Home > XE Currency Blog > US Bond Market Week in Review: Fed President Williams Is a Bit Too Mechanical


XE Currency Blog

Topics4778 Posts4823
By HaleStewart October 22, 2017 7:26 am
  • XE Contributor
HaleStewart's picture
HaleStewart Posts: 712
US Bond Market Week in Review: Fed President Williams Is a Bit Too Mechanical

      There are several ways to classify Federal Reserve governors.  The most common is obviously hawks and doves, with the former arguing for rate increases to pre-emptively slow inflationary pressures while the latter are willing to let the economy run a bit hotter, allowing inflation to increase.  But other classifications are possible.  Over the last few years, I’ve noticed an emerging line between the “modelers” – those who base their interest rate policy determinations on the Fed’s macroeconomic model -- and those who rely on data.  Yellen and Fisher are examples of the former while Bullard and Brainard are in the latter camp.  Fed President John Williams is also in the data camp.  While I disagree with his eventual policy recommendations, they are rooted in data.  He recently returned to one of his favorite themes – R-star (r*), or the interest rate that would neither constrict the economy nor cause inflation. 

            While r-star cannot be “seen,” it can be extracted from common economic data.  In previous research, Williams has noted that r-star is now very low:

Empirical evidence of a sizable decline in the natural rate of interest in the United States has accumulated over recent years. The shaded region of Figure 1 shows the range of estimates of the natural rate for 1986 to 2016 based on studies by Laubach and Williams (2003), Kiley (2015), Lubik and Matthes (2015), Johanssen and Mertens (2016), and Holston, Laubach, and Williams (2016). Although individual estimates differ at any point in time, the range of estimates has clearly moved down over the past decade, and all estimates are below 1% at the end of the sample, the third quarter of 2016. The solid line in the figure plots the average of these five estimates over time. This average fluctuated between 2 and 2½% in the 1990s through the mid-2000s, and then plummeted to about ½% around 2009, where it has remained through 2016.

Here is the accompanying chart:

Other, observable data supports this analysis.  The following long-term chart of the 10 and 30-year CMT shows clear, long-term declines:

The shorter end of the curve exhibits the same behavior:

            Williams notes in a later speech that, assuming r-star is accurate, that the Fed should raise rates to about 2.5%:

Economists tend to think of r-star in terms of the real, or inflation-adjusted, interest rate. My own view is that r-star today is around 0.5 percent. Assuming inflation is running at our goal of 2 percent in the future, the typical, or normal short-term interest rate would be 2.5 percent.1 That’s a full 2 percentage points below what a normal interest rate looked like 20 years ago.

This is a standard formulation of the appropriate interest rate. 

            While I respect Williams use of data, I think his methodology overlooks a few key points.  First, there is no discussion about why inflation is currently so low, completely overlooking a potentially massive economic problem.  If weak inflation is structural, then raising rates could further depress prices, potentially leading to deflation.  Secondly, although he admits the US and the other developed countries are currently stuck in low growth, he doesn’t acknowledge that raising rates to 2.5% could further weaker growth.  Third, there’s a very mechanical quality to William’s analysis that is reminiscent of the arguments supporting the Taylor rule; Williams is treating Fed policy as a simplistic function rather than a nuanced discipline.

Hale Stewart is a tax attorney in Houston, Texas and a financial adviser with Thompson Creek Wealth Advisors.  He is also the author of the Lifetime Income Security Solution.









Paste link in email or IM