| Summary: | 碩士 === 中國文化大學 === 經濟學系 === 98 === Tourism is a rapid growing multi-national industry. It can promote economic growth by enhancing domestic employment opportunities and generate income from the domestic expenditure of foreign tourists. Since 1956, Taiwan developed its tourism by positively revised the relevant laws with policies of openness. The number of foreign tourists in Taiwan increased, since then. Currently, innovation and further development in Taiwan’s tourism market is one of the most important issues for Taiwan’s economic growth. The government promote and market Taiwan’s tourism overseas. The number of Japanese tourists visited Taiwan had exceeded one million in 2005.
This study conducted a time series analysis and forecasting the number of foreign tourism for Taiwan. A number of tourism literatures are related to the demand analysis and univariate analysis. Alternatively, this study paid attention to do study on the number of Taiwan’s main foreign tourists and do the forecast. Five main foreign regions or countries are included in this study: (1) Hong Kong and Macao, (2) Japan, (3) South Korea, (4) Southeast Asia and (5) the United States. Both monthly and quarterly data from 1991 to 2009 are collected to simulate 4 time series models: (1) ARIMA model (Autoregressive Integrated Moving Average), (2) ARIMA with seasonal adjustment by X-12 method (3) ARIMA with seasonal and shock adjustment by dummy variables, and (4) SARIMA model (Seasonal Autoregressive Integrated Moving Average).
The four models are simulated by using data from 1991 to 2004, and forecast ability is examined by comparing the forecast and actual values by using data from 2005 to 2009 with criteria of the Mean Absolute Percentage Error (MAPE), Root Mean Squared Error (RMSE) and Theil Inequality Coefficient (Theil's U). The predictive capabilities suggest that the 4 models are all satisfying with both monthly and quarterly data.
The study found that a variety of models has a good predictive ability, among which the SARIMA model simulated by monthly data is the most accurate one. This most accurate model was used to forecast the number of foreign tourism from 2010 to 2011. Comparing the actual values in 2008 and 2009 and the sequential forecast values in 2010 to 2011, a slightly declining number occurs in Hong Kong and Macau, South Korea and Southeast Asia, while a growth trend was found in Japan and the United States. For the declines, domestic adjustment is demanded.
|
|---|
