Saturday, June 27, 2015

You Can Save During Retirement, Too

My last few posts have encouraged readers to develop a reasonable spending budget in retirement each year by aggregating specific expense components.  For example, in our post of June 7, Don’t Just Tap your Savings; Develop a Reasonable Spending Budget, our hypothetical retiree, Mary, separated her total budget into the following components:
  • Essential non-health-related expenses
  • Essential health-related expenses
  • Bequest motive/end-of-life long-term care
  • Other unexpected expenses, and
  • Non-essential expenses
Mary had different investment and desired payout strategies for these various components.   Therefore, she treated them differently in her planning. 

After our post of June 12, Good Time for Retirees to Stress Test Their Investment/Spending Strategies, in which we discussed some statistics that predict a bear market will likely be coming at some point in the future, I received an email from Colin from Massachusetts suggesting that it might be wise for Mary and other retirees to possibly “beef-up” their “other unexpected expenses” budgets during bull market periods in anticipation of future bear markets.  I agree.

Even though retirement is generally a period of “decumulation” rather than savings accumulation, there is no requirement for you to spend all your experience gains during good times.  The concept of having a “rainy day” fund still applies to retirees.  Therefore, if you do experience gains from investment experience or from spending less than your budget, you may want to consider where you stand in the current economic cycle and put some of those gains aside in the “other unexpected expenses” component of your total spending budget in anticipation of a future bear market.

Thanks again to Colin from Massachusetts for his suggestion.  I am always happy to pass along good ideas from my readers.

Wednesday, June 24, 2015

Will “Ratcheting” the 4% Rule Make it Less Insane?

Long-time readers of this blog know that I have no love for the 4% Rule.   It has a number of deficiencies when compared with the Actuarial Method advocated in this website, including:
  • It doesn’t coordinate well with other fixed-dollar sources of retirement income such as pension income, immediate income annuities or deferred income annuities. 
  • It doesn’t anticipate a specific bequest motive.
  • There is no spending flexibility from year to year.
  • There is no adjustment process for actual experience, spending deviations or changes in assumptions about the future.
  • After the initial year, it is not based on how much assets you have or on how long you expect to live.
  • It is a “set and forget” strategy that requires the retiree to have faith that it will work in the future based on analysis of historical returns that may have absolutely nothing to do with future returns. 
  • It doesn’t accommodate a retiree’s desire to have different spending pattern objectives for different components of the retiree’s overall spending budget.
  • It requires the retiree to invest at least 50% of accumulated assets in equities, irrespective of the retiree’s risk preferences.
Notwithstanding recent research from David Blanchett, Michael Finke and Wade Pfau showing that a safe withdrawal rate for a 30-year retirement based on a forward looking model (rather than historical returns) that reflects current low interest rates and relatively high price/earnings ratios in equities is closer to about 2.4%-2.8% for a portfolio with 60% equities, Michael Kitces is back with his variation of the 4% Rule based on historical returns.  In his new article entitled, “Ratcheting the 4% Rule for saner retirement spending”, Mr. Kitces notes that based on his analysis of historical periods, the 4% Rule would have generally resulted in significant underspending that in most cases would leave “a huge amount of wealth left over.”  As a result, he proposes that a retiree increase spending by 10% of the spending called for under the 4% Rule whenever the retiree’s account balance exceeds more than 150% of the initial account balance.  He further proposes that if the account balance continues to remain high thereafter, the retiree can continue to apply further increases every three years.  He indicates that these spending increases can be “ratcheted” up without much concern about subsequent declines. 

Thus, rather than looking at the current economic environment, Mr. Kitces looks into his rear-view mirror and encourages a new retiree to invest at least 60% of her portfolio in equities and withdraw 4% of her assets in her initial year of retirement with inflation increases each year thereafter.  She can blissfully ignore actual investment experience.  If her assets become too high in the future, she can increase her spending, apparently without much concern for having to reduce spending later on. 

