The paper by Messrs Abbott, Clarke and Treen on “Some Financial Aspects of a General Insurance Company”, submitted recently at a meeting of the Institute, dealt at some length with the accounts of an insurance office and the analysis thereof.

A reinsurance office operating in the London reinsurance market differs in so many ways and tends to be so specialized that many different aspects of analysis arise and a somewhat altered approach is essential. Lest it be thought that the area of operations under consideration is too small to warrant much attention by actuaries, let it be pointed out that recent figures completed by the B.I.A. and by Lloyd's would seem to indicate that the reinsurance offices, once Lloyd's itself and the Home–Foreign departments of the large direct-writing insurance office are included, in all of which administration is carried out in the United Kingdom, involves an annual premium income very roughly equal to that of all direct general insurance business written in the U.K. and each is roughly equal to the U.K. premium income of the life offices.

The general literature on the topic of Modern Portfolio Theory (M.P.T.) is now quite copious but in order to make this paper more self sufficient than it might otherwise be I make no apologies for repeating what is available outside United Kingdom actuarial literature. There are not very many actuarial papers advocating the use of M.P.T. which might suggest that many actuaries practising the techniques of M.P.T. have not been convinced that their work is conclusive. Moore's paper (1) in 1972 laid the groundwork for discussion of the models of M.P.T. by the profession. In 1977 Holbrook (6) discussed in his more general paper on pension fund performance the relevance of risk and return by summarizing the work of Treynor, Sharpe and Fama. There have been two recent papers from north of the border. The 1980 paper by Pountain and Fitzgerald (12) is the earlier. Clarkson's paper (16) to the Faculty contains a particularly interesting section in which he compares his own model for managing an ordinary share portfolio with the methodology of M.P.T.

The Insurance Companies Act 1981, after a relatively uncontroversial passage through both Houses of Parliament, received the Royal Assent on 2 July 1981. The main purposes of the Act were:

(a) to implement the provisions of the E.E.C. Life Directive of 5 March 1979,

(b) to enact in primary legislation the provisions of the E.E.C. Non-Life Directive dated 24 July 1973 (the provisions of this directive had already been implemented by statutory instruments under the European Communities Act 1972), and

(c) to remedy some defects in the Insurance Companies Act 1974 which had come to light in the course of day-to-day supervision, and to make other changes consequent on the implementing of the directives. Changes of particular interest to actuaries include:

(i) a requirement for annual actuarial investigations,

(ii) where any rights of long-term policyholders to participate in profits relate to the profits of particular parts of a long-term fund, a requirement to determine a separate surplus for that sub-fund,

(iii) some closing of loopholes in the requirement that no transfers out of the long-term fund should be made without an actuarial valuation and without meeting the conditions of §25 of the 1974 Act.

Group life insurance has attracted little formal discussion in British actuarial circles. Perhaps this is not so surprising. The technical problems posed by group life insurance seem relatively straightforward. Furthermore, there is little doubt that the evolution of present practice has been moulded more by pragmatism than strict theoretical development.

None the less, group life insurance is an important sector of our life insurance industry. Figures produced by the Life Offices' Association reveal that in 1980 approximately 7½ million people were covered under group insurance contracts for a total sum assured of the order £63 billion.

The development of the group insurance industry in this country has been closely associated with the expansion of private pension provision. Group life cover was seen as an inseparable, but very much secondary, adjunct to the more lucrative field of pensions business. However, over the last decade it has become more likely that a group life scheme will be tendered independently of any related pension business. Thus, the point has long been reached where the underwriting of group life business must be considered as a subject in its own right.

Unit-linked business has been with us for about 20 years. The original concept consisted of a life assurance policy under which the benefits were linked to the performance of a Unit Trust and the definition of unit pricing was therefore of no direct concern to the Insurance Company.

Subsequently, contracts were issued where the benefits were linked to the performance of an internal Unit Fund and the definition of unit pricing then became the direct concern of the Actuary.

This paper adopts a simplified approach to unit pricing and reviews some of the issues to be considered when calculating unit prices for an internal fund.

The paper does not consider the question of Unit Trust pricing nor have I attempted to provide a summary of historic and current practices or to provide any hard and fast rules, the aim being to concentrate on the principles involved.

This paper describes the use of time series analysis in the solution of a problem arising in social insurance. As part of a model which estimates the future cost of unemployment benefit the Government Actuary's Department (GAD) is required to forecast the proportion of the unemployed in future calendar quarters, who are male. The format of the paper is to describe forecasting in general terms in §1 and the particular problem under consideration in §2. In subsequent sections, the data available (§ 3), the existing forecasting model (§ 4) and alternative time series models (§§ 5–8) are described.

The everyday job of the actuary involves the estimation of a future series of events. Examples include the estimation of future streams of liability outgo and asset income in life assurance, the run-off of outstanding claims in nonlife insurance, and the future numbers of persons in a subgroup of the total population. This estimation can be qualitative or quantitative, short-term or long-term, deterministic or stochastic and will involve the establishment of a mathematical-statistical model, and the determination of the relevant parameters by an analysis of the data available.