A Comparison of Two Econometric Models (OLS and SUR) for Forecasting Croatian Tourism Arrivals
Keywords
tourism demand, forecasting
Tourism receipts have a large impact on the Croatian economy. The large inflow of foreign exchange during the summer season provides not just income but also a stabilising effect on the local currency, the kuna. The author compares two demand models using OLS and SUR estimation techniques. The model is a system of equations covering five countries, which represent around 72%-78% of total foreign annual arrivals. The model describes arrivals to be a function of the home country's real GDP and the real exchange rate. Based on estimates of forecasting accuracy, it seems that the SUR model yields more precise predictions of foreign arrivals.