I have no problem with dynamic spending strategies (ones that can go up or down).  I advocate a dynamic strategy.   If you are not going to insure a significant portion of your wealth through lifetime annuity products, and you invest in risky assets, I believe you are going to have to live with a certain amount of variability in spending.  But I don’t think in today’s current economic environment, you can withdraw something like 4% of your initial accumulated savings (with subsequent inflation increases) over a 30-year period and realistically expect to never have to decrease your spending.   Retirees should be very cautious about using the 4% Rule with Mr. Kitces’ ratcheting modification.  The research by Blanchett, Finke and Pfau noted above showed a less than 60% success rate (i.e., a greater than 40% failure rate) over a 30-year period for the 4% Rule based on their forward looking model and 60% investment in equities, and that is before application of any “ratcheting” increases advocated by Mr. Kitces. 

Thursday, June 18, 2015

Managing Risks in Retirement through Diversification of Retirement Income Sources

Many of the posts and articles on this site (as well as retirement research) suggest retirees consider managing risks in retirement by adding a combination of (i) lifetime pension/insurance annuity products and (ii) withdrawals from a managed portfolio of assets to the benefits received from Social Security.  Because annuity products such as single premium immediate annuities (SPIAs) and deferred income annuities (DIAs) pool mortality risk and share mortality gains among survivors in the pool, these vehicles can be more efficient in managing longevity risk than withdrawals from a managed portfolio.  These products can also be more effective in managing investment risk.  On the other hand, withdrawals from a managed portfolio can preserve spending flexibility, provide liquidity to meet unplanned expenses, serve as a hedge against inflation, and can provide a bequest motive, which may not be as easily accomplished with annuity products.   Therefore, depending on the relative size of a retiree’s accumulated savings, the existence or non-existence (and relative size) of a pension benefit and the amount of the retiree’s Social Security benefit, the best approach to managing retirement risks may be this partial annuity (or pension)/partial withdrawal approach. 

As illustrated in several of my most recent posts, retirees may want to separate their total spending budget into several smaller components, including for example, essential health-related expenses, essential non-health-related expenses, other unexpected expenses, non-essential expenses and bequest motives/other end of life expenses.   Investment and payout strategies for these various component budgets could easily vary.  For example, retirees may select more conservative investment (such as annuity products) and payout strategies for assets supporting essential expenses while employing more aggressive investment and payout strategies for assets supporting non-essential expenses.

If a retiree elects to use a partial annuity (or pensions)/partial withdrawal approach and either does or does not separate total spending into smaller components, it is important that the withdrawal strategy applied to invested assets be appropriately coordinated with payments expected to be received under the pension/annuity in order to meet the retiree’s spending objectives.  Note that with the exception of the Actuarial Approach recommended in this website (and possibly the actuarial approach developed by David Blanchett, John Mitchell and Larry Frank), none of the more commonly-advocated withdrawal approaches make any attempt at all to properly coordinate with fixed dollar pensions or lifetime annuity products that a retiree may have.  This failure to properly coordinate is, in my opinion, a serious deficiency for withdrawal strategies such as the 4% Rule (or any of its many variations), the required minimum distribution rules (RMD), any safe withdrawal rate (SWR) approach, the Annually Recalculated Virtual Annuity (ARVA) approach or even the Guyton Decision Rules inexplicably touted by Dr. Wade Pfau, the author of the “Efficient Frontier” retirement research noted above which advocated the combined use of annuities and withdrawals. 

Two recent articles got me started on this post.  The first article, Government Policy on Distribution Methods for Assets in Individual Accounts for Retirees--Life Income Annuities and Withdrawal Rules, by my friend and former business colleague, Mark Warshawsky, compares historical outcomes of 100% investment in annuities vs. 100% investment in a balanced portfolio of equities and fixed income assets with withdrawals made under the 4% Rule.  Based on historical simulation of asset returns, interest rates and inflation, Mark concludes that the 100% annuity strategy beats out the 100% investment strategy.  He uses this conclusion to make several retirement-related government policy recommendations.  For the most part, I found this to be good research, but I would have liked Mark to include comparisons of combined partial annuity/partial withdrawal strategies, and I am not a big fan of the 4% Rule. 

