Death or Bust? The Risk With Post-Retirement Models

CheckRisk has released ground-breaking research on Post-Retirement modeling. 
For further information on the paper please contact CheckRisk on +44 (0)1225 422 583

With an aging population the risk of surviving longer than your pension pot is increasing

Ross Pepperell, Director Research CheckRisk, who led the research team on the project says that “Our research has exposed that there is a significant risk of retirees running out of money before death; the Death or Bust problem. Urgent attention is required by the industry and regulators to deal with the cause of this fundamental problem by considering the current methods used by the financial industry in post retirement calculations.

Nick Bullman, Chairman CheckRisk, said “there is too little or no pre-retirement interaction with advisers and yet retirees are then expected to be fully informed rational investors post retirement. With an aging population the risk of surviving longer than your pension pot is increasing.” He added, “the use of non-fit for purpose models to calculate post retirement
funding is common place, retirees must be made aware of the issues in terms that allow them to make responsible and informed investment decisions.”

The issue is pressing as the population ages, and greater numbers reach over 90 years old. According to a government paper, between 2015 and 2020, a period when the general population is expected to rise 3%, the numbers aged over 65 are anticipated to increase by 12% (1.1 million); the numbers aged over 85 by 18% (300,000); and the number of centenarians by 40% (7,000). (source:

Early findings show:

  • There are various techniques used to plan for income in the post-retirement years. Some of the models are not fit for purpose and may be giving a false sense of security. A lack of consistency exists across the pension planning industry.
  • Simple methodologies and rules of thumb lead to poor outcomes. Specifically deterministic projections that use fixed rates of return (for example a 5% constant rate of fund growth) for illustrations do not adequately express the risk to clients and may give a false sense of security.
  • Simple analytic formula and historical backtesting are poorly suited to income risk modelling as they are unable to accurately model the stochastic nature of portfolio returns by underestimating the risk (probability) of ruin (running out of money) and the age at which this happens. In the case of analytic formulae, they are too simplistic and distort the perception of reality in to predictable behaviour by the use of average returns. The world does not work like this.
  • As for historical backtesting, one issue is that you have no way of knowing whether the future will be like the past – so called ‘Knightian uncertainty.’ The other issue is that backtests can be manipulated to give results that suit the experimenter’s own bias. This may be subconscious, but is easily and commonly done. The sample size on which the backtest is performed is key.
  • Both of these methods also do not take into account the ‘sequence of returns’ risk. This is the risk of achieving a string of poor returns, especially in the first ten years post-retirement, that has a significant impact on the size of the pension fund and future expectations of income. In the case of historical backtesting, you only see the sequence of returns of the past. Unfortunately, both of these methods are still routinely used and misused in the financial industry.
  • The risk to retirees of surviving longer than their pensions appears to be much higher than previously estimated.
  • Product providers have a role in innovating new ideas to help clients and advisers to achieve better outcomes, and in mitigating these risks.
  • The consequences to society of getting Post-Retirement income calculations wrong are severe.
  • Regulators should, as a matter of urgency, set model boundaries within the current framework of competition, to ensure that the risks to retirees are not being misrepresented. Regulators may wish to review the use of non-fit for purpose methodologies which are in use and often common place.

CheckRisk has concluded an industry-wide research paper considering post-retirement model risks. The research indicates that in many instances the models currently in use are not fit for purpose and underestimate the risk of retirees surviving longer than their pensions. The consequences are far-reaching, not only for the retiree who goes bust before death but also for the State, the regulator, and pension providers and planners.

CheckRisk’s study considered eight key methods, currently in use for post–retirement planning that ranges from simple analytic formula models to complex Economic Scenario Generators. CheckRisk also identified areas where those involved in pension planning and provision can significantly enhance their approach. In addition, CheckRisk believes that the regulator may wish to consider regulation in the area of model use as there is a wide disparity of outcome depending which model is used that might impact individual retirees.

For further information on the paper please contact CheckRisk on +44 (0)1225 422 583