Final Thoughts on the Pensions Story
In the final analysis, the actuary’s traditional objective in pension fund management of estimating the stable long-term contribution rate that was expected to be required to fund ultra-long-term and economically complex forms of liability cashflow under a given risky asset strategy was an extremely challenging task. Their conclusions were highly sensitive to guestimates of non-stationary economic variables such as price inflation, real salary growth and long-term asset returns. Different economic environments came and went, each time prompting fundamental reviews of actuarial practices and assumptions. In times of particular economic uncertainty, this could become publically embarrassing. We saw how the five-year period between 1977 and 1982 produced actuarial estimates of the long-term real asset return ranging from less than zero to more than 4 % per annum. The implications of these different assumptions for the health of pension funds and their ongoing funding needs was so great as arguably to be a discredit to actuarial expertise.
Given the scale of inevitable uncertainty in the above estimates, it is perhaps surprising in retrospect that no coherent risk management framework or philosophy emerged as an intrinsic part of the apparatus of actuarial theory and practice in defined benefit pensions. By the late 1950s, actuarial thinking in the life sector, propelled by the papers of actuaries such as Redington, Hayes, Kirton, Anderson and Binns, was producing and applying new technical ideas about asset-liability risk management—matching, immunisation, ‘portfolio insurance’-style thinking in dynamic asset allocation and the stress testing of risky asset values relative to guaranteed liabilities all received serious consideration in this rich era of development of actuarial thought on financial risk management in the life sector. Pension actuarial thinking of that era appeared oblivious to such ideas and unwilling to consider how they could be applied in the particular context of defined benefit pension funds. It is worth pausing to ask why this was the case. It is perhaps linked to the historical lack of clarity and consensus on what advance funding of pension liabilities was for in the first place. In particular, a lack of focus on member security as the overriding purpose of the pension fund and on sponsor credit risk as a potential threat to that member security meant that pension actuaries perhaps did not have the same clarity of purpose as their life brethren when it came to thinking about managing financial risk.
The actuarial focus on the estimation of the stable long-term contribution rate arose in nineteenth-century actuarial practice. It was accountants and regulators that drove actuaries to consider other relevant financial objectives and metrics such as funding for security against near-term sponsor insolvency and measuring the cost of liability accrual. The profession struggled to establish in its thinking how these different demands were distinct questions that could be consistently reconciled within a wider technical actuarial framework.
Again with the benefit of hindsight, it now seems clear that the path the actuarial profession took to ensuring consistency between asset and liability valuation—off-market liability valuations with consistently off-market asset valuations, first proposed by Puckridge in 1948 and enthusiastically refined over the following 15 years—was a choice that created unnecessary confusion both inside the profession and in its dealings with others for half a century.
The treatment of equity investments permeated much of the above thinking, and was a particular source of complexity and, at times, controversy. Prior to the 1960s, the valuation treatment of equities was relatively straightforward, if basic. Equities would be valued at the lower of market value and book value, and the liability valuation paid no heed to the pension fund’s asset allocation. The papers of Heywood and Lander, and Day and McKelvey in the early 1960s strove to create an equity valuation framework that was consistent with the established off-market liability valuation basis. Things were made more complicated still when Heywood and Lander proposed that equity allocations should permit the liability discount rate to be adjusted upwards from the actuarial estimate of the long-term risk-free rate in anticipation of the long-term excess returns of equities. This contributed to the actuarial perception of a ‘free lunch’ in long-term equity investment that was ultimately only dispelled in 1997 by Exley, Mehta and Smith.
The UK dividend experience of the 1970s showed that the original actuarial rationale for pension fund equity investment—stability in real growth in dividends—was misplaced. But actuarial confidence in their ability to look through short-term irrational market value volatility for the benefit of their long-term clients was difficult to entirely dispel, despite the arguments of Arthur and Randall and others in the 1990s. However, this actuarial perspective was not so far away from contemporary financial economics as the actuarial modernisers would lead the profession to believe. The empirical program of financial market research of the 1980s developed by leading academic financial economists such as Jensen, Shiller and Roll could lend some serious intellectual weight to the traditional actuarial disdain for short-term market price volatility. Curiously, this point was entirely missed in the actuarial tradi- tionalist/financial economist moderniser debate of the 1990s. That is perhaps reflective of how this debate could generate more heat than light, and could at times enflame personal passions at the expense of objectivity and intellectual openness.
Rather like with-profits in the life sector, from the late 1990s defined benefit pension funds would start on a path of terminal decline, a structure that no longer met the needs of the modern workforce, and whose untenable economic cost was becoming increasingly apparent to employers. By the twenty- first century, British actuarial debate on how best to manage open defined benefit pension funds had become increasingly moot.