The second article, by the aforementioned Dr. Pfau, entitled 7 Risks of Retirement Income Planning, indirectly makes a pretty compelling argument for combining annuity products and investments to address all of the risks discussed in this article. 

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 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. 

Sunday, June 7, 2015

Don’t Just “Tap Your Retirement Savings”; Develop a Reasonable Spending Budget

This post is a follow-up to my post of August 31 from last year, “Managing Your Spending in Retirement—It’s Not Rocket Science” in which I set forth a four step process for managing spending:
  • Step #1 --develop a reasonable spending budget. 
  • Step #2-- determine your living expenses/needs.
  • Step #3—compare results of Steps #1 and #2 and make necessary adjustments
  • Step #4—follow this process at least once a year
In this post I will use an example to point out the benefits of following this process as opposed to applying some overly simplified spend-down strategy to your accumulated savings.  I will do this by revisiting Mary, the hypothetical retiree we looked at in the May 21 post from this year. 

This past week Anne Tergesen of The Wall Street Journal (WSJ) summarized several popular dynamic withdrawal strategies currently advocated by experts as an alternative to the 4% Rule in her article, “A Better Way to Tap Your Retirement Savings”.  None of the withdrawal strategies discussed in this article consider the retiree’s living expenses/needs or suggest that the retiree make necessary adjustments if expenses/needs don’t line up with the retiree’s spending budget.  They all involve relatively simple mathematical adjustments designed to spend down accumulated savings, with such adjustments applied the same to all retirees, independently of different life expectancy expectations, different other sources of retirement income, different desires for wealth transfers, different desires for future budget patterns, etc.  In my opinion, each of the approaches proposed in this article are clearly inferior to the four step process outlined above using the Actuarial Approach discussed in this website to develop the retiree’s spending budget/budgets. 

As discussed in our May 21 post, Mary is a 65 year old single female retiree with $1,000,000 in accumulated savings, a fixed-dollar pension benefit of $15,000 per year and she is receiving a Social Security benefit of $20,000 per year.  Mary has determined that her essential expenses are $50,000 per year.  She wants to leave $500,000 to her daughter or have that money available for long-term care or other expenses near the end of her life.   She also wants to set aside $100,000 of her accumulated savings for unexpected expenses.

In addition, for purposes of today’s illustration, we are going to assume that $7,000 of Mary’s essential annual expenses are attributable to medical expenses and prescription drugs and that Mary will be establishing a separate budget (within her total spending budget) for these expenses (because Mary expects them to increase at a faster rate in the future than her other essential expenses).  Also, we are going to assume that Mary desires to be more conservative with respect to the investments supporting her essential expenses (both her health-related expenses and her non-health related essential expenses) and has therefore decided to purchase an immediate annuity of $13,000 per year (with $6,000 per year allocated to her health related budget and $7,000 per year allocated to her non-health related essential expense budget).  

Mary’s Desired Budget Patterns

Mary expects her medical expenses to increase by inflation plus 2% and she plans to have these expenses payable for 30 years (or her life expectancy if greater), the recommended payment period under the Actuarial Approach. 

Mary expects her non-medical essential expenses to increase by inflation each year and she also plans to have these expenses also payable for 30 years (or her life expectancy if greater).

With respect to non-essential expenses, Mary is comfortable assuming that these expenses will not increase with inflation and will be based on her life expectancy (initially 24 years).  She realizes that this desired pattern will produce a lower non-essential expense budget (in both real and nominal terms) for her as she ages if all assumptions are realized. 

Number Crunching Under the Actuarial Approach

As noted above, Mary decides to buy an immediate annuity that gives her $13,000 in annual benefit.  This annuity costs her $205,629 at current annuity prices and leaves her with remaining accumulated assets of $794,371 ($1,000,000 - $205,629).

