Expected inflation and interest-rate dynamics in the COVID era: evidence from the time–frequency domain

Abstract

The onset of the COVID-19 pandemic in the United States may have led investors or other individuals to expect sharp drops in output and rising prices, as well as drastic changes in fiscal and/or monetary to deal with the crisis. This paper analyses the co-movement between expected inflation and interest in the U.S. by using a battery of wavelet tools over the period from January 21, 2020 to March 28, 2022. Wavelet methods are used to examine the linkages between expected inflation and nominal interest rates of varying terms, focusing on the direction of co-movement and their sub-time horizons. Both bivariate wavelet and partial wavelet models that incorporate daily COVID-19 case counts or a financial stress variable find that the relationship holds primarily in the longer short-run (more than 6 months), with connections stronger for maturities of 5 years than for 1 year or less. The expectation related to the ‘inflation–interest rate’ nexus and type of bond maturity seem to be significantly shaped by the pandemic peak and anticipated duration of the disease. More precisely, the longer the anticipated duration of the pandemic is, the higher the expected inflation rate, bond yield rate, and maturity are. The interaction between expected inflation and interest seems to be very sensitive to pandemic and financial stress in terms of lead-lag status, in the very short to short-run, for 5 and 10 years bond maturity. This seems to be explained by investor hazard to a new particular unknown stimulus caused by the pandemic and its socio-economic consequences.