Assessing the term structure of expected inflation using treasury inflation—protected securities: near real-time measures for those who need quick estimates

Business Economics, Jan, 2003 by Albert E. DePrince, Jr.

The inversion of the term structure of short-term inflation expectations intensified as time progressed during 2000. In other words, the inflation-protected-based expectations model indicated that, in early 2000, markets expected inflation to be high in the near term, but receded as the horizon lengthened to ten years. This methodology put short-term expected inflation at around 2.25 percent, which was high compared with the Federal Reserve's price stability objective.

Of all the inflation estimates generated from the TIPS, the least reliable may well be the one-year. This was generated from a five-year inflation-protected note that had aged four years. Its wide swing could well reflect a demand for very short-term conventional securities (in the aftermath of the September 11th attacks) that did not spillover onto the inflation-protected note.

The Term Structure of Expected Inflation Using Survey Results

This section briefly examines the term structure of expected inflation obtained from the Survey of Professional Forecasters. This survey is conducted quarterly by the Federal Reserve Bank of Philadelphia and collects inflation forecasts over one- and ten-year forecast horizons. Results are reported in Figure 1 for the 1991Q4-2002Q4 period.

As with the TIPS-based expectations data, the forecasts show that over time, short- and long-term forecasts generally move together. However, the short-term (one-year) forecast tends to cycle around the long-term (ten-year) forecast, reflecting the fact that short-term forecasts are more responsive than long-term forecasts to incoming information.

The cycling suggests that the term structure of expected inflation varies from positive to negative, depending on the evolution of near-term information. Interestingly, the same theme is found in the TIPS-based data. Short-term (two- and three-year) TIPS-based expected inflation cycled above intermediate-term expectations for most of 2000 and the early part of 2001. Though the inversion in the survey data started at about the same time, it was shorter in length.

While the cycles are qualitatively the same, there is a large quantitative difference in the shifts in the term structure of inflation expectations between survey results and the TIPS methodology. A basis for this difference is examined in the next section, which suggests that the TIPS-based inflation forecasts respond more quickly to new information than do the survey results. This characteristic, along with their real-time availability, could make the TIPS-based forecasts important inputs into decisions requiring views of inflation.

The Economic Characteristics of Expected Inflation: SPF versus TIPS-Based Estimates

The ten-year SPF and TIPS-based inflation forecasts are graphically compared in Figure 5. Over the common sample period, both declined from early 1997 through mid-1998 and leveled off since then. The TIPS-based forecast fell further than the SPE estimate in 1998 and has since remained beneath the SPF forecast. Also, it has been more volatile than the SPF forecast throughout the common sample period. Shen and Coming (2001) argued this reflected shifts in the liquidity premium in the inflation-protected security and various premiums in the conventional security.