Friday, September 25, 2015

Another Look at Deferral of Commencement of Social Security Benefits

My retired neighbor, Leon, is turning age 62 in the near future.   A couple of days ago, when I was out performing what our dog believes is my primary purpose in life (being his personal bathroom attendant), Leon asked me for my thoughts about whether he should defer commencement of his Social Security benefit.   Like everyone else, he had read many articles from experts who strongly recommend that retirees defer commencement of their Social Security benefits.  Leon pointed me to Michael Kitces’ April 2, 2014 post where Michael said, “the decision to delay Social Security actually represents an astonishingly valuable ‘investment return’.”  Leon had also done his “breakeven” calculations. 

I told Leon that while deferring commencement of Social Security could be financially advantageous, I believed (and my prior posts on this subject have indicated) that deferral is not necessarily a “no-brainer.”  The effectiveness of the deferral strategy depends on a number of considerations, including:  1) how long you will live, 2) how much savings you will use to “bridge” the period of deferral, 3) what investment return you could earn on your savings and 4) the rate of future inflation. 

The table below shows the increase/(decrease) in the present value of a retiree’s spendable income associated with deferring a $750 per month Social Security benefit payable at age 62 until age 70 vs. commencing the benefit at age 62 assuming various ages of death.  The table uses the same assumptions and hypothetical retiree as used in Mr. Kitces’ article.  The calculations were performed using the Social Security Bridge spreadsheet from this website.  

(click to enlarge)

The table shows that under these assumptions, individuals who live longer will benefit financially by deferring commencement of the benefit until age 70 vs. commencing at age 62, while those with shorter lives will benefit financially by commencing the benefit at age 62.   It also shows that even those individuals who choose to defer commencement until age 70 and live until age 97 are not expected to be huge winners under the assumptions used to develop this table. 

The table also provides survival probabilities to the various ages based on the Society of Actuaries’ 2012 Individual Annuitant Mortality Table with 1% mortality improvement.  This mortality table (and the probabilities of survival) is available in our website in the “other calculators and tools” section.  It should be noted that this table represents mortality experience of individuals who purchased annuities and as such is more conservative (longer life expectancy) than general population mortality.   The probabilities of survival to age 97 for both males and females were not available from the tool on our website and have been estimated by me. 

One of the big differences between Mr. Kitces calculations and mine has to do with the amount of money spent by the hypothetical retiree from his accumulated savings during the eight year deferral period.  Mr. Kitces assumes the retiree will spend $750 per month in the first year of deferral and $922 in the seventh year ($750 increased with 7 years of inflation at 3% per year), whereas I have assumed that the retiree does not want to have a big jump in spendable income in year 8 and will spend the same real dollar amount of $1,672 the retiree expects to receive at age 70 during each year of the deferral ($1,320 per month in the first year and $1,624 per month in the 7th year).  The cost of deferrals (the present value of withdrawals from savings) under Mr. Kitces methodology is $65,258 while under my methodology, it is $114,853.  This is why deferring commencement looks so much more favorable in Mr. Kitces article (if you spend less today, you can spend more later all things being equal).  Of course, it would look even more favorable if the hypothetical retiree decided not to withdraw any amounts during the deferral period. 

As I told Leon, I’m not going to make a recommendation one way or another on whether retirees should defer commencement of Social Security.  It is an individual decision that involves many factors.   If you are willing to defer and don’t spend too much of your accumulated savings during the bridge period, you can generally increase your spendable income in your later years. 

Tuesday, September 22, 2015

Retirement Planning in an Uncertain World--Part 2

This is a follow-up to our post of August 11, 2013.  In that post, I included what I considered to be one of the most helpful (and most succinct) pieces of advice I have ever read regarding managing the risks involved in financial planning for retirement in today’s world.  In his MoneyWatch article of August 7, 2013, my friend and fellow actuary, Steve Vernon said,

"Step 1: Plan to support the life you want, using your best estimate of the future regarding the economy, capital markets, your life expectancy and so on.

Step 2: Be prepared in the event that your forecasts are wrong."

While the simple spending budget calculation spreadsheets (Excluding Social Security and Social Security Bridge) included in this website include a Runout tab (and an inflation adjusted Runout tab) that shows future year’s expected results based on exact realization of all the input assumptions, no changes in assumptions and spending each year exactly equal to the total spendable amount, retirees who use these spreadsheets should have absolutely no expectation that these projected future year’s results will actually occur.  They are primarily provided to show the user that the math works, and that under these totally unrealistic conditions, the amount left to heirs at the end of the expected payout period will equal the amount the user inputted on the input page. 

The fact that the future numbers in the Runout tabs will be wrong, however, does not invalidate the approach recommended in this website to determine a retiree’s spending budget.  The Actuarial Approach anticipates that a retiree’s assets and liabilities will be re-measured at least once a year to adjust the retiree’s budget for differences between actual and assumed experience, for differences between actual and assumed spending and for changes in assumptions.  This re-measurement process is essential for keeping the retiree on track.  I view this as part of Step 2 in the process Steve Vernon outlined above. 

