1) The ARDL bounds test is used to investigate the long-run relationship and short-run dynamics between variables. It has three benefits over other cointegration tests: it does not require variables to be cointegrated of the same order, it is more efficient for small samples, and it allows an unbiased estimate of the long-run model.
2) Granger causality within a VECM identifies the direction of causality between variables. The existence of causality is shown through F-statistics and the lagged error-correction term indicates long-run relationships. The short-run effect is shown through F-statistics on independent variables and the t-statistic on the error-correction term indicates long-run
1) The ARDL bounds test is used to investigate the long-run relationship and short-run dynamics between variables. It has three benefits over other cointegration tests: it does not require variables to be cointegrated of the same order, it is more efficient for small samples, and it allows an unbiased estimate of the long-run model.
2) Granger causality within a VECM identifies the direction of causality between variables. The existence of causality is shown through F-statistics and the lagged error-correction term indicates long-run relationships. The short-run effect is shown through F-statistics on independent variables and the t-statistic on the error-correction term indicates long-run
1) The ARDL bounds test is used to investigate the long-run relationship and short-run dynamics between variables. It has three benefits over other cointegration tests: it does not require variables to be cointegrated of the same order, it is more efficient for small samples, and it allows an unbiased estimate of the long-run model.
2) Granger causality within a VECM identifies the direction of causality between variables. The existence of causality is shown through F-statistics and the lagged error-correction term indicates long-run relationships. The short-run effect is shown through F-statistics on independent variables and the t-statistic on the error-correction term indicates long-run
According to the Belloumi, M. (2013) this test is used to investigate the long-run relationship and short-run dynamic co-integrating among the variables. It encompasses of three benefit if compare with other traditional co integration measurements. Firstly, ARDL is not necessary that all the variables under the research to be inter-connected at the same order. It can also applicable if those variables are cointegrating in different order such as in order one or order zero. Secondly, ARDL is relatively more efficient and suitable with small and finite sample size. Lastly, an unbiased investigation of the long-run model can be obtained by applying the ARDL techniques. Other works using ARDl such as (Yue,Y. and Serge, N.B.Z., 2010; Aseidu, M.K., 2013; Sakyi, D., Commodore, R., Opoku, E.E.O., 2015). 2.2.3 Granger-short run and long-run causality tests Granger causality within VECM is applied to identify the flow of direction of causality between the variables. The existence of at least one direction which is denoted by the F-statistics and the lagged errorcorrection term is the criteria of long run relationship between variables. The short run causal effect is denoted by the F-statistics on the independent variable meanwhile the t-statistics on the coefficient of the lagged error-correction term depicts the long-run relationship. Other works using Granger causality are (Yue, Y. and Serge, N.B.Z., 2010; Kakar, Z.K. and Khilji, B.A., 2011).