I am conducting a research using times series data using ARDL bounds test to co-integration in which long run and short run relationship among variables of interest are examined. In the short run relationship the obtained result shows that the coefficient of ECT is greater than 1. How can it be justified? Someone who has an idea on the issue can help!