Enough Already With the 4% Rule

In David Ning's October 23, 2003 blog titled, The 4 Percent Safe Withdrawal Rule Declines to 3 Percent,Mr. Ning says that most investors will "do fine by sticking with a flexible version of the original 4 percent retirement rule."  His proposed modification to the 4 percent rule to make it "flexible" is to "pause the inflation adjustment when markets decline."  Of course, he doesn't say how long the pause will be required.  It is now about 5 years since the stock market crash of 2008.  Is it now ok, under Mr. Ning's proposed modification to recommence inflation adjustments?

As I have said many times in this blog, the 4% Rule is far from an optimal decumulation strategy.   And making unspecified modifications to it to address some of its flaws does not make it appreciably better.  Rather than rely on set and forget strategy that is supposed to be "safe" with respect to the risk of outliving one's assets (but may result in significant underspending), you need to periodically crunch your numbers based on your situation.  The spreadsheets and actuarial process set forth in this website make this task relatively easy.

Let's look at three different retirees, each with $500,000 of accumulated retirement assets.  Based on the spending spreadsheet in this website and the recommended assumptions discussed in my previous posts,  I will show you that if you desire constant inflation adjusted spending in retirement, there is no x% withdrawal rate that will work for all situations.

Robert retires at age 65, has no bequest motive and no other sources of retirement income (other than Social Security).  Using the Excluding Social Security 2.0 spreadsheet and the recommended assumptions discussed in my previous post, Robert can withdraw $21,725, or 4.3% of his accumulated savings in his first year of retirement.

Ray retires at age 75.  He has a deferred annuity starting at age 85 of $35,000 per year and no bequest motive.  Inputting his information into the spreadsheet and the recommended assumptions shows that Ray can withdraw $40,220, or 8.04% of his accumulated savings in his first year of retirement.

Richie retires at age 60.  He has an immediate life annuity from his company's pension plan of $20,000 per year and he wishes to leave $250,000 (in nominal dollars) to his heirs when he dies.  The spreadsheet says that he can withdraw $11,049, or 2.21% of his accumulated savings in his first year of retirement.

So, while Robert might be ok using the 4% rule, Ray may be significantly underspending and Richie may be significantly overspending if they use it.

Reasonable Assumptions and Algorithm for Simple Actuarial Process

Several individuals have suggested that I provide some guidance with respect to assumptions and the algorithm to be used with the spreadsheets and process described in this website.  The following are my brief thoughts on possible inflation and investment return assumptions to use, reflecting the current economic environment, as well as a possible algorithm to use to adjust future withdrawals for actual experience as it emerges.  Earlier this year (see post of July 12), I recommended that you plan to live until 95 (unless you are already over 85).

Inflation:  In light of an estimated investment return assumption on bonds imbedded in current immediate annuity purchase rates of a little bit more than 4% per annum (as discussed in our post of September 22, 2013) and the long-term relationship between bond returns and inflation, I would assume inflation of something in the neighborhood of 3% per annum.

Investment return:  Given expectations of inflation of 3% and expected bond returns of 4+%, I probably wouldn't use an investment return assumption much higher than 5% per annum (and would use a lower rate if most of my investments were in bonds or other fixed income or I just wanted to be more conservative in my retirement budgeting).  I know that some will argue that equities have historically yielded higher real rates of return, but they also carry higher risk of loss that I would reflect by using a more conservative assumption.

Algorithm for adjusting future withdrawals for actual experience:  I like the approach of increasing last year's withdrawal with actual inflation and then testing the result against a corridor around this year's calculated value.  The example that follows uses a 10% corridor. 
In year 1, Joe determined his withdrawal to be $10,000.  Inflation during year 1 was 3%, so his preliminary year 2 withdrawal is $10,300.  Let's assume actual experience was favorable and his calculated value (using the spreadsheet and revised input items) at the beginning of year 2 is $11,000.  Since the preliminary withdrawal of $10,300 is 94% of the calculated value, Joe would budget a withdrawal of $10,300 in year 2.

Let's assume inflation of 2% in year 2, so Joe's preliminary withdrawal for year 3 would be $10,506 ($10,300 x 1.02).  Let's assume that Joe's calculated value for year 3 is $12,000.  Since the preliminary value of $10,300 is less than 90% of the calculated value, Joe would budget a withdrawal for year 3 of $10,800 (90% of the calculated value of $12,000).  Joe's preliminary withdrawal for year 4 would be $10,800 plus inflation during year 3.

Use Our Spreadsheet to Estimate How Much Deferred Annuity to Purchase

In his September 24, 2013 article, "Why Retirees Should Choose DIAs over SPIAs", Dr. Wade Pfau reaches the conclusion that "Deferred-income annuities (DIA's) work even better than SPIA's, by providing more liquidity and better longevity protection at a lower cost."  Previously,  combinations of single premium income annuities (SPIAs) and equity investments had been Wade's Efficient Frontier champions.

Wade mentions two risks using DIAs:  1) Over or underestimating future inflation and the associated impact on a fixed annuity amount payable many years in the future and 2) running out of accumulated savings prior to commencement of the deferred annuity.

If you decide to purchase a deferred annuity with some of your accumulated savings and self-manage the remainder, you can use the "Excluding Social Security V 2.0" spreadsheet on this website to help you model future experience and coordinate the withdrawal of your self-managed assets with the fixed amounts payable from the annuity to try to achieve constant real dollar total withdrawal/annuity payments.  The Runout tabs show total combined withdrawal/payments each year under the input assumptions.   The risk of buying too little or too much deferred annuity can also be mitigated to some degree by spreading the purchases over a number of years.