If You Think Bonds Are Signaling Lower Inflation, You're Looking At The Wrong Bonds

Recently I was on a panel discussion talking about the economic outlook for America. One of the first questions was about inflation. I said that I thought inflation risk were rising, but the moderator countered by talking about how low treasury bond rates are. But do low bond rates imply that the market thinks there will be low inflation?

There is at least some reason at the surface level why low treasury yields might imply low inflation. Yields are the interest rate divided by the price you pay for the bond. It's the annual return on investment if you hold the bond to maturity.

Here’s the definition of Bond Yield according to Investopedia:

“Bond yield is the return an investor realizes on a bond. The bond yield can be defined in different ways. Setting the bond yield equal to its coupon rate is the simplest definition. The current yield is a function of the bond's price and its coupon or interest payment, which will be more accurate than the coupon yield if the price of the bond is different than its face value.”

Since bonds are generally what's called "fixed income," i.e., the amount of the annual interest check does not change, then that return cannot vary with inflation rates, therefore the yield needs to be high enough to convince investors that they are being adequately compensated for the risk of future inflation.

Think of it this way: If you invest 100 dollars in a bond and that bond pays you 5 dollars a year, thus giving you a 5% yield, do you actually get a real return? Not if the inflation rate is 5% or more. In that case, your "return" is being eaten up with inflation. Your interest checks are not keeping up with the rising cost of living. So, if you think that inflation is going to be rising, you will not be willing to invest in bonds at an interest rate which doesn't compensate you for that risk.

So, treasury bond yields do have something to do with inflation.

But treasury bond yields also have a lot to do with other things, as well. For example, those yields can be affected by the yields of other kinds of bonds. If the other bonds start to offer more attractive yields, then treasury bonds might have to go up to in order to compete with the alternative. And it's not just other bonds; even stocks compete with bonds for your investment dollar. Plus, on top of all of that, our central bank has, ever since the period shortly after The Great Recession, been buying treasury bonds, which means our own Fed has become a bond investor and it is competing with us to purchase these bonds, which drives the yield down. Even the risk of default (not a big factor probably with treasury bonds, although one of the rating agencies did downgrade treasury bonds, indicating a slightly higher risk of default) is part of the mix, especially as debt levels rise to historically high levels.

Because of all these other factors which affect the yields on bonds, bond yields are not just reflections on what investors think inflation will be. Inflation is one of many factors which is having an effect.

So, is there a way of accounting for all these other factors in such a way as to isolate the risk of inflation? There is, and it's called the 'TIPS spread' or 'the breakeven rate'.

As defined by Investopedia:

“TIPS spread is the difference in the yields between U.S. treasury bonds and Treasury Inflation-Protected Securities (OTCPK:TIPS) and is a useful measure of the market’s expectation of future Consumer Price Index (CPI) inflation.”

Let's look at two sets of Treasury Yields. Both are five-year duration: the blue one is just a regular treasury, but the red one is from TIPS.

(Source: St. Louis Federal Reserve, as of 7/31/2021 or most recently available)

The regular yield has to compensate investors for inflation risk. But TIPS work differently. They compensate the investor for inflation by repaying more than the original principal at the end of the five-year period. One compensates for inflation through the yield, the other does it differently. This makes the difference between the two types of bond yields a reflection of the expected inflation.

Here's that spread below over the same time period:

(Source: St. Louis Federal Reserve, as of 7/31/2021 or most recently available)

We can see that it’s been rising since the COVID crisis and has risen year-to-date.

Zooming out on the same measurement for the entire time period available in the data, we can see that the recent spike puts this inflation expectation near historic highs:

(Source: St. Louis Federal Reserve, as of 7/31/2021 or most recently available)

Of course, this is a measurement of expectations (actually, it's not a perfect measurement even of that, there are other things distorting it, which we'll discuss in a future blog), but expectations can be wrong. Does this indicator accurately predict inflation? If we compare the five-year inflation expectation (on the right-to-left axis) to the actual inflation over the ensuing five years (on the top-to-bottom axis), it looks like this:

(Source: St. Louis Federal Reserve, as of 6/30/2021 or most recently available)

Every dot above the line is an instance in which actual inflation was higher than a model based on the Inflation Expectation would predict. Please note that the higher-than-expected inflation, the group of dots in the upper right of the chart, tends to occur when both Inflation Expectation is higher than usual (meaning towards the right of the chart), and ensuing inflation is higher than normal (meaning towards the top of the chart). So, historically when TIPS spreads have been high and subsequent inflation turned out to be high, then at such times, TIPS have tended to underestimate inflation. We know that we're currently over on the right side of the chart, with high expectations. Time will tell whether the pattern holds and subsequent inflation turns out higher than expected.