Date
Safe Withdrawal Rates in an All-Bond Portfolio
For my analysis, I use a portfolio consisting of 50/50 municipal bonds and TIPS. The actual mix should be based on forecasted inflation rates.
As of March 19, 2009, AAA-rated 30-year general obligation (GO) municipal bonds yield 4.99%, equating to a taxable-equivalent yield of 6.93% (based on a 28% tax bracket).
No AAA-rated state issuer has defaulted in modern history, so I will ignore default risk. A more conservative approach, which I will leave to future researchers, would assume a non-zero default rate and a recovery rate in the event of default.
Skeptics may point out that many states, such as California, face fiscal challenges. I agree, however, with Tom Doe of Municipal Market Advisors, who argues that it is “virtually impossible” for a state to default, since state governments have many options to raise revenue to meet debt payment obligations, and the Federal government has a strong incentive to prevent defaults by states.
I assume a highly diversified portfolio of these bonds, which are typically callable in 10 years at 101 or 102.
For TIPS, I use the current yield to maturity (2.22%) on the 20-year issue on January 15, 2009, which carries a 2.5% coupon. I assume that these bonds can be purchased at par with a 2.22% coupon and that there is no risk of default.
I also assume that TIPS and municipal bonds are annually repriced based on their current spreads relative to the CPI index (i.e., a 100 basis point increase in the CPI index results in a 100 basis point increase in bond yields). For the municipal bonds, I price these bonds to maturity and to a call price of 101, and use the lower of the two prices. If a bond is called, I assume that new bonds are purchased at then-prevailing interest rates, adjusted based on changes in the CPI.
I assume that any excess returns (when income exceeds withdrawals) are reinvested in the portfolio and, conversely, that shortfalls are funded by selling assets. This is consistent with previous studies, although a more rigorous approach would take into account the reinvestment risk associated with these cash flows.
I have not included any advisory fees.
Aside from the possibility of the muni bonds being called, the cash flows from this portfolio are known with complete certainty. Only the rate of inflation must be forecast to determine the safe withdrawal rate.
I analyzed this portfolio under the following scenarios:
- Constant inflation rates of -2%, -1%, 0%, 2%, 3%, 4% and 5%
- Historical inflation rates from 1966-1996. Consistent with prior research, I chose 1966 as the starting point because it represents the start of the most severe period of inflation in US history. It was also the start of a secular bear market, so this scenario is the worst case based on empirical data
- Inflation equal to twice the previous scenario, in order to simulate a hyperinflationary environment
- The inflation rate necessary to allow for a 4% withdrawal rate
Display article as PDF for printing.
Would you like to send this article to a friend?
Remember, if you have a question or comment, send it to .
