.png){kind=icon}
.png){kind=icon}
Return
Testing for frequency domain causality in the vector error correction model
.png){kind=icon}
PRE
AI
DOI：10.1080/03610926.2026.2669818.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
In this article, based on the frequency domain causality measure of Granger and Lin [. Econ Theory 11 (3), 530-536] and Hosoya [2001. J Time Series Anal. 22 (5), 537-554] in the vector error correction model, we propose a frequency domain causality test procedure for cointegrated time series which is very intuitive and simple. The test is based on a set of linear hypotheses on the loading and short-run coefficients of the vector error correction model. Thus, the null hypothesis corresponding to the proposed test can be tested easily by the usual Wald test statistic. The power properties of the test are theoretically analyzed based on a simple model. Meanwhile, the finite sample properties of the test are examined by Monte Carlo experiments. Finally, as an illustration, we investigate the predictive power of dividends for stock prices in the frequency domain. We find that dividends can predict S&P 500 stock prices not only at certain low frequencies, but also at certain high and business cycle frequencies.
Keywords:
C12
C32
Frequency domain causality test
cointegrated time series
vector error correction model
Journal
C
IF:
0.8
Papers:
211
Citations:
0