Bond Market Useless in Predicting Inflation
A common argument from people who belive inflation is not a problem in America is to point to the low spread between regular bonds and TIPS (Treasury Inflation Protected Securities). I've explained before why bond investors seem to systematically underpredict future inflation (see here, here and here).
Here is more concrete example of how the TIPS-regular bond spread is not a reliable inflation indicator and how it systematicallym underestimate future inflation. The Federal Reserve statistics on historical TIPS only go back to January 2003, even though TIPS have existed longer. Anyway, now 5 years have passed since then and with today's inflation statistics we can compare the TIPS-regular bond spread for 5 year securities with the actual inflation since then.
In January 3 2003, the yield on the 5 year nominal security was 2.91%. Meanwhile, the 5 year TIPS yield in January 3 2003 was 1.75%. This means that the spread was a mere 1.16% of the time, implying an average expexted inflation of 1.16%.
However, during these 5 years, the consumer price index rose from 180.9 to 210.0, an increase of 16.1%, which translates into an average annualized rate of inflation of 3,03%, or 1.87%:points higher than the rate predicted by the bond market. Which means that bond investors forecasts of inflation was 1.87%:points lower than the actual number.
This again illustrates how irrational anyone who buys U.S. government bonds is, particularly the buyers of the nominal ones. Indeed, note how the yield for the suckers who bought 5-year securities of 2.91% was lower than the rate of inflation of 3.03%. So, even before taking taxes (which are based on nominal, not real, returns)into account, they experienced a negative real return.
So, considering how they've been consistently wrong and consistently losing money, we certainly shouldn't consider their irrational investment decisions as evidence of low future inflation.
Here is more concrete example of how the TIPS-regular bond spread is not a reliable inflation indicator and how it systematicallym underestimate future inflation. The Federal Reserve statistics on historical TIPS only go back to January 2003, even though TIPS have existed longer. Anyway, now 5 years have passed since then and with today's inflation statistics we can compare the TIPS-regular bond spread for 5 year securities with the actual inflation since then.
In January 3 2003, the yield on the 5 year nominal security was 2.91%. Meanwhile, the 5 year TIPS yield in January 3 2003 was 1.75%. This means that the spread was a mere 1.16% of the time, implying an average expexted inflation of 1.16%.
However, during these 5 years, the consumer price index rose from 180.9 to 210.0, an increase of 16.1%, which translates into an average annualized rate of inflation of 3,03%, or 1.87%:points higher than the rate predicted by the bond market. Which means that bond investors forecasts of inflation was 1.87%:points lower than the actual number.
This again illustrates how irrational anyone who buys U.S. government bonds is, particularly the buyers of the nominal ones. Indeed, note how the yield for the suckers who bought 5-year securities of 2.91% was lower than the rate of inflation of 3.03%. So, even before taking taxes (which are based on nominal, not real, returns)into account, they experienced a negative real return.
So, considering how they've been consistently wrong and consistently losing money, we certainly shouldn't consider their irrational investment decisions as evidence of low future inflation.
posted by stefankarlsson at 10:26 PM
1 Comments:
I call the people who believe that bond markets can predict inflation "market mystics". Still, ceteris paribus, higher bond rates, ceteris paribus, normally express fears for inflation.
Post a Comment
<< Home