Followers of the market for U.S. government bonds have started tossing around the term “bond bubble” with increasing frequency. The concern is that interest rates on long-term Treasuries have gotten so low that investors face high risk (if interest rates rise, the value of securities will fall) at low return.

My favorite academic bond analyst, J. Huston McCulloch, seems to share this view.
In the May 30 edition of his web page, he writes,

10-year TIPS [inflation-indexed Treasury securities] therefore have a higher expected return, in either real or nominal terms, than nominal notes of similar maturity, for every one of the 34 forecasters polled. At the same time, the indexed notes are essentially risk-free to their respective maturities, while the inflation risk on 10-year nominals is considerable. This is what is known in the economics literature as second order stochastic dominance. This means that no informed rational risk-averse investor with these inflationary expectations should be investing in the nominal notes when the indexed notes are available at current rates.

The way I interpret McCulloch’s analysis is that the long-term expected inflation rate that is implicit in Treasury bonds is implausibly low.

For Discussion. In theory, if McCulloch is correct, it should be profitable to buy TIPS and short nominal Treasuries. What are the risks and costs of doing so?