A Primer on Safe Withdrawal Rates (Part 1 of 2)


By Tom Wall


Retirement income planning is a relatively new field.  Historically, workers would go to work for a company that offer a pension in return for years of service.  It was up to the company to make investments on behalf of the worker to ensure they could fund those promised benefits.  Social security would eventually help as well, and a retiree may have other modest savings.  If the mortgage was paid off by retirement, that could combine to be enough.  Around the 1980s era, the country began shifting from defined benefit (pension) plans toward defined contribution (401(k)) plans.  This resulted in employee contributions to individual accounts replaced employer-funded plans.  Employers traditionally offer matching contributions, but that is generally far less expensive than providing a pension.  But even with that shift, retirement planning wasn’t historically all that difficult.  When one could get reasonably high interest rates on CDs, savings bonds, and money market funds, retirees could invest their retirement funds in conservative vehicles and live off the interest.  If they did not invade their principal, they were generally expected to be alright.

When rates began to fall below levels that would provide acceptable conservative investment income in retirement, some advisors began to study whether it may be prudent to take systematic income out of volatile investment portfolios, regardless of the interest rates or returns they were earning.  In 1994, an advisor by the name William Bengen published his seminal study that led to what has become commonly known as the “4% rule” of retirement income.  Using stock, bond, and inflation data that began in 1926, he sought to determine what the maximum sustainable starting withdrawal rate was for each year, assuming retirement lasted 30 years.  He dubbed this rate the “safemax,” and it represented the highest amount one could draw from a portfolio in their first year, after which they would adjust it for inflation.  The hypothetical investment portfolio used was 50% stocks and 50% bonds, rebalanced at the end of each year to keep the allocation constant.  The results varied quite a bit year to year, but the worst scenario, which was a retirement beginning in 1966, showed a “safemax” rate of just over 4%.  Therefore, historically speaking, if you started with income that represented no more than 4% of your starting balance, you could have increased it each year at the actual inflation rate and still had money left in all scenarios without worrying where interest rates were or what markets were doing.  Bengen’s study was never meant to imply that 4% or any other number should be viewed as safe or tell present-day retirees what to do, but the elegance and simplicity of the “4% rule” quickly took hold as a starting point and assumption in the budding field of retirement income planning.

Since then, countless iterations of his research have been conducted to look at different portfolio composition, inflation assumptions, durations of retirement, inclusion of fees, etc.  My own research in Figure 5-1 below shows similar results, with the key differences being a 60/40 stock/bond allocation and the inclusion of common investment fees that an investor would pay today.  What stands out is how different the results were in various periods of time.  The problem with retiring in the 60’s is that it was right before a period of very high inflation, poor bond returns due to rising rates, and a stock market that was stagnant.  Assuming the need for inflation-adjusted income meant increased stress on the portfolio with ever-increasing income being pulled from it.  The result was a safe withdrawal rate of only about 3.5%, or $35,000 of income from a million-dollar portfolio.  It’s hard to feel like a millionaire living on less than $600 per week after taxes.  On the other hand, a retiree in 1982 at the beginning of the biggest bull market in bonds ever, coupled with decreasing inflation and a strong stock market could have had triple the safe starting income of the 1960’s retiree.

Figure 5-1

While this is all fun to talk about, an investor facing retirement has no clue what their scenario will end up looking like.  History can only serve as a general guide of where to start, and the investor needs to check and adjust their spending at least annually to ensure they remain on track.  That said, understanding what caused the negative scenarios to be so stark is the key to developing a strategy that can maximize retirement income while being able to weather any financial storm.

Thinking in terms of probability, the chart below can also be shown as a representation of the amount of time a given initial withdrawal rate would have been successful.  Again, success is defined as still having money left in the retirement account at the end of 30 years.

A Realistic Approach to Inflation

Recall that one of the main reasons for the safe withdrawal rates of the 1960s was due to high inflation in the period that ensued over the 1970s.  But what if, as discussed before, an affluent retiree who naturally decreased discretionary spending over time didn’t have the need for inflation-adjusted income?  Instead, they could start income much higher and keep it relatively level as inflation ate away at purchasing power while their appetite for purchasing similarly waned.  What impact would that have had on safe withdrawal rates?  As you can see in my research below, the impact is dramatic, with safe withdrawal rates increasing by 2-3% on average.  That translates to much higher initial withdrawal rates at a time in retirement that the retiree is most likely to be able to enjoy that income to its fullest potential.  As you’ll note, the “safemax” is still somewhat low, but has shifted to 1929.  This was the great depression era.  So assuming we don’t repeat the great depression, major improvements can be made to retirement income for more affluent clients who are more able and willing to make adjustments to spending along the way if necessary.

Viewed in terms of probability of success like before, one can see that a 6.5% initial withdrawal rate with level income provides a similar probability of success as 4% with level income.

A large body of research demonstrates that retirees tend to spend less over time.  The degree to which spending declines depends largely on income levels, with more affluent retirees realizing greater declines over time as discretionary spending goes down.  To provide some perspective on the numbers above, you can see in the chart below how much lifestyle one would be forfeiting by inaccurately assuming they need inflation adjusted income.  In two scenarios with similar historical probabilities of success (4% inflated vs. 6.5% level), it takes 18 years at 3% inflation for annual income to catch up to what one would have from a level source.  In periods of high inflation this crossover would occur earlier, but in most historical cases it would not have.

Taking this discussion a step further and looking at cumulative income, the inflation-adjusted scenario takes more than 30 years to catch up.  This is what I describe when I say that the financial services industry is grossly under-serving its retired clients because too often the advice is to play it safe and operate from a place of fear.  Managing client incomes based on the worst cases that history has provided means that in that vast majority of real-life outcomes, the retiree would have been drastically under-enjoying the retirement they spent decades working toward.

Still not sold on the lack of need for inflation protected income from your portfolio?  Don’t forget that almost all retirees will also enjoy income from social security.  Social security is an inflation-adjusted income source which can help mitigate the risk of high inflation in retirement, particularly for those who elect to delay benefits until later years.  In an upcoming post I’ll discuss social security and how you can leverage it for life sales.

This research is also very supportive of annuitized income, which tends to come with higher initial cash flows and level income, but with the benefit of lasting forever.  If there are other assets available to address liquidity concerns, and life insurance to guarantee any desired legacies, you can give your clients permission to annuitize.

In my next article, I’ll discuss the volatility buffer concept and how we can build on this research for meaningfully better client outcomes.