How to handle your retirement financially
A model to manage your retirement funds
How much of your investment portfolio can you withdraw and spend annually? This is a key question for retirees and even working individuals aiming for financial planning.
The most well-known answer is the 4% rule, introduced by William P. Bengen in 1994. It suggests withdrawing 4% of your investment portfolio in the first year of retirement, and then adjusting that amount annually for inflation. During periods of low interest rates, some have advocated a more conservative 3% rule.
The 4% rule and similar guidelines rely solely on historical returns and inflation data, yet no one can guarantee the future will mirror the past. Moreover, withdrawal rates should reflect individual portfolio factors like asset allocation, investment income, taxes, and real estate holdings. A simple, "one-size-fits-all" rule may provide only a rough estimate. The biggest flaw, in my view, is the static nature of these rules. Financial decisions are typically dynamic, influenced by short-term feedback and changing circumstances. Why should retirement planning be any different?
I propose a practical model I originally developed for myself, but it may benefit others as well. Unlike static rules, this model is dynamic, adapting to changes in personal circumstances and the broader financial environment—such as shifts in interest rates, inflation, taxes, and asset valuations.
Let’s illustrate this model with a hypothetical retiree, "Joe," and his investment portfolio in the table below.
Table 1
Joe divides his assets and income sources into four categories:
High-Quality Liquid Assets (HQLA): Cash, short-term Treasuries, and liquid bonds from high-quality issuers with below 3 years maturities.
Fixed Income: Primarily investment-grade bonds with longer maturities and high-quality preferred stocks.
Stocks: A broad category that includes additional types of risky and volatile assets, such as MLPs, stock mutual funds, ETFs, and potentially risky preferred stocks and junk bonds.
Income from Non-Liquid Sources: Social Security, pensions, net rental income, fixed annuities, and similar stable income streams.
Joe excludes job income, viewing it as temporary and unreliable for long-term planning. However, you may include it in the 4th category if desired. For simplicity, Joe includes rental income in this category (perhaps he doesn’t receive any), but investors could separate it into its own category if relevant.
For each category, Joe notes its current asset value and income in two columns. Then, he calculates Reduced Income, assuming certain assets are liquidated in sequence: first HQLA, then Fixed Income, and finally Stocks. This sequence isn’t necessarily optimal and used solely for modeling.
Joe is now ready to assess his withdrawal levels, as shown in the table below.
Table 2
The top row shows a range of potential expenses. With his income from Non-Liquid Sources identified in Table 1, Joe can now calculate how much he needs to withdraw from his investments to cover these expenses (shown in the second row). For instance, if Joe’s expenses are $60,000, he needs to withdraw $41,908 from his investments, given his Non-Liquid Income of $18,092.
In the next three rows, Joe analyzes each expense level by calculating:
Withdrawal Rate: The percentage of his investment portfolio he needs to withdraw.
Investment Yield: The income generated by his investments.
Reliance on Capital Appreciation: The percentage of Stocks Joe would need to sell to make up the shortfall between investment income and projected expenses.
The key insights are in the last three rows. With the Reduced Income (excluding HQLA contributions) from Table 1, Joe can estimate how many years his HQLA assets alone would last. For example, with expenses of $60,000 and a Reduced Income of $53,752, Joe’s $176,500 in HQLA could sustain him for 176,500/(60,000−53,752)=28 years.
In calculating the final row (Years on All Liquid Assets, or portfolio longevity), Joe assumes a conservative scenario where the future value of Stocks is halved. Why this assumption? This is where we utilize historical data: from 1928 to 2018 the worst 1-, 2-, 3-, 4-, 5-, and 6-year returns from stocks are correspondingly negative 44%, 58%, 62%, 65%, 49%, and 48% (7-year and longer stock returns are significantly better).
Before proceeding, we’ll review the favorable (i.e., making the model more conservative) and unfavorable (i.e., making it less conservative) assumptions we used together with Joe.
Favorable Assumptions:
We ignore stock dividend growth, typically around 6% annually.
We assume stocks will depreciate by 50% after holding them long-term.
We disregard income from assets while selling them; for example, we assume that HQLA sustains us for 28 years without producing income during that time.
Income from illiquid sources can grow over time—e.g., rental appreciation, Social Security inflation adjustments, etc.
Unfavorable Assumptions:
Inflation: A significant factor, especially long-term.
Taxation: While generally mild on investment income, some transactions—like IRA withdrawals or sales of highly appreciated stocks—may trigger higher taxes.
Fixed Income Defaults: Low risk for high-quality assets.
Dividend Cuts: Always a possibility, though diversification across high-quality stocks mitigates this risk. In 2009, for example, S&P 500 dividends dropped by 21% but recovered quickly.
Life Events (or Vicissitudes of Life): Unexpected circumstances underscore the need for a healthy margin of safety in the model.
Overall, I believe the favorable factors outweigh the unfavorable ones, making the model fairly conservative. This model doesn’t aim for precision but provides a flexible outcome range that adjusts dynamically.
The most important results appear in the last row of Table 2, shaded in yellow. Years on All Liquid Funds (portfolio longevity) should be compared to expected life expectancy. If portfolio longevity significantly exceeds life expectancy, you’re in good shape. For instance, if Joe is over 60, he could safely spend $60,000 with a solid margin or even increase it to $75,000 with a reduced margin of safety. Spending $90,000, however, would be risky.
Interestingly, the $60,000–$75,000 spending levels roughly align with 3–4% withdrawal rates, echoing the 4% rule’s utility as a starting point. However, our model is investment-income sensitive, unlike the Bengen model. Joe can adjust his income by reallocating funds—for instance, moving from non-dividend stocks like Berkshire Hathaway (BRK.A/BRK.B) to high-dividend stocks, or vice versa. Such adjustments could create notable differences between our model and the 3–4% rule.
Beyond portfolio longevity, other figures in Table 2 are equally valuable. For instance, Joe prefers a 5-year HQLA reserve to avoid selling bonds early or stocks at low prices. He also values many stock-holding years to maximize potential capital appreciation. While there are many ways Joe could leverage the table for financial decisions, this article won’t delve into them.
The proposed model’s strength is its dynamic nature—allowing you to generate and adjust forecasts in real-time, say yearly. It factors in current asset allocation and valuation, aiding financial decisions in various scenarios. For example, if stocks drop significantly, should you buy more? Or is there too much invested in stocks? How might selling your rental property impact your finances? This model allows straightforward evaluations of these scenarios, unlike fixed-percentage rules.
The model doesn't dictate specific asset allocations or decisions but assists in risk assessment, providing a versatile tool for managing your financial future.