WHAT DO THE YIELD CURVE AND SOME OTHER FACTORS TELL US ABOUT THE POSSIBILITY OF A U.S. RECESSION COMING SOON?

 

Will the U.S. have a recession soon?  At this point, while it is possible, I think that it’s difficult to conclude with a high degree of confidence that the U.S. will soon have a recession. Based on information to follow, we may question whether recent U.S. yield curve behavior is signaling a recession in the near future.

It is well known that changes in the slope of the U.S. Treasury yield curve, which is the difference between a longer-term interest rate and a shorter-term interest rate (perhaps for the same borrower), have foretold U.S. recessions in the past.  Normally, longer-term interest rates are higher than shorter-term rates, perhaps because lenders require a premium for lending their funds for a longer period.  All else the same, the longer the term to maturity, the higher is the interest rate for the loan.  However, when shorter-term interest rates on U.S. Treasury debt have exceeded longer-term interest rates, such inversions have often been associated with a U.S. recession on the horizon.  The thinking is that an inverted yield curve implies that people expect lower interest rates in the future (due to expectations of lower real GDP growth, lower inflation, and/or higher money supply growth), to the point that these expectations dominate the liquidity premium and result in a yield curve that is at least partially downward sloping where a longer term to maturity is associated with a lower interest rate for the downward-sloping part of the yield curve.

Many have been trumpeting the inversion of the yield curve using the two-year interest rate and the ten-year interest rate of U.S. Treasury debt as foreshadowing a recession coming in the U.S.  If I remember correctly, then some even pointed to the recent inversion of the five-year and ten-year U.S. Treasury yield curve as foretelling a recession.

However, I think that the certainty of a U.S. recession in the very near future is ambiguous at best for at least two reasons.  First, the aforementioned two-to-ten-year and five-to-ten-year yield curves were no longer inverted based on data that I saw recently on CNBC.  I am not certain that very short-term inversions of the yield curve have accurately forecasted recessions in the past.

Second, Bauer and Mertens (2018), perhaps using monthly interest rate data, note the high degree of forecasting accuracy of recessions in the U.S. when the ONE-to-ten-year U.S. Treasury yield curve inverts.  My memory seems to tell me that their work found that (a) the one-to-ten-year yield curve for U.S. Treasury debt inverted less than twenty-five months before ALL U.S. recessions in recent decades up to the time of their study, and (b) the one-to-ten-year yield curve produced a ‘false positive’ result (where the yield curve inverted but no recession ensued within two years in the U.S.) only once during their sample period.  The ONE-TO-TEN-YEAR U.S. Treasury yield curve has NOT COME EVEN CLOSE TO INVERTING this year and has probably not come close to inverting since soon after the periods in which it was inverted in 2019.  Readers may want to note that a U.S. recession started in early 2020, which obviously is less than two years after the 2019 inversion of the U.S. Treasury one-to-ten-year yield curve.

Given that (1) recessions have periods of decreasing real gross domestic product (real GDP) and (2) real GDP is comprised of real consumption (by households), real investment in physical capital (largely if not completely by firms), real government purchases (by various domestic government sectors), and real net exports (or exports minus imports, both measured in ‘real’ or constant-buying-power currency); a U.S. recession must have AT LEAST ONE of the following:  decreasing U.S. consumption, decreasing U.S. investment, decreasing U.S. government purchases (including state and local government purchases in the U.S.), or decreasing U.S. net exports.  U.S. retail sales have been strong at least until recently based on reports from the U.S. Census Bureau.  If U.S. consumption continues to increase due to (a) the stimulus efforts that were implemented and (b) pent up demand following the COVID-19 shutdowns, then a recession would need to be due to a decrease in investment in physical capital by firms, decreases in government purchases by the government sectors, and/or a decrease in net exports, and the decrease or the decreases must be sufficiently large to offset the increase in consumption spending (and other increases that may occur).

Does the failure of the ONE-TO-TEN-YEAR U.S. Treasury yield curve to invert thus far in the year 2022 indicate that bond market participants are forecasting continued growth in the sum of U.S. consumption spending, U.S. investment in physical capital, U.S. government purchases, and U.S. net exports?  Does this mean that the U.S. economy is not headed for a recession within the next two years?  To the extent that the findings of Bauer and Mertens (2018) hold, a U.S. recession in the short term becomes much more likely if the one-to-ten-year U.S. Treasury yield curve inverts.

Of course, a negative aggregate supply shock such as the recent surge in oil and gasoline prices could lead to a recession.  Readers may remember that the U.S. experienced recessions in the middle of the 1970s and the early 1980s around the times of energy crises.  If my memory is correct, then with the goal of reducing inflation, the Fed tightened monetary policy before both of those recessions, and the one-to-ten-year U.S. Treasury yield curve inverted before both of those recessions.

The Federal Open Market Committee (FOMC) of the Federal Reserve Bank has already begun raising interest rate targets with the goal of reducing inflation as inflation has dramatically increased recently in the U.S.  If the Fed raises shorter term rates substantially and longer term rates do not increase sufficiently, then that could cause the one-to-ten-year U.S. Treasury yield curve to invert, which Bauer and Mertens (2018) may interpret as a clear signal for a recession.  But, if Bauer and Mertens’ (2018) finding (or suggestion or conclusion) that the one-to-ten-year U.S. Treasury yield curve is really the one to watch, then at this point the yield curve has not yet signaled a recession coming in the U.S. within the next two years (if not longer).  Will further tightening by the FOMC, removal of expansionary fiscal policy, changes in the supplies of and the demands for U.S. Treasury debt instruments, a decrease in businesses’ and households’ desires to spend, other factors, or some combination wind up resulting in (a) an inverted one-to-ten-year U.S. Treasury yield curve and (b) subsequently a U.S. recession?  Is the one-to-ten-year U.S. Treasury yield curve really the one to watch?  Vigilant eyes cast on the one-to-ten-year U.S. Treasury yield curve, other yield curves, and other data will find out in time.

 

REFERENCE

Bauer, M. D. and Mertens, T. M. (2018). "Economic Forecasts with the Yield Curve." March 5, 2018. FRBSF Economic Letter. 2018-07. Available online at:  https://www.frbsf.org/economic-research/publications/economic-letter/2018/march/economic-forecasts-with-yield-curve/ 

JUNE 21, 2022 ADDENDUM

I noticed today on CNBC that the five-to-ten-year U.S. Treasury yield curve was reinverted.  I do not know when that yield curve reinverted.  Please accept my apology for any inconvenience or confusion.  

However, if the one-to-ten-year U.S. Treasury yield curve is the most reliable in predicting upcoming U.S. recessions, the main message of this blog post remains unchanged.  We can continue monitoring data to improve our perception of the direction of the U.S. economy.