Short-term forecasting of tourist arrivals in Malta with google trends data

Abstract

The aim of this dissertation is to assess whether a model with Google Trends search query data is able to provide more accurate forecasts of tourist arrivals in Malta than a model without such data. Google Trends search query data is collected following a careful methodological process to select queries that are related to travelling to Malta. A Seasonal Autoregressive Integrated Moving Average model is employed as a benchmark model. Principal Component Analysis is conducted on the Google Trends data collected and the principal components are added as predictors to the benchmark model to obtain the competing Google Trends model. Pseudo-out-of-sample one-step ahead forecast simulations are carried out from both models estimated on the period January 2004 – December 2016 and their forecasting performance is tested on the sample of observations from January 2017 – December 2019 using measures of the mean error. The results reveal that the model with Google Trends data generates better forecasts as the forecast accuracy metrics show a lower error for the forecasts on the testing sample. The robustness of these results is checked via the Clark & West (2007) test for nested models and another pseudo-out-of-sample one-step ahead forecast simulation from a model with European Gross Domestic Product and the Real Effective Exchange as income and price variables, respectively, which are typical explanatory factors of tourism demand. The forecasts from the Google Trends model are robust against these checks which allowed for a demonstration of a practical use of this model in a nowcasting exercise. Tourist arrivals for July 2022 and August 2022 were nowcasted from the model with the Google Trends data which included a dummy variable for April, May and June 2020 to control for the period in which the airport was closed as a measure against the spread of Covid-19. These findings elicit important implications for private sector businesses in the tourism industry, public sector institutions such as the Malta Tourism Authority, national institutions concerned with projections of the Maltese macroeconomy like the Central Bank of Malta and various other sectors in the economy which are impacted by the indirect and induced effects of tourist activity.