This well-written book is a good read that should be of interest to all British actuaries.
Turnbull sets out developments in actuarial work from the earliest life tables in the late 17th century through a series of chronological chapters. 20th-century developments are arranged by practice area: life, pensions and general insurance.
The early chapters cover the origins of actuarial work in pure probability, through to Edmond Halley’s ground-breaking mortality table and William Morgan’s introduction of with-profits policies. We read of the intellectual steps towards the first select mortality tables for assured lives in the 19th century. One omission, surprising given the author’s Edinburgh connections, is a discussion of Robert Wallace and Colin MacLaurin’s early actuarial work on the Scottish Ministers’ fund.
From the 19th century, the main referenced sources are successive volumes from the Transactions of the Faculty of Actuaries and Journal of the Institute of Actuaries. There was much debate more than a century ago regarding the higher yields on illiquid fixed-income assets compared to more liquid assets, and the extent to which this might be recognised in liability valuations. In the late 19th century and early 20th century, mortgages (secured mostly on agricultural land) comprised up to half of life offices’ assets, with the discussions about the valuation implications recorded in the pages of our actuarial journals. It is interesting to read how our forebears grappled with issues that once again challenge actuaries in the 21st century.
Older actuaries still remember the inflation of the 1970s, the erosion of the value of government gilts and the emergence of equities as a popular investment for insurers and pension funds. Turnbull reveals that that, too, is not new. Equity allocations also rose in the 1920s after the burst of inflation and currency debasement that followed the First World War.
Turnbull’s account of early debates is largely uncoloured by his own personal views. That restraint recedes in the discussion of financial economics, beginning with a startling account of Sidney Benjamin’s 1971 paper to the Institute of Actuaries. Benjamin’s random walk approach highlighted the lack of diversification between unit linked policy guarantees, and the significant chance of insurers’ failure as a result of guarantees coming into the money on multiple policies simultaneously. The then President retrospectively re-classified the sessional meeting as a private gathering, in order to bury both the paper and the subsequent heated debate. The industry, however, took fright and ceased writing this type of business. After the roller-coaster stock market of the mid 1970s, the Profession’s procrastination paid off as strong growth lifted guarantees out of the money. By 1980 a working party, including Benjamin and David Wilkie, re-worked Benjamin’s approach, with an updated stochastic model. By this stage, the guarantees had mostly expired, or were sufficiently remote, that a scientific approach to reserving was no longer embarrassing. Subsequent generations of actuaries were allowed to grow up under the impression that the Maturity Guarantees Working Party had been the first word on the matter. Wilkie’s further development of that model heralded a golden era of actuarial stochastic modelling in which both Turnbull and the writer of this review began their careers.
During the 1990s, the actuarial profession’s embracing of financial economics was surprisingly contentious. Turnbull’s personal convictions come through clearly in his account. Increasing use of portfolio theory, option pricing theory, application of Modigliani and Miller’s irrelevance propositions, and market-based values were “intelligent and open-minded”. Actuarial criticism of these developments was “perfunctory and dismissive”. The book finishes with three major areas of actuarial work: life, pensions and non-life all submitting to the inexorable logic of market-consistent valuation.
Actuarial history is a large subject; any attempt to condense it to 300 pages must involve difficult choices. In his account of the last 30 years Turnbull’s treatment is most thorough for the developments that affected his own area of work – economic scenario generators. Notorious debacles including guaranteed annuity options and the Minimum Funding Requirement are also covered in some detail. The perceptive commentary showcases the author’s expertise as an actuary as well as a historian.
No space is given, however, to insurance risk segmentation using generalised linear models, which has transformed actuarial work in personal lines insurance and is now starting to affect annuities. Innovations in longevity modelling, including the discovery of the cohort effect and the introduction of stochastic mortality improvement models are barely mentioned. Actuarial efforts in banking, infrastructure projects, health care and other wider fields are perhaps too recent to belong in a history book.
On my shelf, next to Turnbull’s tome, is a school textbook from the 1970s – the Illustrated History of the Soviet Union. It paints a picture of progress from feudalism to capitalism and then finally to socialism: that crowning achievement of human endeavour. We cannot know whether the authors privately foresaw what would happen in the following 20 years, but to put those thoughts in writing would surely have consigned their text to the same fate as Benjamin’s 1971 paper.
Like that soviet text, Turnbull’s final chapter on the future of actuarial thought veers at time into self-congratulation. We are now in the after-shocks of a 10-year-old crisis, in which hundreds of thousands lost their homes, financial institutions collapsed or were nationalised while central banks created huge quantities of virtual money in a quantitative easing gamble whose long-term effects are still unknown. Market-consistent methods assuming risk-free bonds, frictionless trading and transparent prices are confronted with the realities of sovereign risk, illiquid assets and rigged interest rates. History repeats itself. Like the British maturity guarantee writers of the 1970s, some of Europe’s insurers again find themselves with unexpectedly onerous guarantees, subject to painful valuation theories tempered by regulatory forbearance. I should have liked to hear Craig Turnbull’s view on how our existing actuarial thought will respond these multiple challenges. Maybe that will be in Volume 2.
