This study aims to evaluate the performance of various time-series and econometric models' long-term forecasts of US visitor arrivals to China. The autoregressive lag distributed model, time varying parameter model, vector autoregressive model, exponential smoothing model, autoregressive integrated moving average model and the na ve model are included in the comparison. A wide range of forecasting horizons, from one year ahead to seven years ahead, are considered in this exercise. In addition, the performance of forecast combinations is further examined over all horizons under consideration. The empirical results indicate the overall best-performing model for forecasting the demand for China tourism by US residents. The ability of forecast combination in improving forecasting accuracy is assessed too.
|Title of host publication||[Missing Source Name from PIRA]|
|Number of pages||4|
|Publication status||Published - 2013|
|Event||CAUTHE Annual Conference - |
Duration: 1 Jan 2013 → …
|Conference||CAUTHE Annual Conference|
|Period||1/01/13 → …|