A Mystery Solved: Why Real Yields are falling despite higher growth
All the major economies are doing somewhat better, but US “real yields” — as seen in Treasury inflation-protected securities, or TIPS — have moved sharply lower after a jump following November’s presidential election.
In the chart below, we see the two components of the five-year US government note yield: the inflation-indexed TIPS yield, and the difference between the TIPS yield and the overall coupon yield. This difference is called “breakeven inflation,” because if the US Consumer Price Index is equal to this level over the next five years, investors in TIPS and investors in ordinary coupon securities will get the same return, or breakeven.
Economists often think of real yields as the “real interest rate,” or baseline rate of return, in a macroeconomic model. From this standpoint the low level of TIPS yields is a mystery: when economic growth is rising, the real interest rate should rise. The expected short-term interest rate has been rising as the Fed sets about normalizing rates, and the rising short-term rates affect real yields. The fall in TIPS yields in the face of Fed tightening and stronger growth presents a double challenge to the conventional wisdom.
The conventional way of looking at real yields ignores the way markets treat risk. Government debt (and particularly the government debt of the United States) is not just a gauge of economic activity, but a kind of insurance. If the world comes crashing down, you want to own safe assets. Investors hold Treasuries in their portfolios not just for the income, but as an insurance against disaster. And TIPS offer a double form of insurance: If economic crisis takes the form of a big rise in the inflation rate, TIPS investors will be paid a correspondingly higher amount of principal when their bond matures. That explains why TIPS yields sometimes are negative: investors will accept a negative rate of return at the present expected inflation rate in return for a hedge against an unexpected rise in the inflation rate.
The yield on TIPS has tracked the price of gold with a remarkable degree of precision during the past 10 years, as shown in the chart below. Gold tracks the 5-year TIPS yield with 85% accuracy. That’s because both gold and TIPS function as a hedge against unexpected inflation.
During the past year, for example, we observe that the relationship between gold and the 5-year TIPS yield has remained consistent, while the relationship between the expected short-term rate (as reflected in the price of federal funds futures for delivery a year ahead) has jumped around. There are lots of local relationships between federal funds futures and the TIPS yield, but the overall relationship is highly unstable.
The breakeven inflation component of the TIPS yield, by contrast, tracks current input prices, for example, the price of oil.
Two different sorts of expectations are at work. A 40-year-old family man doesn’t expect to die anytime soon, but he probably will pay up for life insurance in case the unexpected happens. If he has high blood pressure, the insurance company might charge more for insurance – even though the likelihood of his death is quite small. If he is fit and blood pressure is normal, the insurance company will charge less.
Something like this is at work in the interest rate market. During the first 10 months of 2016, the world was at risk of deflation. The oil price was low, and the market’s best estimate of inflation reflected in Treasury breakeven yields was quite low. That is the region shown in the left hand part of the chart below. Note that as breakeven inflation fell, the yield on TIPS went up. That is, when expected inflation fell, the cost of insurance against unexpected inflation also fell. The risk of unexpected inflation is lower in a world that is worried about deflation. (Rising TIPS yields indicate a lower cost of insurance for unexpected inflation. When investors really worry about unexpected inflation, they will accept a lower, indeed a negative yield for inflation-protected securities).
Starting in October, though, there was a jump in the expected inflation rate. This corresponded to a rise in the price of oil. The relationship between expected inflation and TIPS yields “jumped” to a different level as TIPS yields fell (which is to say that the cost of insurance against unexpected inflation went up).
It turns out that the price of expected inflation (or breakeven yields) and the cost of insurance against unexpected inflation (for example the gold price) together give us a very high degree of explanation of the yield on inflation-protected securities.
The chart below shows the same regression fit without including breakeven inflation as an explanatory variable.
By taking into account the cost of insurance against unexpected outcomes in an uncertain world, we show that there is no mystery at all to the recent fall in real yields. The problem lies in thinking of the TIPS yield as the real yield in an abstract macroeconomic model based on fixed expectations.
Originally published on the Asia Times.