The Survey of Professional Forecasters (SPF) February release and the January five-year breakeven inflation rate suggest accelerating inflation – up to about 2.1%-2.2%.

(Click on image to enlarge)

Figure 1: Year-on-year expected inflation from University of Michigan survey median response (blue), from Survey of Professional Forecasters median response (red +), and implied five-year average inflation from spread between five-year constant maturity Treasurys and TIPS (green). NBER defined recession dates and peak in gray. Source: U.Michigan via FRED, SPF via Cleveland Fed, and Treasury via FRED, NBER, and author’s calculations.

The five-year breakeven has risen further, to 2.37% as of yesterday (2/16). The latest observation for the Michigan survey of households (as opposed to economists) is for December and comes in at 2.5%. Interestingly, households have consistently higher expectations of year-ahead CPI inflation, relative to forecasters. See Coibion and Gorodnichenko (2015) for further discussion of the differences, and which one better fits the expectations augmented Phillips Curve.

Where will year-on-year inflation be in a year from now? Forecast errors for the Michigan and SPF (both based on medians) are shown below.

(Click on image to enlarge)

Figure 2: Forecast error for year on year expected inflation from University of Michigan survey median response (blue), from Survey of Professional Forecasters median response (red +).NBER defined recession dates and peak in gray. Source: U.Michigan via FRED, SPF via Cleveland Fed, NBER, and author’s calculations.

Over the sample period displayed in Figure 2, The average forecast error is -0.96 ppts for the Michigan survey (i.e., households overpredict inflation), while that for the SPF is -0.06 ppts. (The five-year breakeven underpredicts on average 5-year inflation by 0.11 ppts for 2003-2015, although there is a lot of upward skew, so median error is overprediction.)