My interest in this area was originally stimulated by David Blake in the late 1990s, when he suggested that we should build a ‘PensionMetrics’ model - a stochastic simulation model of a defined-contribution (DC) pension scheme. These are pension schemes in which the individual pension plan member bears all the risks on their own, as opposed to defined-benefit (DB) schemes in which pension benefits are typically determined by years served and final salary.
I then started working with David and Andrew Cairns and we published the first (we believe) PensionMetrics model in 2001 (“PensionMetrics 1”) in 2001. This was a model of the accumulation phase (or pre-retirement phase) of a DC pension plan, taking the plan member up to the point of retirement and assuming that the plan member simply converts his/her pension fund into an annuity at the point of retirement.
Other papers then followed:
Alistair Byrne then joined us and two further papers followed:
The methodology of DC modeling is set out in "Good Practice Principles in Modeling Defined Contribution Default Funds" published in March 2013.
More information on PensionMetrics can be found in our PensionMetrics website, www.penionmetrics.net.
A working demo can be found on the PensionMetrics website at www.pensionmetrics.net/online-demo.html.
Different versions of the PensionMetrics model have been presented at seminars at the BSI-Gamma Foundation, the OECD, and at various major UK financial institutions. PM models have also been used in consultancy projects involving the Armenian Central Bank, ITN, the Money Advice Service, the World Bank and other organisations.