Thursday, March 31, 2016

Planning for Constant Real-Dollar Spending in Retirement—Is It Setting the Bar Too High?

Determining how much assets you will need in order to afford to retire is not an easy exercise.  It involves making many assumptions about the future, and as Yogi Berra said, “it is tough to make predictions, especially about the future.”

In his article, Why Most Retirement-Income Plans Lead to Over-Saving and Over-Working, Jonathan Guyton argues that the use of static withdrawal approaches that are designed to keep withdrawals constant in real dollar terms from year to year overstate future spending requirements/ future desired spending levels and thus overstate how much savings is needed to retire and to meet future spending objectives.  Guyton points to research by David Blanchett that shows real dollar spending generally decreases with age at least until very old age when medical related costs kick in.

While over-estimating future spending requirements/desires is certainly one way to inflate estimated assets needed for retirement, there are other ways to do this, including:

  • Underestimating future investment returns 
  • Overestimating future inflation 
  • Overestimating the period of retirement 
  • Underestimating the value of retirement assets 
  • Underestimating actual future spending vs. budgeted spending
Thus, it can be somewhat risky for a retiree to focus on one planning assumption as a justification to reduce total retirement savings.  Starting retirement with lower assets and not adjusting spending will increase the retiree’s chances of having to reduce spending levels later on.  Having said that, I don’t have a big problem with Mr. Guyton’s proposed solution (Approach #1) of using different assumptions for the future expected growth (and possibly the expected period) of essential and non-essential expenses when calculating the assets a retiree may need to support such expenses.  In fact, the new tab in our Actuarial Budget Calculator spreadsheet, “Budget by Expense Type”, allows users to do just that (and not just at retirement, but throughout retirement as conditions change). 

Mr. Guyton brushes aside the issues of how possible long-term care expenses and increased medical costs might affect current spending budgets, and he ignores unexpected expenses completely.  It is important for retirees to make sure that these possible costs are adequately addressed before moving on to how to deploy their assets for essential and non-essential budgets.

Readers may wish to use our spreadsheet to model the example Mr. Guyton’s includes in his article.  If I input $1,750,000 in accumulated savings, $40,000 in Social Security benefits, our recommended assumptions and no bequest motive, the spreadsheet calculates a present value of future spending budgets of $2,669,494.  Going to the Budget by Expense Tab and inputting 0s for long-term cost and unexpected expense reserves, $80,000 for essential non-health expenses (and 2.5% annual increases), $15,000 for essential health expenses (and a 3.5% annual increase), the remaining present value for non-essential expenses would be $437,712, or $25,715 per year (with 0% increases) over a 30-year period.  Thus, the total initial spending budget would be $120,715, which is approximately the couple’s initial desired spending budget.

In order to reach the couple’s initial desired spending budget with initial accumulated savings of $1,600,000, I would have to lower the future increase assumption for essential health related expenses from 3.5% to 2.5% and reduce the period of non-essential expenses from 30 years to 20 years.  As indicated in our post of November 30, 2015, it is certainly possible to change the assumptions about the future to achieve/rationalize a higher current spending budget.  In the end, however, it is up to you to decide what is reasonable for your particular situation and how much risk you are willing to assume.