ABD’nin Ekonomi Politikası Belirsizliği ve Finansal Baskı Endeksleri Arasındaki İlişkinin Araştırılması: Fourier Serisi Yaklaşımı Yöntemlerinden Kanıtlar

Abstract

We investigate the relationship between economic policy uncertainty (EPU) and St. Louis Fed's financial stress (FS) indices for the US by using monthly data for the period 2013:1 -2019:6 and employing linear (conventional) as well as nonlinear (exponential) unit root tests; nonlinear (exponential smooth transition autoregressive-ESTAR) cointegration test initially introduced by Kapetanios, Shin, and Snell (2006) (KSS) and residual-based Fourier cointegration test suggested by Yılancı (2019); conventional and Fourier Granger causality tests as well as asymmetric causality tests. Empirical findings from these procedures can be classified into three major categories: (i) The results from the KSS and residual-based Fourier cointegration analyses confirm each other that a long-run equilibrium exists between EPU and FS. (ii) Estimations from the Fourier Granger causality test that allows for structural breaks of unknown number and form unveiled that there is a one-way causality running from FS to EPU, a finding that contrasts with the one from the conventional procedure which shows a two-way causality. (iii) Finally, the findings from the asymmetric causality testing procedure verified that while two unidirectional causalities exist running from the negative and positive components of FS to the negative and positive components of EPU, respectively; we found no evidence for such asymmetric causality running from EPU to FS. These robust findings we believe shed a bright light on a major policy suggestion. The US policy makers should design policies and regulations aiming at lessening the stress on the financial markets in order to leash the uncertainty associated with economic policies.