Modelling the real exchange rate misalignment in the presence of outliers for developing countries

Persistent real exchange rate misalignment is sometimes explained by the presence of a nonlinear adjustment process towards the long-run equilibrium. However, although the nonlinearity may be a feature for some real exchange rates, outliers and nonlinearity may easily be confused. The main purpose of this paper is to make the distinction between nonlinearity and outliers using robust estimation methods of maximum likelihood. Also, the nature of the real exchange rate misalignment is specified, based on the value of fundamentals, for a sample of African countries during the period from 1960 to 2015. According to the robust estimation procedures, the results of linearity tests are particularly interesting. Indeed, the apparent nonlinearity is significantly reduced when considering outliers, especially for South Africa, Ghana, Madagascar and Morocco. We can conclude that the source of the nonlinearity of the real exchange misalignment rate for the countries already mentioned is the existence of outliers in the considered series.