On the Cyclical Properties of Hamilton's Regression Filter and Refinements

We document shifts in the lead‐lag properties of the U.S. business cycle since the mid‐1980s. Specifically, (1) the well‐known inverted leading indicator property of real interest rates has completely vanished; (2) labor productivity switched from positively leading to negatively lagging output and labor inputs over the cycle; and (3) the unemployment rate shifted from lagging productivity negatively to leading positively. Many contemporary business cycle models produce counterfactual cross‐correlations revealing that popular frictions and shocks provide an incomplete account of business cycle comovement. Determining the underlying sources of these shifts in the lead‐lag properties and their consequences for macroeconomic forecasts is therefore a promising direction for future research. (JEL E24, E32, E43)

show abstract

“…Schüler () presents a detailed assessment of the Hamilton () filter and discusses some of its shortcomings.…”

mentioning

confidence: 99%

The Shifts in Lead‐lag Properties of the U.S. Business Cycle

Brault

Khan

2019

Economic Inquiry

View full text Add to dashboard Cite

show abstract

“…The HP filter is not, of course, the only method of extracting unobserved trends from time series data, and is not without critics (Hamilton, 2018). Alternatives proposed by critics of the HP filter have, themselves, been subject to criticism, however, resulting in renewed enthusiasm for the HP filter in some quarters (Schüler, 2019). In what follows we set aside debate over the properties of alternative methods of filtering data.…”

Section: Methodsmentioning

confidence: 99%

Hysteresis in the Normal Rate of Capacity Utilization: A Behavioural Explanation

Setterfield

Avritzer

2019

SSRN Journal

View full text Add to dashboard Cite

Kaleckians describe a normal rate of capacity utilization that is subject to hysteresis effects. This means that the normal rate varies directly with the actual rate of capacity utilization, ensuring that steady-state equilibrium conditions in the Kaleckian model are fully adjusted (the actual and normal rates of capacity utilization are equalized) but without this last condition implying that the rate of capacity utilization is constant in the long run. The relationship between distribution and growth unique to the Kaleckian model is thus preserved. The hysteresis mechanism has been criticized from various quarters, however, these criticisms focusing on its alleged lack of behavioural foundations. This paper shows that consistent with the stylized facts, variation in the normal rate of capacity utilization in response to variation in the actual capacity utilization rate can be derived from the links between both variables and the volatility of the macroeconomic environment-volatility, in the presence of fundamental uncertainty, being an important reason why firms deliberately under-utilize capacity (even in the long run) in the first place. The result is an empirically-grounded behavioural foundation for hysteresis in the normal rate of capacity utilization.

show abstract

“…The process of extracting unobserved trends from raw data is a vexed and vexing issue, as recent contributions focused on this issue attest (Hamilton, 2018;Schüler, 2019). The process of then imbuing these trends with behavioural meaning -as when the trend in measured capacity utilization rates (however extracted) is interpreted as the normal rate of capacity utilization -further complicates matters.…”

Section: Estimating the Normal Rate Of Capacity Utilizationmentioning

confidence: 99%