Lawler: Is The “Natural” Rate of Interest Back to Pre-Financial Crisis Levels?

CR Note: This is related to my note earlier this week: The "New Normal" Mortgage Rate Range

From housing economist Tom Lawler:

The resiliency of the US economy in the face of the “aggressive” Federal Reserve rate hikes since last March, combined with an increase in inflation-adjusted Treasury (TIPS) yields to levels not seen in almost 15 years, have led a growing number of analysts to conclude that the so-called “natural” rate of interest (often referred to as “r-star”) has increased significantly. By “significantly” I don’t just mean back to the historically low levels just prior to the pandemic, but rather to levels closer to the period right before the financial crisis of 2008 that led to the worst recession since the Great Depression.

To be sure, many other analysts and policymakers disagree with this view. For example, in May of this year the Federal Reserve Bank of New York relaunched its widely followed measure of r-star (the model generating r-star “did not work” during the pandemic), and the results suggested that r-star at the end of last year was not much different from where it was just prior to the pandemic. Moreover, the estimate of r-star from the so-called “HLW” model for the first quarter of this year was actually below pre-pandemic levels, leading NY Fed President Williams to conclude that “there is no evidence that the era of very low natural rates of interest has ended.”

On the other hand, estimates of the natural rate of interest from Lubik and Matthes (LM) of the Federal Reserve Bank of Richmond using a different/more flexible modelling approach suggest that the natural rate of interest had increased from a pre-pandemic level of about 1 ¼% at the end of 2019 to a little over 2% in the latter part of 2022 and the first quarter of 2023.

In addition, “real” interest rates as measured by “Treasury Inflation Protected” (TIP) yields have increased recently to their highest levels in about 15 years, with 5-year TIPS yielding about 2.2% and 10-year TIPS yielding almost 2%.

Moreover, the economy this year has been much stronger than most forecasters and policymakers had expected, which would also be consistent with (though by no means proof of) a “natural” rate of interest that had increased.

Back in June Goldman Sachs released a research report questioning the reliability of the low r-star estimates generated by the HLW model, and suggested that r-star (which is stated in real terms) was probably closer to 1 to 1 1/4%, implying a nominal “neutral” Fed funds rate if we have a soft landing and inflation returns to 2% of 3 to 3 ¼%.  Since that report “market-based” measures suggest a higher value of r-star.

It’s worth noting that in a 2015 Brookings Working Paper Laubach and Williams (The LW in HLW) noted that the decline in TIPS yields and the LM model results were “broadly consistent” with their estimates of the decline in r-star through 2015.  That is not the case today.

Here is a graph showing the HLW and LM estimates of r-star from 1990 through the first quarter of 2023.

Note that both of these estimates show a sharp decline in r-star right around the same time as the beginning of the financial crisis that led to the worst recession since the Great Depression.  For folks who didn’t live through this period, it was a period of great fear, and there were serious concerns that the financial system might collapse.  There were unprecedented actions taken by the government, but fear remained at elevated levels for years.  These fears were not just limited to the US:  in Europe multiple countries either faced banking crises, sovereign debt crises, or both and there were unprecedented actions taken by governments/central banks but fear remained for many years as well.  And it was during those “fear years” that estimated levels of r-star remained at extremely low levels.  Discrete changes like those observed were not related to longer-term trends such as demographics.

Note that in the last several years of last decade r-star estimates had begun to move gradually higher from the incredibly low levels (more so in the LM model than the HLW model) as fear began to recede, but then the pandemic hit, bringing more fear and uncertainty, though fortunately for a much shorter period than during the Great Recession/financial crisis/sovereign debt crisis.

By the latter part of 2022, however, these fears had larger abated, and the fears associated with the earlier financial crisis were long forgotten (save for a very brief and not very intense regional banks’ “scare” this March).

In sum, despite what NY Fed Governor Williams said, there is in fact evidence that the natural rate of interest, or r-star, has in fact risen over the last year, and by some measures it would appear as if r-star may be at levels not that far from the those seen just prior to the financial crisis of 2008.  If that were the case, then a “soft-landing scenario where inflation fell back to 2% would be one where the Federal Fund Rate fell back to the 3 ½-4% level, rather than the 2 ½-2 ¾ % level implied by the HLW model or the FOMC’s “dot plots.”  In addition, if r-star has risen significantly, then recent Federal Reserve policy has not been as “restrictive” as many have suggested.  If that were the case, then one would have expected the economy to have performed better than the “consensus” forecast, which in fact has been the case.

Of course, that does not mean the funds rate won’t go below 3 ½-4% even if r-star has increased.  If we were to have a recession, then the FOMC would probably decide that an “accommodative” policy would be appropriate, meaning a funds rate below the “natural” nominal rate.

But absent a recession and/or another sustained financial market “scare,” it seems that interest rates will be significantly higher than most people had been predicting for a sustained period of time.