One of the factors to be considered when planning for how long your retirement savings might need to last, is to consider both life expectancy and healthy life expectancy. The World Health organisation has published some data on Life expectancy and Healthy life expectancy by country. For retirement planning purposes it is also important to look at these figures at age 60, not 'at birth'. The relevant figures for males in various developed countries are listed below:
yrs => age:
Australia 18.2 78.2
New Zealand 17.9 77.9
Japan 18.8 78.8
UK 17.6 77.6
Germany 17.0 77.0
France 18.5 78.5
USA 15.6 75.6
Canada 18.2 78.2
yrs => age:
Australia 24.4 84.4
New Zealand 23.8 83.8
Japan 23.9 83.9
UK 23.0 83.0
Germany 21.9 81.9
France 23.3 83.3
USA 21.8 81.8
Canada 23.8 83.8
So for the average Australian male aged 60 who intends to retire at 65, it would be reasonable to expect (on average) to have 13.2 years (to age 78.2) of healthy, or 'go-go' years where spending needs might be somewhat higher to fund travel, hobbies etc. Then a slightly lower spending rate during to slow-go and no-go years to age 84.4 (ie. another 6.2 years).
Of course past lifestyle, genetics, and behaviour during retirement and luck (eg. terminal illness) could make actual healthy and total lifespan vary considerably from 'the average'. So it is really just a reasonable 'guestimate'. I find it quite amusing that financial planners often plug in national life expectancy (sometimes with an extra 5 years added on 'just in case') when doing Monte Carlo simulations to provide a 'probability of success' figure for a specific retirement starting balance, asset allocation, and estimated average return and std dev. The 'simulation' of 1,000 'runs' will then spit out a figure like '96%' probability of success, meaning that in 4% of the simulations the retirement funds would be exhausted before the 'end date'. But if one also takes into account the uncertainty in the 'end date' (ie longevity risk) such probability figures are really have massive 'error bars'. The plots of the simulations often show the average final balance and the range (and perhaps top and bottom quartiles), but generally do not show the impact the variability in life expectancy will have on the range of 'projected' outcomes.
Subscribe to Enough Wealth. Copyright 2006-2024