Nonlinear Taylor rules: evidence from a large dataset

Abstract

In this paper we estimate nonlinear Taylor rules over the 1986–2008 sample time period and augment the traditional Taylor rule by including principal components to better model Federal Reserve policy. Including principal components is useful in that they extract information about the overall economy from multiple economic indicators in a statistically optimal way. Additionally, given that uncertainty may influence Federal Reserve decisions, we incorporate an uncertainty index in the reaction function of the Federal Reserve. We find substantial evidence that the Federal Reserve responded to increases in macroeconomic uncertainty by cutting the Federal Funds rate over the sample period. We also find evidence that the Federal Reserve responded aggressively to increases in capacity utilization, especially when the inflation rate was above 2%.

Conventional and unconventional monetary policy reaction to uncertainty in advanced economies: evidence from quantile regressions

“…On the contrary, Giannoni (2002) and Sonderstrom (2002), among others, have presented evidence that supports an aggressive reaction of monetary policy under uncertainty. More recently, Caggiano, Castelnuovo, and Nodari (2018) and Ma, Olson, and Wohar (2018) found strong evidence indicating that the Fed responded to increases in financial and macroeconomic uncertainty respectively, by cutting the Federal Funds rate. The fact that the Fed did respond to macroeconomic uncertainty in the pre-zero lower bound (ZLB) period has also been earlier confirmed by Evans et al (2015), when estimating modified Taylor-type monetary policy rules.…”

“…In the process, we, for the very first-time, contribute to the existing literature on quantile-based estimation of Taylor rules by analysing the behaviour of major central banks to not only inflation and output gap, but also to uncertainty, across the periods of conventional and unconventional monetary policy decisions. In sum, while studies like Evans et al (2015), Caggiano, Castelnuovo, andNodari (2018), andMa, Olson, and have included uncertainty in the Taylor-rules, their sample have been restricted to the pre-ZLB era only and results derived based on a conditional-mean model, we use a quantiles-based approach that allows us to study the entire conditional distribution of the interest rate response. In the process, we also add to the literature on quantiles-based Taylor rule of Chevapatrakul, Kim, and Mizen (2009), Wolters (2012), Chen andKashiwagi (2017), andLiu (2018) by incorporating the role of uncertainty in the model, which has been shown to be important in interest rate setting behaviour by Evans et al (2015) and Ma, Olson, and Wohar (2018) in conditional-mean based monetary policy rules.…”

Monetary Policy Setting in Australia, Canada and the Euro Area: Insights from the Shadows

“…To evaluate similarities or differences in preferences, we estimate a monetary policy reaction function of the form.Where i is the policy interest rate; I is a country/economy indicator; t indicates the month; refers to the central bank, SMPC, or an individual member of the SMPC; ; r is the long‐term real neutral interest rate; is the inflation gap; and is the output gap. Variants of Equation () to incorporate the possibility of nonlinearities or additional determinants in the policy rate setting decision have been proposed (e.g., Ma et al , , and references therein). However, versions of Equation () remain popular in the literature.…”

Unconventional monetary policy reaction functions: evidence from the US

“…Indeed, despite numerous extensions of Taylor rule, 4 these have typically been formulated in standard monetary policy frameworks (where short-term policy rates respond to fluctuations in the output gap and the inflation rate) 5 or "leaning against the wind" monetary policies (where policy rates are set at a somewhat higher level than what would be required for pure inflation targeting reasons to safeguard financial stability). 6 More recently, Caggiano, Castelnuovo, and Nodari (2018) and Ma, Olson, and Wohar (2018) consider the response of central bank to heightened financial and macroeconomic uncertainty, and show that it normally encompasses aggressive policy rate cuts. This result is also uncovered in previous studies, such as Giannoni (2002) and Sonderstrom (2002), but Kuttner and Posen (2001), Bernanke andEvans et al (2015) point to an asymmetric reaction by the central bank depending on whether policy rates are near or far away from the zero lower bound.…”