Premium
A Note on the Power of Money‐Output Causality Tests
Author(s) -
Cheung YinWong,
Fujii Eiji
Publication year - 2001
Publication title -
oxford bulletin of economics and statistics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.131
H-Index - 73
eISSN - 1468-0084
pISSN - 0305-9049
DOI - 10.1111/1468-0084.00219
Subject(s) - arch , heteroscedasticity , econometrics , causality (physics) , autoregressive model , economics , monte carlo method , power (physics) , sample (material) , statistics , mathematics , engineering , civil engineering , physics , chemistry , chromatography , quantum mechanics
This study suggests that some empirical findings against money‐output causality can be the consequence of ignoring autoregressive conditional heteroskedastic (ARCH) errors. Monte Carlo results confirm that ARCH effects drastically reduce the power of the standard causality test. The maximum likelihood approach allowing for ARCH effects, on the other hand, provides a good power performance. Using different specifications and sample period, Friedman and Kuttner (1993) and Thomas (1994) report limited evidence of money causing output. We detect significant ARCH effects in the models considered by these studies. Once ARCH effects are explicitly accounted for, we find that the monetary effect is significant though its magnitude is quite small.