A Causality Test in Time Series Analysis

DOI IR (HANDLE)

Bibliographic Information

Other Title
  • 時系列解析における因果検定

Search this article

Description

In a practical application of Sims causality tests based on Granger definition of causality, some improvements have been attempted with respect to the stationarizing process of time series data, the estimation of the orders of optimal lags and leads, the correction of serial correlations through optimal filtering with the aid of Box = Pierce χ^2 tests, and the judgement of causal intensity. The test has been applied to the relationships among macroeconomic variables and its results are almost acceptable. The test seems to acquire improved precision, while the other methods cannot remove serial correlations almost perfectly nor bring about acceptable results.

Journal

Keywords

Details 詳細情報について

Report a problem

Back to top