On My Mind: (r)-stargazing

What is r-star, or r*, and how much does it matter for the outlook for bond yields and market interest rates?

It matters because in setting monetary policy, the US Federal Reserve’s (Fed’s) estimate of r* also guides it. The Fed’s estimate of r* is the real policy rate (as opposed to the nominal fed funds rate) that would be neither contractionary nor expansionary when the economy is at full employment and inflation at its 2% target.¹ Call it the “real neutral” rate. The consensus view is that r* had experienced a secular decline and might have moved back up somewhat in recent years—but is that view correct?

Fed Governor Christopher Waller and former New York Fed President William Dudley recently gave some contrasting perspectives.² Of the two, Dudley’s comes much closer to what I have been arguing for the past 2-3 years—namely that the neutral rate is a lot higher than the Fed and the markets think. In this note, however, I want to go further and offer a different perspective on how to look at the past decades and what they mean for the future outlook on rates.  

R*, is unobservable, but, as Waller notes, we have two distinct ways of gauging it. The first is to look at the inflation-adjusted 10-year US Treasury (UST) yield:

At first glance, this seems to indicate a downward trend in r* since the early 1980s, followed by a rise in the last four years. However, a different gauge of the neutral rate is the real return on capital, which shows no declining trend whatsoever.

Waller focuses on the real UST yield as the most meaningful proxy of r*, and reasons that something must have happened to drive a divergence between the real return on capital and the real yield on USTs. Waller argues that five factors boosted demand for USTs: globalization of capital markets, a buildup of official reserves, demand from sovereign wealth funds, population aging, financial regulation, and the Fed’s purchases of USTs. Waller reasons that all these five factors will continue to keep r* low, with loose fiscal policy increasing UST supply the only contrast.

Dudley dismisses out of hand the idea that this is a useful way to look at r*, since by definition r* must be closely linked to economic fundamentals, such as productivity. As I have been arguing for quite some time now, Dudley also observes that since growth has remained robust even as the Fed hiked the fed funds rate, r* must have risen, so that the current stance of monetary policy is only mildly restrictive.

But here is where I want to offer a different perspective. Both Dudley and Waller buy into the idea that r* was in secular decline until recently; and yet neither one of them can point to any change in underlying growth. In fact, changes in growth fundamentals, such as productivity, should have impacted first and foremost the real rate of return on capital, which as we saw did not budge. I will instead suggest, somewhat controversially, that r* might in fact never have been in secular decline. Let’s take a second look at the real UST yield chart:

1. Notice, first of all, that the downward trend is artificially accentuated by the huge jump at the very beginning of the sample period, from 2% to over 9%. Longer-term estimates of r* indicate that from the 1950s to the global financial crisis (GFC) it averaged about 2%, so this spike is an outlier (see my previous article "On My Mind: The structural shift that wasn't." The real UST yield then stepped down during the 1980s as policymakers tamed high inflation; throughout the 1990s it was stable in the 2%-4% range.
2. There is another discrete step down between 2000 and 2003, from about 3.5% to about 1%. This coincided with a major easing of monetary policy in response to the popping of the dot-com bubble and ensuing recession. The real UST yield then trended back up to about 2% until …
3. … the GFC triggered another massive expansion of monetary policy, and the real UST yield fell again from over 2% in 2007 to below zero by 2013. R* then rose back until…
4. … we got the COVID-19 recession and another round of even more dramatic monetary policy expansion, which sent the real UST yield back down.
5. Finally, once monetary easing first stopped and then slowly reversed, the UST yield rate went back up quite sharply, until it is almost back to where it was at the start of the sample.

I am not arguing that r* was always stable; it fluctuated with economic fundamentals. But underlying economic fundamentals did not change enough to push the real neutral rate down from 2% to near-zero in a durable, structural manner. What we see in the UST real yield chart, in my view, is mostly the impact of repeated rounds of massive monetary policy easing, boosting demand for USTs and driving their real yield down. In other words, massive Fed purchases have historically played a defining role in the movements of real UST yields. I think the neutral rate, the famous r*, is now probably at 2%-2.5%, in line with its long-term average from the 1950s to the eve of the GFC—especially given that we are seeing a rising investment trend and some signs of a pickup in productivity.

This has two implications. First, that current Fed policy is not overly restrictive and disinflation progress will remain gradual. Second, that the next Fed rate-cutting cycle will likely be short and shallow.

Now let’s think about the outlook for nominal UST yields. If I am right, and r* is about 2%-2.5%, then once inflation is at the Fed’s target (2%), the fed funds rate should be at 4%-4.5%. Then we have to add term and risk premium,³ which averaged about 1.3 percentage points in the two decades preceding the GFC. And now we also have a massive persistent fiscal deficit that the congressional budget office projects will keep boosting UST supply for the next decade. After the next Fed easing cycle has played out, therefore, in the medium and long term I expect UST yields will drift back up, with levels greater than 5% looking more than plausible.