in the computer and producing automatically all documents needed.

3. The system of compact coding has been based on the possibility in the computer to interpret any combination of binary units. Since in this type of storing information data have no fixed localisation the normal programming system where operands are supposed to be found in a known address has to be adapted.