Measuring the strength of cointegration and Granger-causality

Abstract

Standard-Nutzungsbedingungen:Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden.Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen.Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may

Causes and hazards of the euro area sovereign debt crisis: Pure and fundamentals-based contagion

“…The results (not shown here to save space but available from the authors upon request) render the same qualitative conclusions as when 200 observations were used.22 Since the estimations are based only on the most recent portion of the data, our rolling window approach accommodates parameter variability and is robust to small samples and presence of multiple structural breaks and nonlinearities, providing evidence of existence or otherwise of temporal causal relationship (in-sample predictability over time) between the variables under study.23 We perform formal tests to evaluate whether the series have the same mean during the detected episodes and the rest of the observations. The results of these tests (not shown here, but available from the authors upon request) strongly reject the null hypothesis of equal mean across sub-samples, and provide additional support for the presence of increased Using the framework for grading the strength of the Granger-causality relationship proposed byAtukeren (2005) we obtain the same classification of episodes of causality intensification Atukeren (2005)Jeffreys (1961)'s Bayesian concept of grades of evidence.…”

Christmas cards, Easter bunnies, and Granger-causality

“…63-64) for a discussion on causality and significance testing. See also Atukeren (2005) for a Bayesian posterior odds ratio-based alternative to classical significance testing in the context of Granger-causality tests. 1951 1957 1963 1969 1975 1981 1987 1993 1999 2005 ESDI Index…”

Informed Trading, Flow Toxicity and the Impact on Intraday Trading Factors

“…Hence, 1 < 0 . As discussed in Atukeren (2005) the question arises as to how confidently it can be suggested that causality exists when the condition 1 < 0 is met. Conventionally, Ftests, likelihood ratios or Walds tests for joint significance testing are applied.…”

“…However, the application of classical statistical significance testing with the use of a Bayesian cost function for lag selection is conceptually problematic. Therefore, in order to consistently apply SBIC throughout Granger causality testing, Atukeren (2005) suggests the employment of the framework by Poskitt and Tremayne (1987). This framework originates from Jefferys' (1961) concept of 'grades of evidence', which questions the uniqueness of the best model.…”

“…As discussed in various econometrics literature (see Hsiao (1979), Kang (1989), Thornton and Batten (1984) and Atukeren (2005)), lag specification greatly affects the results from Granger causality testing via a VAR framework. In this paper, existing research is extended by employing a robust measure explained in Atukeren (2005), combining Hsiao's (1979) lag selection procedure and Poskitt and Tremayne's (1987) posterior odds ratio test based on the Bayesian Information Criterion to test for Granger causality rather than conventional Walds tests used in the existing literature. The authors believe this flexible Bayesian approach based on "grades of evidence" provides a sound framework for linear time series identification.…”