Mary uses the Excluding Social Security spreadsheet available in this website to determine that $107,500 of her accumulated savings plus a $6,000 per year fixed income annuity is expected to give her an annual spending budget of $6,988 per year increasing by 4.5% per year (based on the recommended assumptions and no amounts left to heirs).  This is close to her $7,000 per year estimate of health related expenses that are expected to increase by inflation plus 2% each year. 

Mary uses the spreadsheet again to solve for how much of her accumulated savings she will need to cover her non-health related essential expenses and her bequest motive/long-term care reserve.  Her target for this calculation is about $23,000 (to which she will add her Social Security benefit of $20,000).  She enters $22,000 for the fixed annuity ($15,000 pension plus the $7,000 recently purchased annuity allocated to non-health essential expenses), the recommended assumptions and $500,000 to heirs and determines that $288,000 of her accumulated savings will give her a constant real-dollar spending budget of $23,012 annually, to which she will add her Social Security benefit of $20,000 per year to give her an expected non-health essential budget of $43,012 per year (and $500,000 to her heir) if all assumptions are realized and Mary spends exactly her budgeted amount each year. 

Mary also has a separate budget for unexpected expenses in the amount of $100,000.

After budgeting for her future health related expenses, her non-health related essential expenses (including the desired amount to be left to her heir) and her future unexpected expenses, Mary has $298,871 of accumulated savings left ($794,371 - $107,500 - $288,000 - $100,000) for her non-essential budget.  She enters that amount in the spreadsheet with a 24 year payout period and 0% annual increases (and the other recommended assumptions) to develop a non-essential spending budget for her first year of retirement of $19,730. 

Her total spending budget for her first year of retirement, then, is $69,730 plus whatever amount she decides to spend from her unexpected budget account.  Assuming no withdrawals from her unexpected budget account, her $69,730 budget for her first year of retirement comes from the following sources:

  • Social Security: $20,000 
  • Pension:           $15,000
  • Life Annuity:    $13,000
  • Savings:           $21,730
  • Total:               $69,730

The total amount Mary expects to withdraw from her accumulated savings of $21,730 represents 2.74% of her accumulated assets of $794,371 (but only 0.35% of her non-health essential accumulated savings, only 0.92% of her health budget accumulated savings and 6.60% of her non-essential spending accumulated savings. 

By comparison, because the 4% Rule is just an accumulated savings withdrawal strategy (and not a budget setting strategy) that ignores the existence of Mary’s Pension and Life Annuity, it would suggest that Mary withdraw $31,775 from savings in addition to amounts she expects to receive from her pension, annuity and Social Security.   This would result in too much withdrawn from Mary’s essential budget accounts and too little withdrawn from her non-essential budget account to meet her spending objectives.  Using the Guyton Decision Rules with a 5% initial withdrawal rate would make things even worse in terms of meeting Mary’s objectives. 

Here is Mary’s Actuarial Balance Sheet under the Actuarial Approach.  The assumptions used to determine present values are the same as the assumptions used to develop the individual budget amounts.

(Click to enlarge)

Future Adjustments

In the future, lots of things will change.   These changes will include investment returns different from expected, spending different from budgets, inflation different from assumed, changes in mortality expectations, different expectations with respect to essential and non-essential spending, different expectations with respect to desired patterns of future budgets, different desires with respect to amounts left to heirs or expectations regarding long-term care, etc.  Since Mary is  re-visiting her retirement budget every year, she will be able to change her budget to accommodate these changes.  For example, if investment experience is more favorable than assumed, she can choose to beef up her amount left to heirs/reserve for future long-term care budget if that makes sense to her rather than increase her non-essential budget.  Unlike under the spend-down strategies featured in the WSJ article, Mary will have a significant amount of flexibility dealing with these changes in the future. 

Can These Other Approaches Do What The Actuarial Approach Does?

The short answer to this question is a resounding NO.  The approaches outlined in the WSJ article are spend-down strategies.  I recommend that they not be used if you want to develop a spending budget that reflects your individual situation and spending objectives.