The Case for a 0% Allocation to Bonds in Retirement
An in-depth case study on how volatility does not equal risk
Should you use your age to determine your asset allocation? Conventional wisdom says yes. We say no — emphatically. Using some simple simulations and predicated on the critical assumption that a retiree has set aside two years of living expenses in cash, we make the case for a 0% allocation to bonds in a retiree’s portfolio. If you have no liabilities, then the corresponding allocation to equities should be 100%.
Simulations showing how various portfolio allocations performed for someone retiring at a stock market peak, such as 2000 or 2007, can be misleading and can favor bonds for two reasons. First, the unprecedented 30-year bull market in bonds beginning in the mid-1980s, during which interest rates dropped from 15% to 3%, has fully run its course. Total returns during this period skewed so far from their longer-term averages that they are useless as guides for prospective returns. Second, these simulations typically apply the “4% rule” as the amount of the portfolio to be liquidated annually — meaning equities are sold automatically under all circumstances.
A reserve fund approach is a better solution to effectively solving what is known as the “sequence of return” problem. Setting aside two years’ worth of living expenses enables the suspension of equity withdrawals if a long or steep bear market strikes relatively early in retirement. This is explained in detail below using a case study titled “Volatility Does Not Equal Risk.”
In modern portfolio theory (MPT), risk equals volatility, as measured by the standard deviation of returns. From a theoretical perspective, the issue boils down to a simple question: Do you believe the central underpinning of MPT, that volatility and risk are one and the same? We do not.
In this case study, we will provide a different perspective and definition of risk. This definition of risk is derived from two core principles: One, money is purchasing power; and two, the safety of an asset class is its ability to preserve purchasing power over time. Thus, the relative safety or riskiness of a particular security or asset class can be measured by the reduction or increase in purchasing power over time, not by short-term fluctuations in price.
To maintain purchasing power, and assuming the cost of living increases 2.25% per year due to inflation, a 65-year-old spending $100,000 per year after taxes would need $149,094 at age 75 (a 25% increase), $187,261 at age 85 (a 56% increase) and $233,927at age 95 (a 95% increase). This is the quantification of longevity risk — an increasingly common concern — and ignores the accompanying potential peril of an increase in inflation.
Using financial planning software, we ran simulations to compare the hypothetical ending portfolio values for three different asset allocation scenarios. (For your own retirement, you’d want your financial to run these scenarios based on your unique facts and circumstances.)
We begin by making a series of assumptions. James and Penny Bond are married, both 65 years old, in good health and recently retired. Based on their actuarial joint life expectancy, the Bonds need their money to last for nearly 30 years, to age 94.
We further assume that James and Penny each have $1 million in a retirement account, for a total of $2 million, and they have no other assets or liabilities. Their annual living expenses are $120,000, they receive $24,000 in combined annual Social Security income and both these amounts increase annually at a rate of 2.25% due to inflation. Finally, we assume the net annual return of intermediate-term bonds and stocks are 5% and 9%, respectively.
The tables accompanying each scenario below detail the inflows (Social Security income and investment earnings), the outflows (taxes and living expenses) and the corresponding ending portfolio values, displayed in 10-year increments until death. The portfolio withdrawal rate is the sum of Social Security income less outflows, divided by the preceding year’s ending portfolio value.
(Note: the tax calculation is robust, and the annual amount is largely a function of the annual withdrawal from the tax-deferred accounts, which is treated and taxed as ordinary income.)
Scenario 1: 40% Stocks/60% Bonds (6.6% net hypothetical return)
The “100 minus age” rule of thumb is intended to systematically reduce portfolio “risk” by steadily reducing the allocation to stocks over time, with 100 – age being the percentage allocation to equities. For simplicity, this scenario keeps the allocation at 40% stocks, typical for a 60-year-old, rather than reducing it as time goes on (as is the case in a “target-date fund”).
In this simulation, the Bonds would run out of money at age 92. If they followed the “100 – age” rule, their allocation to bonds would be progressively higher over time, their return lower and they would run out of funds even sooner.
Scenario 2: 60% Stocks/40% Bonds (7.4% net hypothetical return)
Here, we flip the allocations, and the Bonds end up with a final portfolio value of $755,667. Their money outlives them, but just barely. Considering their ending portfolio withdrawal rate of 24.4%, it’s dangerously close.
Scenario 3: 90% Stocks/10% Cash (8.3% net hypothetical return)
In this portfolio, the Bonds have set aside 10% of their $2 million in a cash account. This strategy results in an ending portfolio value of $2,639,536 — nearly 3.5 times the ending value of the 60% stocks/40% bonds portfolio. If the market takes a sudden downturn, they can draw their living expenses from that account rather than from their equities fund (though this is not modeled here).
Some people don’t need to set 10% of their portfolio aside in cash. For those with a pension or other income that covers their living expenses, a 100% equities portfolio might be more appropriate. This same $2 million portfolio allocated 100% to stocks would have an ending value of more than $4.2 million.
The charts below illustrate the magnitude of differences in the ending portfolio values under the three scenarios.
Bond Return Assumption Is More Than Double Current Yields
Based on research from Vanguard and Princeton professor/author Burton Malkiel, and looking at rolling 10-year periods going back to 1906, about 90% of the subsequent returns on bonds are explained by their initial yields. In “A Random Walk Down Wall Street,” Malkiel says, “Over the long run, the yield that a bond investor receives is approximated by the yield to maturity of the bond at the time it is purchased.” Notwithstanding the recent bond bull market outlier over more than 100 years, capital gains and losses have had almost no impact on return.
The Barclays U.S. Aggregate Bond Index, the benchmark for publicly traded government and corporate bonds, had a yield of 2.14% on August 30, 2019. In our scenarios, we assumed a net bond return of 5% — more than double the current yield of the bond index.
Revisiting the Definition of Risk: Outliving Your Money
Adding bonds to an equity portfolio lowers the portfolio’s volatility as well as its overall return, and these differentials correspond precisely to the differentials between bonds and equities in volatility and returns. Remember, volatility is not risk; and a temporary decline only becomes a permanent loss if you sell. True risk is the possibility of outliving your money. Steadily increasing your allocation to bonds based on your age can be a recipe for financial ruin over the long term.
Timothy Keating is the president of Keating Wealth Management, a financial planning and investment advisory firm. He has 34 years of Wall Street experience, previously serving as the CEO and founder of a publicly traded closed-end fund focused on pre-IPO investing.