Resumen de Research note:Forecasting tourism demand by disaggregated time series - empirical evidence from Spain
Glauber Eduardo de Oliveira Santos
Frequently, tourism demand can be disaggregated into different components according to variables such as country of residence, purpose of the trip, type of transport and accommodation. However, researchers generally develop forecasts of the total tourism demand without considering the existence of its disaggregated components, which might have independent behaviour. As an alternative, the disaggregated approach models each component first and then sums these individual forecasts in order to obtain aggregate forecasts of the total tourism demand. This paper compares the aggregated and disaggregated approaches by ex post forecasting international tourist arrivals in Spain. The total tourism demand is disaggregated into 12 different origins. The HEGY test is used to check for regular and seasonal unit roots in each time series and SARIMA models are used to develop single forecasts. This empirical study finds slightly more accurate results using the disaggregated approach in multi-step, outof- sample forecasting.
