I am confused by the "10-Year Breakeven Inflation Rate" derived by the Fed. The Federal Reserve Bank of St. Louis maintains a chart at https://fred.stlouisfed.org/series/T10YIE which is designed to represent "expected inflation derived from 10-Year Treasury Constant Maturity Securities and 10-Year Treasury Inflation-Indexed Constant Maturity Securities". If I read this right, it currently (April 1, 2022) shows an expected inflation of of 2.79% over ten years.

Many expect inflation to be much higher, at least over the short term. Also, it shows very little inflation was expected during the GFC in 2008 and at the start of the lockdown in 2020. Does this represent something wrong with the metric? Or overall failure of the markets to predict inflation?

  • 2
    $\begingroup$ The value comes from traded securities. Hence, the value is what many expect (and directly trade). quant.stackexchange.com/a/65713/54838 has some links and a general explanation for this metric (and its 5y5y fwd). $\endgroup$
    – AKdemy
    Apr 2, 2022 at 16:22
  • $\begingroup$ I suggest you make a spreadsheet to compare the expectations against the running averages of the subsequent outcome. $\endgroup$
    – H2ONaCl
    Apr 2, 2022 at 21:49

1 Answer 1


Is the "10-Year Breakeven Inflation Rate" useful?

Yes it is a standard proxy for inflation expectations used across research and also by central banks. It is actually one of the most commonly used measure even though other measures such as survey of experts exist (see this Brookings explainer Powell & Wessel).

Many expect inflation to be much higher, at least over the short term. Also, it shows very little inflation was expected during the GFC in 2008 and at the start of the lockdown in 2020. Does this represent something wrong with the metric?

First, I am not sure who are the "many" you refer to there are still discussions among economists whether current high inflation is transitory or not it is possible that next year or already year after inflation will be low (see this FT article). You can have 10% inflation for 2 years and then 1% inflation for 8 years and it still averages at 2.8%. Second, as I explain little bit more below, inflation expectation of 2.79% is not the same as saying the realized inflation will be 2.79%, inflation expectation just says that is the most probable inflation rate, but of course 2.8%, 2.9%, 3% or 3,5% might have still very high probability just not as high as 2.79% as perceived by market participants given information they have available today.

There is no indication that there is something wrong with the metric itself just because the inflation expectations are lower than what you believe they should be.

Or overall failure of the markets to predict inflation?

Markets cannot predict inflation in a sense that they can know beforehand with certainty what the inflation will be. They could at best only predict it in the statistical way. Expected inflation rate is the anticipated inflation market participant believe will occur conditional on the information they have.

Let me give you example. Suppose we have 10 fair coins so chance of getting head or tail is exactly 50%. Given this trivially we can calculate that the expectation for the number of heads from an experiment where we throw all ten coins is to get 5 heads. Getting 5 heads would be verifiably the correct expectation.

Yet despite that 5 heads is the correct expectation for number of heads its not that unprobeable we will observe 3,4, 6, or 7. So despite that the 5 heads is correct expectation there are lot of different outcomes that can happen and although the 5 will be correct on average over long term where I repeat experiment large number of times, in the short term you might see even 10 heads from some experiment. That does not make the expectation of 5 heads wrong (again in such simple experiment we can mathematically prove that expected value is 5 since for binomial discrete variable expectation is given by $E = np$, where $n$ is number of trials (in our case coins) and $p$ probability of success (in our case heads) and in our case $n=10$ and $p=0.5$, implying that the correct expectation is $E=5$).

Hence you should not confuse expected inflation with some sort of precognition of what correct inflation will be. Indeed inflation expectations are not good predictors of one step ahead actual inflation, but that does not mean there is something wrong with inflation expectations. This because expected value of something will often be different than the actual value of something (as the coin tossing experiment indicates), and in addition economies are continuously getting hit by inherently unpredictable random shocks (e.g. Covid19).

  • $\begingroup$ Obviously markets cannot predict the future perfectly, but we have markets designed to predict the future to guide capital allocation (e.g., commodity futures). If I were to take a long position in inflation-adjusted bonds like TIPS and go short in nominal bonds, wouldn't I expect to match the inflation rate predicted by the Fed? $\endgroup$ Apr 2, 2022 at 16:42
  • $\begingroup$ @GeneMcCulley I am not sure if I follow your example. Look expected inflation is like expected IQ of a random person from a city. The expected IQ will be 100 but chances that if you pick single random person her or his IQ is 100 are near zero. It just mean the 100 is the most likely out of all other numbers (ie even though the chance of picking someone with IQ 100 is almost zero chance of picking someone with IQ 101 is even closer to zero $\endgroup$
    – 1muflon1
    Apr 2, 2022 at 18:37
  • $\begingroup$ But I expect the settlement of a futures market to be close to the actual price on the settlement date and I expect a sophisticated futures market to be a good predictor. Likewise, I expect the ratio of TIPS bonds to other long term bonds to reflect the expected inflation. $\endgroup$ Apr 2, 2022 at 19:06
  • $\begingroup$ @GeneMcCulley but the prices on such market will be set in a such way that there is no ex ante arbitrage opportunity. In that sense the expected prices are also correct ex ante. However, the actual realized (ex post) price that is observed at some future date won’t likely be equal to the expected price. That would require precognition in sufficient number of market participants. That’s impossible $\endgroup$
    – 1muflon1
    Apr 2, 2022 at 21:26
  • 3
    $\begingroup$ CLH22 (WTI march 22 future) traded well below 60, sometimes below 40 for over two years. Yet, the oil price ended up at 92. Following your logic, commodity futures are grossly mispriced. Just because on lockdown day the value is low does not mean much. No one was able to predict the outcome of Corona. In the short term, a lockdown is always deflationary and what was clear is that it was very hard to get inflation going for years. Also, it is the 10 year rate on average, not what happens now for the last few months. For example, U.S. inflation rate for 2020 was the lowest in several years. $\endgroup$
    – Alex
    Apr 2, 2022 at 23:31

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.