Our disciplined friend calls us, somewhat distraught, to ask us how to estimate the size of the portfolio of financial assets that will support spending during retirement, and how to determine the maximum withdrawal rate year after year.
Before we get into the details, let us take this opportunity to explain that simulation models have limitations as they try to capture the markets’ future behavior, the longevity of individuals, their spending needs, and other variables that are uncertain. However, the models are sufficiently explanatory to point us in the right direction, and from there, it is up to us to adapt it to the circumstances and decide whether we want or should be more conservative.
For this exercise, the first step would be to estimate the level of expected spending during retirement. The actual pre-retirement spending level can be used as a baseline, in this example we will use $99,000. From this expense we must subtract what is expected to be contractual income such as social security (e.g., $11,400), private pensions (e.g., $8,600), director fees, annuities, rents (e.g., $8,000), royalties, etc. Some people, especially in years past, were able to cover all their expected retirement expenses with this type of income. Nowadays, however, it is common for such contractual income to be insufficient, being necessary to resort to a portfolio of investments that was accumulated before retirement.
Once we have the level of expense not covered by contractual income (in our example, $71,000), we can use the 4% rule as a first approximation. The 4% rule was developed by Bengen (1994) through an exercise in which he used historical returns to estimate the initial maximum withdrawal rate that would be sustainable for 30 years. This rule, static and simple as it is, requires that for the first year the withdrawal should not exceed 4% of the value of the portfolio and, from there, the amount will increase according to inflation each year, that is, the amount is fixed in terms of real or purchase value. To estimate the size of the portfolio we take the net spend and multiply it by 25 (i.e., the inverse of 4%), resulting in $1,775,000 for our example.
Since its publication, the 4% rule has been tested, updated, and refined, proving to be quite stable: estimated at 3.3% in 2021, 3.8% in 2022, 4.0%, in 2023 according to Arnott, Benz and Rekenthaler (2023) . Their methodology included using Monte Carlo simulations and forecasted rather than historical returns with a 90% probability of success.
Now, is this the most sophisticated thing we can do in terms of a rule or method for setting maximum portfolio withdrawals? The answer is no. The 4% rule is just a starting point. Over time, modifications have been added that could, for example, include a diminishing level of consumption as the years go by, that is, the level of consumption at 60 years of age is not the same as at 70 or 80 years of age, since consumption tends to decrease with the passage of time. Another refinement of this rule assumes that, if the behavior of the markets has been negative, the following year’s spending will not receive the full benefit of the increase due to inflation, thus mitigating the level of erosion of the portfolio.
Guyton and Klinger (2006) added flexibility by proposing adjustments to the level of withdrawal. In those cases where the amount of projected spending is more than 20% away from its initial percentage, that is, if the value of the portfolio falls and the amount of projected spending represents a percentage that is more than 1.2 times the initial percentage, the withdrawal is penalized by 10% , and vice versa when the market rises and the percentage of projected spending is less than 0.8 times the initial percentage. This rule is also known as the Guardrails Rule because it imposes a corridor for withdrawals depending on market conditions and the remaining portfolio value. By being flexible, this method is more efficient and allows for higher levels of maximum initial sustainable withdrawals.
Another method with added flexibility requires the use of mortality tables to calculate the amount of the annual withdrawal (RMD-style), considering both the value of the portfolio and life expectancy at the time.
The study by Arnott, Benz, and Rekenthaler (2023) compares several methods and reaches the following conclusions:
- The highest values in the maximum initial sustainable withdrawal percentages are observed in portfolios with equity exposures between 20 and 40%. However, portfolios with a higher percentage in equities may end up with higher residual values after 30 years.
- Flexible rules allow for higher levels of maximum initial sustainable withdrawals, but the annual spending is more volatile, which can be a problem for people looking for stability, and the value of the remaining portfolio is lower, which can run counter to a desire to leave a bequest.
- Rigid rules often result in higher leftover portfolio values, favoring people who prefer stability in spending and leaving a bequest.
An additional item to consider is the ability to immunize part of the expenditures with contractual income by converting part of the investment portfolio into annuities (TIAA) or using laddered portfolios of inflation-protected bonds or TIPS [Arnott, Benz, & Rekenthaler (2023)]. The greater the proportion of essential expenditures (household, food, health) that is covered by income not subject to market fluctuations, the greater the inclination to use flexible rules that require adjustments in annual spending. In addition, the existence of reserve assets that have been earmarked to meet medical or long-term care expenses, for example, would facilitate the use of more flexible and efficient rules, or even relax the probability of success for higher maximum initial withdrawals.
In conclusion, the level of maximum initial sustainable withdrawals should be optimized at the individual level (Blanchett, 2024) since it will depend on the expected duration of retirement, the existence of reserve assets, the willingness to adjust the level of spending from one year to the next, the proportion of expenditure covered by income unrelated to the investment portfolio, and the desire to leave a bequest, among other variables.
Sources:
Arnott, Amy C., Christine Benz, and John Rekenthaler. 2023. “The State of Retirement Income: 2023” Morningstar Portfolio and Planning Research.
Blanchett, David. 2024. “Guided Spending Rates: Rethinking “Safe” Initial Withdrawal Rates”. PGIM DC Solutions.
Bengen, William P. 1994 “Determining Withdrawal Rates Using Historical Data” Journal of Financial Planning.
Guyton, Jonathan T., and William J. Klinger. 2006. “Decision Rules and Maximum Initial Withdrawal Rates” Journal of Financial Planning.
TIAA. Annuity Paycheck Advantage | TIAA
Disclaimer: The information provided herein is for educational purposes only. Portfolio Resources Group does not guarantee the accuracy of any tax advice, as we do not provide tax or legal advice. Consult a tax professional to make sure the recommendations are appropriate for your situation.