“…Furthermore, Marques et al (2016) points out a fourth advantage, noting that unlike traditional cointegration techniques, the inclusion of dummy variables does not vitiate the efficiency of the model. Therefore, in view to these advantages, the number of studies employing ARDL in estimating relationships amongst various time-series variables of interest has skyrocketed significantly over the last 17 years (example, Tursoy and Faisal, 2017;Sun et al, 2017;Rahman and Kashem, 2017;Marques et al, 2016;Narayan, 2002Narayan, , 2014Murthy and Okunade, 2016;Odhiambo, 2009;Fuinhas and Marques, 2012;Acaravci, 2010, 2011, to mention just very few).…”