Friday, June 12, 2015

Good Time for Retirees to Stress Test Their Investment/Spending Strategies

Thanks to Martin from Maine for pointing me to Bear Markets! Are They a Thing of the Past, by Greg Morris in StockCharts.com and for suggesting that some other retired readers of this website might benefit from doing some stress testing of their investment/spending strategies.  In his article, Mr. Morris quite rightly sounds a warning bell reminding us that bull markets generally turn into bear markets after a period of time and indicates that, based on his analysis, we may be close to the end of the most recent bull market.  He cites lots of scary statistics about bull and bear markets and concludes, “if the average bear market lasts about 26 months and it takes an average of 56 months to get back to where it started, that translates into a little over 5 years of going nowhere and that is not including the 1929 outlier.  Don’t even think about including the declining purchasing effects caused by inflation into the equation, it would only worsen the situation.”

Martin from Maine was sufficiently disturbed by Mr. Morris’ article that he decided to stress test his personal situation by using the spreadsheets on our website and modeling the estimated projected effects on his spending budgets for the next five years assuming a drop in his assets similar to that experienced in 2007-2009 followed by a recovery in equity investments back to “where the market started” as discussed by Mr. Morris above.   This 5-year projection required Martin to perform a little more math, as he had to roll forward his investment portfolio each year to reflect assumed withdrawals as well as assumed investment performance for the 5-year period.  Martin indicated that this extra work was very much worthwhile as he felt his family could survive such a period, albeit with some necessary belt tightening.   As a result of performing several five year forecasts and seeing the results, he claims that he is better able to sleep at night, which is always a good thing.

The rest of this blog provides an illustration of how you might go about performing your own 5-year projection/stress test.  For this illustration, we will look at a retiree named Beth, who is 70 years old.  Beth has determined that her essential expenses in retirement (including taxes) are about $45,000 per annum.  She is currently receiving a Social Security benefit of $20,000 per year, so she needs her accumulated savings to provide an additional $25,000 per year (in real dollars) to cover her essential expenses.  If she has additional assets, she can use them to provide for her non-essential expenses.

Beth currently has $750,000 in accumulated savings and no other sources of retirement income.   She has no bequest motive.  Using the recommended assumptions and the “Excluding Social Security” spreadsheet in this website, she determines that designating $500,000 as Essential Assets, she can expect to withdraw almost $25,000 per year (in real dollars) for the next 20 years.  That leaves her with $250,000 to allocate to non-essential expenses and as a general reserve fund in case her investments don’t do as well as expected.  To develop her non-essential spending budget for a year, she uses all the recommended assumptions, except she assumes that future non-essential expense budgets will remain the same in nominal dollars (she does this by inputting 0% for desired future increases with respect to this component of her total spending budget).

Because Beth has this dedicated non-essential pot of assets, she doesn’t feel the need to use the smoothing algorithm recommended in this website.  Each year, she will use the spreadsheet to determine how much essential assets she will need to support her essential expenses.  If this amount is less than needed, she will transfer money from her non-essential asset account.  If this amount is more than needed, she will transfer money to her non-essential asset account.   She has decided that for years that she has to transfer amounts from her non-essential account, she will not make a withdrawal from that account.  Finally, she invests the two accounts differently.  Her asset allocation for her essential asset account is 30% equities and 70% fixed income, while her asset allocation for her non-essential asset account is 70% equities and 30% fixed income. 

Now let’s take a look at what happens to Beth’s financial situation under conditions at little bit worse than outlined in Mr. Morris’ average bear market/recovery scenario.  Let’s assume that Beth earns the following returns on her equities:  Year 1: -40%, Year 2: -15%, Year 3: 15%, Year 4: 25% and Year 5: 36%.  Under this projected investment scenario, equities will just about get back to even at the end of the five year period.  We will also assume that Beth earns a constant 3% per annum on her fixed income investments and inflation remains at 2.5% per annum for the projection period.  Beth rebalances her investments at the beginning of each year, and for calculation simplicity we will assume that Beth spends exactly her budget amount each year.


(click to enlarge)

As shown in the chart above, Beth’s first year spending budget is $61,110.  This is the sum of her Social Security, withdrawal from essential spending and withdrawal from non-essential spending.  This total is determined at the beginning of the year before the stock market tanks.  We are assuming that Beth will spend exactly this amount, but in the real world, she is likely to cut back her spending somewhat as the year progresses.

At the beginning of year 2, she determines what her essential withdrawal budget is (in this case the $25,000 budget from the previous year increased with assumed inflation of 2.5%, or $25,625).  She uses the spreadsheet on this website to back into how much her essential assets will have to be to produce this amount.  She determines that assets of $497,000 will give her a withdrawal for Year 2 of $25,632.  But, because her essential assets lost $47,027 and she spent $24,976, she needs to transfer $69,003 from her non-essential account to give her assets of $497,000 at the beginning of Year 2. 

She rolls forward her non-essential account by subtracting withdrawals adding investment return and subtracting any amounts transferred to the essential asset account. 

She follows this same process each year of the 5-year forecast, applying the assumed investment returns to the respective equity and fixed income portions of her essential and non-essential accounts.  At the end of the period, her essential assets and essential spending budgets are just about the same as she had expected under the run-out tab of the spreadsheet, but because of the adverse investment experience, she would have to cut back her withdrawals from her non-essential spending account (which under the recommended assumptions was $16,134 each year) and her non-essential assets are about 61,000 lower than expected. 

Beth sees from this exercise that there are no guarantees when you self-insure your retirement and invest in equities.  She also sees that she can weather a fairly significant bear market, but not without some adjustments.  She might consider the purchase of an annuity to cover more of her essential expenses or she may look at alternatives to reduce some of her essential expenses if experience is worse than the assumed scenario. 

This is a website for number crunchers, so I know that you can do this 5-year projection.  And maybe you will make better decisions (or sleep better like Martin from Maine) as a result of kicking the tires on your financial strategies.