Asymmetric panel causality test: An application on stock exchanges of developing countries

Causality which means the relationship between variables or events that have been studied by many disciplines so far. Even if the history of causality begins with philosophy later on the quantitative science such as statistics and econometrics has dealt with this topic. Determining the existence of relationships between economic variables are the basis of many existing theories in the economics literature. Therefore, the importance of causality cannot be ignored by the literature of economics. Econometrics gives the opportunity to researchers testing the validity of economic theory. Many of these theories are explained by the help of causality tests. In this respect, the causality relationships and testing of these are extremely important. The asymmetry has many different definitions. Asymmetry of variables means an economic time series with different responses to positive and negative shocks. Responses to the shocks of the economic variables may vary. The ignorance of these differences can not reveale the relationship existing between the variables. This situation significantly reduces the reliability of the analysis. If we consider the asymmetry, the presence of hidden relationships between variables can be revealed. Especially in the more volatility of variables analyzing asymmetric relations leads to obtain more reliable results instead of symmetric ones. This study aimed to develop asymmetric causality test for panel data because of the lack of literature. In this way, using asymmetrical components of variables between them for each unit in the panel can be analyzed separately. This theory has been applied to investigate the asymmetrical relationship between stock market indices and the foreign exchange rates for developing 11 countries. In our thesis the relations between variables has been examined by symmetric causality test. Finally, the hidden relations which can not be revealed by the symmetric panel causality tests are revealed by newly developed asymmetric panel causality test.