I’d like to share and discuss a FRED graph. To me it shows an alarming similarity between inflation expected over the next five years and inflation over the past year. The title is not uh very useful.
The red line is the nominal treasuries minus TIPS breakeven for constant maturity of 5 years — the difference between the nominal interest rate and the real rate paid by Treasury Inflation Protected Securities. The 5 year constant maturity TIPS rate is calculated by interpolation, because there are only a few TIPS with a few different maturities on the market. The breakeven should be related to expected inflation, but should be lower than expected inflation as the price of regular nominal Treasuries includes a liquidity premium (which was huge during the panic of 2008). The yellow line is the rate of growth over the precedeing year of the geometric mean of the personal consumption deflator and the personal consumption deflator excluding food and energy (AKA core inflation). The blue line is the rate of growth of the price of crude oil in West Texas times 0.03 to fit it on the same graph.
To me the most striking feature of the graph is that back in the good old days before the recession, the looking forward 5 years breakeven is almost identical to the past years geometric mean of inflation and core inflation. This shouldn’t have happened. It corresponds to extremely mechanical adaptive expectations. The correlation is just too strong to be a coincidence. Since the trough the comovement remains impressive but is no longer shocking.
The oil price impressively times the increase in breakeven inflation at the time QE II started. The magnitude is a bit low (earlier the movement of the breakeven is less than 0.03 times the movement of the log price of oil and in late 2010 and early 2011 they are almost identical).
My takeaway from the graph is that foreward looking expectated inflation is extremely similar to past inflation leaving almost no room for the rational expectations based critique of the use of adaptive expectations or any effects from quantitative easing (since QE 1) or forward guidance.
There are two problems. First, my takeaway from the empirical exercise is exactly the same as my bring to the exercise which makes even me suspicious. Also the exercise is neither here nor there — it isn’t a simple obvious FRED graph as the variables are somewhat processed and it certainly isn’t econometrics with confidence intervals and hypotheses tested. Still I think the graph should impress and shock.