After writing about retirement savings and spending for the past couple of years, I thought it was about time to take a look at what others have written on these subjects; here is what I found.

4% Withdrawal Rate

A seminal contribution to this literature was made by William Bengen (1994) with his analysis of the safe withdrawal rate (WR).  Bengen defined the maximum withdrawal rate (MWR) as the largest proportion of financial wealth at retirement that a person could spend each year during a 30-year retirement, adjusting the spending up each year by the rate of inflation, without running out of funds.  Using a portfolio comprised of 60% equity and 40% bonds and historical rates of return for 30-year periods beginning in 1926, Bengen determined that the MWR was 4%.  That is, you can comfortably spend 4% of your wealth during the first year of retirement and then increase this amount in subsequent years by the rate of inflation.  At least, this strategy would have worked for all 30-year periods beginning in 1926 and ending in 1993.

Using a slightly different approach, I have been able to support Bengen’s finding.  My approach is “Monte Carlo” simulation.  Instead of looking at actual historical sequences of annual returns, I obtain stock and bond returns by drawing randomly from an assumed distribution of returns.  When I set the parameters of this distribution (expected return and volatility) to match the historical average returns, I find that the 4% withdrawal rate strategy works fine.  Specifically, I simulate financial wealth for a recent retiree who spends 4% of his wealth the first year of retirement and maintains that level of real (inflation adjusted) spending for 30 years.  I define failure as running out of money within 30 years.  Using historical averages for expected returns on stocks and bonds, I find that the 4% rule fails only 6% of the time.  Naturally, a major issue in this type of analysis is whether it is reasonable to assume that the distribution of future returns will be similar to the historical distribution.

The 4% Rule Won’t Work Today

Indeed, this has been one of the directions for subsequent research.  Researchers have noted the potential weakness of relying on historical returns to determine the safe WR.  The historical record is limited – only two non-overlapping 30-year data points in the sample period.  Second, the US experience may be special.  Wade Pfau and colleagues (2010) have found that the 4% rule would not have been “safe” in other countries.  One extension has been Monte Carlo simulation with an estimated joint distribution of returns (as described above).  What distribution should you use?  In most applications of this approach, including mine, due to relatively low bond yields today and relatively elevated equity valuations as measured by relative high price earnings ratios and relatively low dividend yields, the expected return on stocks and bonds is lower than historical returns.  Using what appear to be more reasonable estimates of future returns, the probability of failure for the 4% withdrawal rate turns out to be rather high (see Finke, Pfau and Blanchett (2013)).  Thus, retirees should not rely on the 4% rule, at least not with today’s capital market conditions.

However, the planning literature seems to assume that even 4% represents a level of spending that is too low for most clients.  Therefore, another direction has been to seek ways to enable larger immediate withdrawals than 4% per year through the use of annuities, dynamic portfolio allocations, reverse mortgages or assuming greater risk of spending cutbacks in future years.

Convert Your DC Plan into a DB Plan

An interesting way to maximize floor income can be found in a pair of articles by Stephen Sexaur (SS) and colleagues in the Financial Analysts Journal (2012 and 2013) on how to turn your Defined Contribution (DC) Plan into a Defined Benefit (DB) Plan.  The idea is to use Inflation Indexed Treasury Securities (TIPS) and deferred life annuities to create a floor amount of retirement spending that is not subject to market volatility.  The TIPS portfolio provides a stable level of real consumption over the first 20 years of retirement, and the deferred annuity takes over from there.  Interestingly, the SS examples indicate that a floor level of consumption greater than 4% of wealth at retirement is feasible.   However, if you adopt this strategy you have given up upside potential or the ability to leave a bequest.

Another part of this literature includes ways to gradually scale down your spending in retirement (see the book The Retrenchment Rule by Gordon Pye (2012)).  In general, subsequent research has made the problem more realistic by including additional asset classes (most notably immediate and deferred annuities) and more complex strategies.  This is valuable for the well-trained financial planner, who now has more tools with which to devise optimal financial plans for clients.  But, it may be confusing to the average person trying to develop a reasonably coherent financial strategy.

Are People Saving Enough?

Economists disagree on the fundamental question of whether people are saving enough or not.  Alicia Munnell and colleagues (2014) argue that more than half of households will have to significantly retrench their spending in retirement.  They show that the age-adjusted ratio of wealth to income is more or less unchanged over the past few decades, but it should have gone up a lot due to several factors including:  fewer DB plans and more DC plans (the latter are included in typical measures of wealth but the former are not), longer life expectancy today, lesser Social Security payouts in future years, and lower expected future rates of return.

