|
This paper critically evaluates the literature on the volatility of tourism demand, especially as it pertains to the Caribbean region with the objective of producing models and forecasts that rectify the major problems highlighted in the literature. ARCH, GARCH, MS, MS-VAR and MGARCH processes are estimated to derive short-run estimates of own market volatility, volatility persistence in the long-run, and cross spill-over short and long-run effects in the markets for the period 1999 to 2010. The results showed the markets displaying the highest own market volatility are Dominican Republic (0.77), The Bahamas (0.36), and Saint Vincent and the Grenadines (0.28), followed by Grenada (0.24). Antigua and Barbuda (0.89), and Anguilla, Aruba, The Cayman Islands and Jamaica (all 0.87) revealed the highest long-run volatility persistence after an unexpected shock.
In assessing the asymmetric models using EGARCH and TGARCH processes, the paper found that with most markets negative shocks increase volatility, and had a greater impact than positive shocks. For Anguilla, Antigua and Barbuda, Dominica, Puerto Rico, Saint Lucia, Curacao and Trinidad and Tobago positive shocks were greater than negative shocks and they increased volatility in these markets.
There was similarity of results between the MS model and CGARCH as Dominica and the Dominican Republic had the fastest mean convergence results. In addition to these markets the MS model found Jamaica (0.70), Anguilla (0.67) and Saint Lucia (0.65) with autoregressive coefficients close to unity. The transitional probabilities showed that Trinidad and Tobago (0.4590) and Dominica (0.4078) were the quickest markets to switch to normal growth after a downturn within a quarter.
|
University College of the Cayman Islands
