The Economist recently published an article debunking the already well-debunked myth that U.S. Treasury bonds are risk-free (see Problem 22 in Chapter 10). While we expect that many people are not aware that Treasury bonds have inflation, or purchasing power risk, we hope that by now you do. It is well-known that there is no true risk-free asset, but Treasuries are used as a proxy for the risk-free rate. What is really disappointing is the time value of money skills displayed by the author. The article notes that the real return in Treasury bonds was a loss of about 2 percent per year, for a cumulative purchasing power loss of 91 percent for the period 1946-1981. And of course, this got us fact checking. Based on the returns reported by Ibbotson, an investor in long-term Treasury bonds would have earned a nominal return of 2.35 percent over this 35-year period, while inflation averaged 4.90 percent per year. This gives a real return of -2.43 percent per year, which results in a cumulative real loss of 57.76 percent. Pretty bad, but not even close to the 91 percent reported.
We would also like to point out dangers in any article that chooses a specific period from a longer dataset. If the investor has stayed in the market for the period 1982-1991, the average real return on Treasuries was 12 percent per year for this period. So, over the entire 45-year period, the real return for Treasury bond investors was .78 percent per year.