Arguing the other side is John Scholtz and colleagues (2006).  Using a formal model of optimal lifetime savings and consumption and a rich data set, they calculate optimal or target wealth for each of a large number of households characterized by age of head of household, marital status, and income level.  They then compare the target wealth levels with actual wealth levels cohort by cohort and claim that only 16% of household have under-saved, and the median savings deficit is modest.  Munnell tries to reconcile these results with her own and concludes that the differences are largely explained by different assumptions; in particular, assumptions about the optimal degree of decline in real consumption as you age.

Ratchet Rule

This brings me to the initial impetus for my interest in this area which was the problem of financial literacy, or more specifically, seeking a cure for financial illiteracy.  As a long-time executive in the financial services industry, and as a part-time university lecturer in finance and economics, it struck me that a major obstacle to financial literacy is the complexity of financial products.  Perhaps no one is fully conversant with all the intricacies and nuances of every major financial product or service.

I became interested in trying to simplify the problem.  I began with the most basic issue:  given your economic resources, how much is it reasonable to spend and save.  My solution was to estimate the total value of economic resources, including the present value of future income (I call this Total Wealth), and to identify a sustainable level of savings as the product of a small fraction times total wealth.  The Bengen 4% rule is a special case of this approach for the retiree for whom total wealth is represented by his financial portfolio, and the “small fraction” is .04.  But the rule applies much more broadly, to the recent college graduate or the mid-career executive, as well as the recent retiree.

The simple solution is a fraction of your total wealth that you can spend this year, and every subsequent year, without a significant chance of failure.  The answer will depend on many factors including your age, your willingness to take investment risk, what you mean by “significant chance of failure,” whether you have a desire to leave a financial legacy, and your expectations about future rates of return.

For example, if you desire flat or rising real consumption over time and you would like your wealth to grow, then the optimal spending rule for you is a very small fraction of total wealth, like 1% or 2%.  It is highly likely that you will be able to earn an after-tax real rate return of greater than 1% or 2% and so your real wealth will grow over time.  Then if you maintain the rule of spending 1% or 2% of wealth, you will be enabled to “ratchet” up your spending whenever wealth reaches a new maximum value.

The downside to this rule is that it implies a hefty savings rate.  Consider a young person starting out in their career.  Total wealth (comprised mostly of the present value of future income) is probably around 35 times current income, so spending 2% of wealth is the same as spending 70% of income.  This is a 30% savings rate as conventionally measured.

Or take another example.  Suppose an individual is comfortable with significantly lower spending in retirement and has no desire to leave a financial bequest.  In this case a much higher level of current spending, and lower savings rate, would be appropriate.

The Sustainable Spending Tool

There is no single strategy that is optimal for everyone.  Accordingly, my colleagues and I have developed a simple simulation tool that you can use to craft for yourself a suitable spending plan.  You can find a beta version of this tool at the CLU website (www.clucerf.org/financial-markets).  You need only input a few assumptions (current age, expected retirement age, current income and net worth, desired spending, desired bequest) and then push “GO” and you will see the probability of plan failure (defined as running out of money in retirement) along with the average bequest (wealth at the time of death).  If the failure probability is very low, then you can spend more, retire earlier or leave a large bequest.  On the other hand, if the failure probability is high, then you may want to lower your spending, delay retirement, or plan on not living so long.  It really is pretty simple to iterate towards a reasonable solution.

 

References

William Bengen.  1994.  “Determining Withdrawal Rates Using Historical Data.” Journal of Financial Planning.

Alicia Munnell, Matthew Rutledge and Anthony Webb.  2014. “Are Retirees Falling Short? Reconciling the Conflicting Evidence.” Center for Retirement Reseach at Boston College.

Wade Pfau.  2010. “An International Perspective on Safe Withdrawal Rates from Retirement Savings:  The Demise of the 4 Percent Rule?” Journal of Financial Planning.

Michael Finke, Wade Pfau and David Blanchett.  2013. “The 4% Rule is Not Safe in a Low-Yield World.”  Journal of Financial Planning.

Wade Pfau.  2011. “Can We Predict the Sustainable Withdrawal Rate for New Retirees?”  Journal of Financial Planning.

Gordon Pye.  2012. The Retrenchment Rule.

John Scholz, Aranth Seshadri and Surachai Khitatrakun.  2006. “Are Americans Saving “Optimally” for Retirement?” Journal of Political Economy.

Stephen Sexaur, Michael Peskin and Daniel Cassidy.  2012. “Making Retirement Income Last a Lifetime,“ Financial Analysts Journal.

Stephen Sexaur and Laurence Siegel.  2013.  “A Pension Promise to Oneself,“ Financial Analysts Journal.