Evidence for Nonlinear Asymmetric Causality in US Inflation, Metal, and Stock Returns

Table 2

Test for nonlinear causality (asymmetric case for positive returns). If 𝑃>.05, then at 5% we accept the null hypothesis that 𝐴 does not cause 𝐵. The parameters for the M-G model were 𝜏₁=3, 𝜏₂=1, and 𝑐₁=𝑐₂=2.

Relation (𝐴→𝐵)	F-statistic	Probability
DLCPI^p→ DLMET	2.6389	0.1049
DLMET^p→ DLCPI	18.9806	1.6E-5