British Actuarial Thought in General Insurance (1851-1994)
General insurance—that is, the insurance of risks other than life contingencies—pre-dates life assurance in its development as an established business activity of critical importance to Britain’s economy. Marine insurance first emerged as a significant commercial activity in early thirteenth- century Italy in ports such as Genoa. The earliest historical records of British marine insurance date from sixteenth-century London. Over the following 200 years, the City of London and its insurance institutions such as Lloyds of London developed into a leading global centre of maritime insurance. These institutions delivered dependable insurance to the cargos and sailing ships whose perilous sea voyages drove Britain’s maritime trading economy and the expansion of its Empire.
The actuarial role in the British general insurance industry has, however, been most notable by its absence for most of this history. It was only in the 1970s that the actuarial profession’s engagement with the industry moved significantly beyond the occasional passing observation. General insurance first appeared on the syllabus of the Institute of Actuaries in 1978. Actuarial professional guidance was not proposed in Britain until 1986. This contrasted with experience elsewhere in the world, most notably in the USA. General insurance actuaries even had their own actuarial professional organisation: the Casualty Actuarial Society, which was established in 1914. This greater actuarial role in US general insurance was prompted mainly by state regulation of premium setting, which created a demand for objective analysis and professional certification.
© The Author(s) 2017
C. Turnbull, A History of British Actuarial Thought, DOI 10.1007/978-3-319-33183-6_7
In comparison with life assurance, most forms of general insurance typically have several fundamental features that, to a greater or lesser degree, complicate and confound the use of the statistical methods pioneered by the likes of Richard Price and William Morgan in mortality modelling and the valuation of life contingencies. These include:
- • Defining the unit of exposure. In life assurance, the unit of exposure is straightforward: person-years. It is unambiguous, intuitive and quite easily observed and recorded. In general insurance, the unit of exposure can be much less obvious: for example, in motor insurance, it could be vehicle- years or vehicle-miles; for fire insurance, it could be building-years, or building-floors-years, or something else. And if there is a theoretically appropriate unit, it may not be easily recorded.
- • The above point is related to the heterogeneity of general insurance experience. There may be a very large number of risk factors that drive general insurance experience. Identifying and observing these risk factors may be a significant practical challenge. And once they have been identified, the resultant homogenous sub-sets of data may be very small.
- • The claims experience incurred in a given insurance period may take many years to ‘discover’. In some forms of general insurance, especially liability business, it may only be many years after the end of the insurance period that it transpires that a claim exists. So-called incurred but not reported (IBNR) claims are inherently challenging to predict or estimate. For some classes of business, the final settlement of the claim size may only occur a further many years from when it is reported.
- • Most classes of general insurance business are inherently less stable or stationary than life assurance and its mortality rates. Social, judicial, technological, economic and commercial changes can drive changes in the expected claims behaviour of general insurance that are typically faster and more material than those in life contingencies. This can place substantial limitations on the usefulness of historical experience data, even where it is homogeneous, plentiful and complete.
As a result of these challenges, general insurance pricing and reserving have historically often been determined on a case-by-case basis by underwriters and other general insurance specialist practitioners using their expert judgement and individual experience. The use of analytical methods and statistical modelling based on empirical data has often been very limited. In this setting, British actuaries have historically been uncertain of what role they can and should play in the industry. We will see below, however, that British actuar?ies have been, at least sporadically, considering how to apply their technical and professional skills to general insurance pricing and reserving since the mid-nineteenth century, if not before. This engagement grew in conviction from the early 1970s, and by the 1990s actuaries in general insurance, having finally established an uncontested industry role, were faced with similar challenges and questions as their life assurance and pensions peers: did their relative lack of technical training in cutting-edge stochastic modelling and the new ideas in financial economics represent an existential threat to their established role? Could they absorb some of these concepts into the mainstream of actuarial thought and practice? First, however, we go back to the 1850s and how that generation of innovative actuarial thinkers tackled the unique challenges of general insurance.