The Fragility of Robust Causality Analysis : Finite-Sample Analysis and the Inflation-Economic Growth Relationship

Surplus-lag testing has been proposed as a means of undertaking persistence-robust causality analysis irrespective of the integrated nature of economic and/or financial series under examination. The present paper examines whether this suggested robustness holds when considering series experiencing changes in their orders of integration or persistence. Using Monte Carlo simulation, the results of finite-sample analysis demonstrate that while surplus-lag tests outperform Granger causality tests in these circumstances, they can nonetheless exhibit severe size distortion. The empirical relevance of this previously undetected fragility is reinforced by an examination of the relationship between inflation and economic growth in the USA. Recognising recent research documenting the extensive application of causality analysis within corporate finance and the noted occurrence of changes in persistence in financial series, the results obtained provide a cautionary note for practitioners anticipating persistence-robust inference when undertaking surplus-lag causality analysis