If the new figure was lower, however, we reduced the spending – but not below a predetermined floor (for example a 2 per cent reduction from the previous year’s level).
The lower the floor, the greater the likelihood of not depleting retirement wealth. But, as we shall see, even a 2 per cent floor makes a sizeable difference over the long term.
This process was repeated every year, with the withdrawals reducing or increasing according to the changes in the portfolio’s value.
Dynamic spending in action
A dynamic-spending approach does mean that retirees may go through leaner times than they might wish when markets are falling – especially when markets decline for several years in succession.
But our analysis of the major bear markets of the 20th century showed that this should be set against the complete elimination of the risk of portfolio depletion.
When we modelled retirement-portfolio performance in all six of the 20th century’s major bear markets, we found that dynamic spending removed the risk of portfolio depletion – even for those who retired into the teeth of the Great Depression or into the tumult of the late 1960s and the 1970s.
These periods represented the greatest risks in the data we looked at.
The dynamic-spending process managed to maintain overall retirement income at largely the same level for both those who were fully exposed to the bear market (that is, those who retired just as it began) and those who were only partially exposed – because they retired either at the tail-end of a bear market or because they retired a year or so before it began and so were cushioned by the last months of the bull market.
This is not to say that dynamic spending entirely offsets the bad luck of retiring into a bear market.
On average, the income available over the 35-year period was around $20,000 after adjusting for inflation, which would be 21 per cent below the initially planned amount of $25,000 per year. This is broadly comparable to the experience of those who used a fixed-spending approach.
But where dynamic spending made a crucial difference was in removing the risk of depletion (that is, running out of money) and in allowing a much higher potential bequest.
Under a fixed-spending approach, our modelling showed that the average ending portfolio value was $44,000 for full-bear-market retirees.
By contrast, the average value for those taking a dynamic-spending approach was $229,000 – 38 per cent of the portfolio’s initial value rather than just 9 per cent.
