Wednesday, December 3, 2014

Revisiting Recommended Assumptions

As discussed in our posts of July 12, 2013 and October 11, 2013, and in the Journal of Personal Finance paper (originally written in February, 2014), we have been recommending the following assumptions for annual budget determinations under the Actuarial Approach: 
  • Expected annual rate of return on savings: 5% (a nominal interest rate) 
  • Annual desired increase in payments/inflation: 3%,
  • Expected payout period (in years): Until age 95 or life expectancy if longer.
In the absence of significant changes in interest rates, we recommend continued use of these assumptions for 2015 retiree budget determinations.

The remainder of this post will discuss the rationale for continuing these recommended assumptions.

Investment Return and Inflation Assumptions

In prior discussions, I have tied the expected future nominal expected rate of investment return on accumulated assets to the approximate interest rate "baked into" immediate annuity purchases.  For this purpose, I have used the immediate annuity purchase rates made available on the Income Solutions website.

According to this website, as of November 26, 2014, a premium of $100,000 could purchase a monthly immediate annuity of $571 for a 65 year old male and $544 for a 65 year old female.  Assuming a life expectancy at age 65 of 22.9 years for a 65-year old male and 24.9 years for a 65-year old female (based on the Society of Actuaries 2012 Individual Annuity Mortality Table with 1% per year mortality improvement, a link to which is available in the "Other Calculators/Tools" section of this website), I have determined that the interest rate inherent in these annuity purchase rates is about 4.3%.  This rate is slightly lower than the 4.6% rate I approximated last February using the same approach.  This change may be due to the use of more conservative mortality assumptions, declining interest rates or some combination of the two.

While I have no problem if a retiree wants to use an investment return assumption lower than 5% (particularly if the retiree is heavily invested in fixed income securities), I continue to believe that an annual 5% nominal return can be reasonably justified by retirees with relatively diversified investment portfolios.  I would caution, however, against assuming higher nominal (or real) investment returns based on increased investment in equities as those strategies carry more risk that should be reflected in the assumption.

Consistent with Wade Pfau's research, I believe budgeting should assume a real rate of return of about 2% per annum, so I am retaining the 3% per annum inflation assumption combined with the 5% investment return assumption as my recommended economic assumptions for 2015 budgeting.   


As discussed above, the Society of Actuaries has released several new mortality tables which show significant mortality improvement has taken place in recent years.  For example, life expectancy for a 65-year old male has increased by about a year under the new Individual Annuity Mortality Table (with 1% annual improvement).  Under this revised table, a 65-year old male has about a 24% probability of surviving until age 95, while this probability is about 33% for a 65-year old female.  Note, however, that this new table is based on mortality experience for individuals who buy immediate annuities from insurance companies and presumably have better than average health.  By comparison, life expectancies under the 2010 Social Security tables (with 1% mortality improvement), with experience based on essentially the U.S. population are 17.6 years for 65-year old males and 20.4 years for 65-year old females, respectively.  While some argument can be made for increasing the "live to 95" assumption by a year (at least for females), I continue to believe that this assumption is reasonable.  Of course, if you are already in your late 80s, you need to look at longer possible payment periods.

If you are the rare retiree who has a good idea when you are going to die, feel free to use your knowledge in your budgeting.  If you are like most of us, you should plan to live longer than your life expectancy, at least relatively early in your retirement.   The problem with doing this is that this increases the probability that you will die with more (unspent) assets than you desired.  To some extent, this is simply the cost of not buying an annuity.

The chart below shows projected budget amounts (in inflation-adjusted dollars) for a 65-year old retiree with $600,000 in accumulated assets under the Actuarial Approach (and recommended smoothing) using two different approaches for determining the remaining payout period.  The first approach uses the retiree's life expectancy (based on the SoA 2012 Individual Annuity Table) while the second approach uses the "live until 95" approach recommended in this website.  Investments are assumed to earn 5% per annum, inflation is assumed to be 3% per annum and the retiree is assumed to spend exactly the budget amount each year.  This chart illustrates the problem with using one's life expectancy each year and having the misfortune? of surviving. 

(click to enlarge)