I am not talking about the NAIRU. I am talking about the ten-year Treasury yield! The decline in the neutral fund’s rate, the decline in the steady-state term premium, and our expectation of a much slower rise toward the steady-state term premium have materially lowered our projection for the ten-year Treasury yield over our forecast horizon, which now extends through 2020. Not all that long ago, we were using 2½% as our estimate of the (long-run) real neutral funds rate and 100 basis points as our estimate of the term premium, implying a ten-year yield near 5½% at full employment and price stability. Estimates of the longer-run real neutral rate have since declined to 1%, or slightly below. And the current, or short-run, real neutral rate is estimated to be slightly negative, with uncertainty about how long it may take for the short-run rate to converge to its longer-run level. Our estimate of the term premium over the forecast horizon has also declined. Today, assuming a normalized balance sheet, it may be close to zero by the end of 2020. That leaves a plausible central tendency range for the ten-year yield in 2020 between 2% and 3%.
The Decline in the Neutral Funds Rate
Estimates of the (longer-run) real neutral funds rate have declined from 2½% when I was at the Federal Reserve Board, and for some time after that, to 1% today, based on the median estimate of FOMC participants in June. In addition, the estimate of the short-term real neutral rate today is lower than its longer-run level, zero or even negative, reflecting the degree to which monetary policy today should be more accommodative than otherwise to offset persistent headwinds originating with the 2007-2009 recession.
In Figure 1, we show participants’ median estimate of the longer-run neutral rate since 2012, when FOMC members began to report the fund’s rate “in the longer run,” and a measure of the real neutral interest rate in the short run. FOMC participants (including Chair Yellen) often turn to the Laubach-Williams r-star, based on a Fed staff research paper coauthored in 2001 by Thomas Laubach, now Director of Monetary Affairs at the Board, and John Williams, now President Williams of the San Francisco Fed.1 That measure, which is publicly available on the San Francisco Fed website, has fallen along with the long-run rate, and is lower than the long-run rate–now zero, or slightly negative–reflecting that monetary policy has to be more accommodative over the years following the 2007-2009 recession than in normal times because of persistent headwinds. In Figure 1, we show the estimate of the long-run neutral rate in real terms, to make it comparable to the Laubach-Williams estimate of the short-run rate. While many participants in June expected the short-run r
star to converge to its longer-run value by the end of 2019, there is uncertainty about whether the short-run r-star might actually be the new normal.
1 “Measuring the Natural Rate of Interest,” FEDS Working Paper 2001-56, Thomas Laubach and John C. Williams, Board of Governors of the Federal Reserve System, November 2001.
The Decline in the Term Premium
In Figure 2, we show the Kim-Wright measure (and proxy) of the ten-year Treasury yield term premium. We can see that, over the last couple of decades, the term premium has declined, reflecting that investors today are less concerned with protecting against inflation risk and more concerned about secular stagnation and deflation risk (see Headed to a Flat Yield Curve?).
There was, in addition, a material decline in the term premium associated with asset purchases. One Fed staff study estimates that the peak cumulative impact was -120 basis points at the end of 2013. Since then, the term premium has been rising during the ongoing normalization of the balance sheet (thus affecting the term premium), though it is still negative. (see Headed to a Flat Yield Curve?). We expect the rise toward the steady-state level will be very slow, and that the term premium could still be zero at the end of 2020.
How Low Could the Steady State Ten-year Yield Be?
To establish a plausible range for the steady-state ten-year yield today, we assume that the real neutral fund’s rate is between zero (the estimate cited by FOMC participants, even though the Laubach-Williams two-sided estimate for 2017:Q2 is -22 basis points) and 1% (participants’ median projection of the real longer-run level); or 2% and 3% in nominal terms. And we assume the term premium will be between zero in 2020. Table 1 shows the range for the ten-year Treasury yield at the end of 2020, reflecting the ranges and estimates we set out above for the short-run and long-run r-star, and a fair consensus like an estimate for the term premium at the end of 2020. Quite a range for the ten-year yield, and a very low one, from 2% to 3%!
A Plausible Range for Neutral Rate and Term Premium in 2020
|Ten-year Treasury Yield in 2020 (%)||Neutral Rate in 2020 (%)||Term premium in 2020 (bps)|
Revisions to Our Path of the Ten-year Treasury Yield
In the just-completed forecast round, we revised our estimate of the neutral fund’s rate down 25 basis points to 2¾% and our projection of the term premium in 2020, also down 25 basis points to zero. The decline in our estimate of the neutral fund’s rate reflects our expectation now that the short-run fund’s rate will not have fully moved to its longer-run level by 2020. The 25-basis-point downward revision to the projected term premium in 2020 assumes a much slower convergence of the term premium to its steady-state level, which may still be positive. That leaves our projected ten-year yield in 2020 at 2.8%, taking into account the compounding of the fund’s rate, down from 3.4% at the time of our last forecast.