There are lots of possible reasons why forecasts made today will be wrong (deviate from expected results) in the future.  These reasons include:

  • Differences between actual and assumed investment returns 
  • Changes in assumed future investment returns
  • Differences in actual or assumed spending
  • Differences in desired amounts to be left to heirs
  • Differences in actual or assumed rates of inflation/desired increases in budgets to keep up with inflation
  • Differences in actual or assumed longevity
  • Differences in sources of income
Each of these differences can increase or decrease the retiree’s spending budget under the Actuarial Approach.  It is important for a retiree to realize that their spending budget can and will go up and down in future years depending on these changes.  As indicated in previous posts, the spending budget determined under the Actuarial Approach (with or without applying the recommended smoothing algorithm) is simply one of several decision factors that a retiree can use in the process of determining his or her actual spending for the year. 

Depending on the proportion of a retiree’s spending budget that is derived from accumulated savings invested in risky assets, differences between actual and assumed investment returns can have a significant effect on the retiree’s spending budget.  In order to give retiree’s a sense of how such deviations from the assumed investment return can affect accumulated savings and spending budgets, we have added a new tab to the “Excluding Social Security” spreadsheet (now called “Excluding Social Security V 3.0”).  The new tab is called “5-year forecast” and the only difference between the results shown in this tab and the results shown in the Runout tab are attributable to different investment returns inputted by the user for years 1-5 at the top of this tab.  If the same assumed investment return is input for each of the 5 years as is input for the assumed investment return in the input tab, the results shown in the 5-year projection will be the same as those shown in the Runout tab.  We have also provided two graphs which highlight the differences in beginning of year account balances and total spendable amounts (excluding Social Security and other inflation indexed annuities) resulting from investment experience different from assumed.  No smoothing algorithm was applied to the results in the 5-year projection. 

We encourage you play with the “actual” investment inputs in the 5-year projection to provide yourself a better sense of the investment risk you are assuming with your current investment strategy (or strategies).  As discussed in recent posts, if you have separate budgets for essential expenses, non-essential expenses, emergency expenses, etc. and different investment strategies for these different categories of expenses, you can “kick the tires” on these separate investment strategies to see if you are comfortable with the risk you are assuming for each expense category.  

Inspiration for this post and the resulting modification of the “Excluding Social Security” spreadsheet came from discussions with John D. Craig and from work by the Pension Committee of the Actuarial Standards Board on exposure drafts of a standard of actuarial practice regarding assessment and measurement of risk associated with measuring pension obligations.  Thanks to both John and the ASB Pension Committee.  Readers who are interested in John’s thoughts on Retirement Planning may find this website to be of interest.

Friday, September 4, 2015

Time for a Mid-Year Spending Adjustment?

While I encourage retirees who use the Actuarial Approach to revisit their spending budgets at least once a year, this doesn’t mean that they can’t be revisited more frequently.  In light of recent equity market volatility, it may make sense to check the status of your accounts to see whether mid-year adjustments to your 2015 spending might be appropriate.  This post will outline a simple way to do this and will illustrate the process with an example.

In addition to showing spendable amounts payable from accumulated savings, both of the spending spreadsheets contained in this website show the amount of accumulated savings expected at the end of the year if the investment return assumption for the year is exactly realized and actual spending exactly equals the spendable amount determined for the year (which is assumed to be withdrawn from accumulated savings at the beginning of the year).  Therefore, any difference between actual and expected end-of-year accumulated savings will result from these two sources:  deviations from expected investment return and/or deviations from expected spending.  If you want to get “back on track”, you need to manage your spending or investments (or transfer money into or out of this account) so that your end-of-year assets in this account approximately equals the expected end-of-year value. 


Mary, from our June 7, 2015 post, had assets equal to $298,871 in her non-essential expenses account with a non-essential spending budget for 2015 of $19,730.  Her expected end-of-year assets in this account equaled $291,702.   As of September 4, Mary notes that the amount of assets in this account is only $260,000.  Some of the decrease in assets in this account resulted from spending a little bit more than 2/3rds of her annual spending budget and the rest resulted from decreases in the investments in her account.  As a practical matter, it doesn’t really matter the exact sources of the decrease.   Her current account balance dedicated to non-essential expenses is lower than her expected end-of-year account.

What can/should Mary do about this situation?

She has a number of alternatives:

  1. She can make no changes in her non-essential spending for the rest of the year.  She will face the issue of a potential lower spending budget next year.   She can hope her investments will rebound by year end.  
  2. She can reduce her non-essential spending budget as best she can for the remainder of the year. 
  3. She can transfer some of her assets in other spending accounts (such as her emergency account) to her essential spending account.
  4. She can pursue a combination of the above actions.
Mary knows that she can spend her retirement money now or she can spend it later (or provide more money to her heirs).  By virtue of going through this exercise, however, she knows that as of September 4, 2015 she was a little bit more than $30,000 under her target level of assets in this account with about a third of 2015 still remaining (including her spending for the holidays).  She can use this knowledge to help her make non-essential spending decisions for the remainder of the year. 

Note that Mary can use same process for years when investment return is more favorable than assumed.   In those years, “excess” assets can be transferred out of her essential spending account to her emergency fund account or some other budget account.