%PDF-1.3
%
1 0 obj
<<
/CreationDate (D:20091118144739Z00'00')
/ModDate (D:20091118144743Z)
/Producer (Acrobat Distiller 5.0 \(Windows\))
/Author (GIG)
/Title (TE8\(1\))
/Subject (TE8\(1\))
/Creator (Adobe PageMaker 7.0)
/Keywords ()
>>
endobj
2 0 obj
<< /Type /Metadata /Subtype /XML /Length 1144 >>
stream
TE8(1)TE8(1)
endstream
endobj
3 0 obj
<<
/Type /Page
/Parent 10 0 R
/Resources 5 0 R
/Contents 4 0 R
>>
endobj
4 0 obj
<< /Length 4012 >>
stream
BT
/F1 1 Tf
9 0 0 9 56.76 655.135 Tm
0 0 0 rg
/GS2 gs
0.0082 Tc
0.1611 Tw
[(T)99.2(ourism Economics)]TJ
/F2 1 Tf
6.72 0 TD
0.0078 Tc
0.1639 Tw
(, 2009, 15 \(4\), 693708)Tj
/F3 1 Tf
18.5 0 0 18.5 89.04 591.415 Tm
-0.0118 Tc
0.1779 Tw
[(Ef)14.1(fect of seasonality treatment on the)]TJ
0.5059 -1.0832 TD
0.1793 Tw
(forecasting performance of tourism)Tj
4.2292 -1.0768 TD
-0.013 Tc
0 Tw
[(demand)-246.5(models)]TJ
/F2 1 Tf
11.65 0 0 10.5 206.16 507.775 Tm
0 Tc
(S)Tj
9.3 0 0 8.4 211.44 507.775 Tm
0.0055 Tc
(HUJIE)Tj
11.65 0 0 10.5 237.48 507.775 Tm
0.1609 Tc
( S)Tj
9.3 0 0 8.4 247.44 507.775 Tm
0.0085 Tc
(HEN)Tj
/F1 1 Tf
10.35 0 0 11.5 95.04 489.175 Tm
0.0028 Tc
0.1879 Tw
[(Institute for T)95.6(ransport Studies, University of Leeds, Leeds LS2 9JT)107.1(, UK.)]TJ
8.1275 -1.1374 TD
0.0029 Tc
0.2608 Tw
(E-mail: s.shen@its.leeds.ac.uk.)Tj
/F2 1 Tf
11.65 0 0 10.5 215.4 450.415 Tm
0 Tc
0 Tw
(G)Tj
9.3 0 0 8.4 223.92 450.415 Tm
0.0068 Tc
(ANG)Tj
11.65 0 0 10.5 244.2 450.415 Tm
0.1564 Tc
( L)Tj
9.3 0 0 8.4 254.88 450.415 Tm
0 Tc
(I)Tj
/F1 1 Tf
10.35 0 0 11.5 94.32 431.695 Tm
0.0027 Tc
0.189 Tw
[(School of Management, University of Surrey)25.9(, Guildfor)25.9(d GU2 7XH, UK.)]TJ
8.9275 -1.127 TD
0.003 Tc
0.2223 Tw
[(E-mail: g.li@surrey)37.8(.ac.uk.)]TJ
/F2 1 Tf
11.65 0 0 10.5 202.44 392.935 Tm
0 Tc
0 Tw
(H)Tj
9.3 0 0 8.4 211.68 392.935 Tm
0.0032 Tc
[(AIY)93.5(AN)]TJ
11.65 0 0 10.5 240 392.935 Tm
0.1712 Tc
( S)Tj
9.3 0 0 8.4 249.96 392.935 Tm
0.0116 Tc
(ONG)Tj
/F1 1 Tf
10.35 0 0 11.5 62.04 374.335 Tm
0.0027 Tc
0.1879 Tw
[(School of Hotel and T)95.5(ourism Management, Hong Kong Polytechnic University)49.1(, Hung Hom,)]TJ
4.7768 -1.127 TD
0.0024 Tc
0.1876 Tw
[(Kowloon, Hong Kong, PR China. E-mail: hmsong@polyu.edu.hk.)]TJ
/F2 1 Tf
10.5 0 0 10.5 94.08 334.015 Tm
0.0042 Tc
0.0905 Tw
(This study provides a comprehensive comparison of the performance)Tj
0 -1.0514 TD
0.1581 Tw
[(of the commonly used econometric and time-series models in fore-)]TJ
0 -1.04 TD
0.0043 Tc
0.091 Tw
(casting seasonal tourism demand. The empirical study is carried out)Tj
0 -1.0514 TD
0.0995 Tw
(based on the demand for outbound leisure tourism by UK residents)Tj
T*
0.0047 Tc
0.2646 Tw
(to seven destination countries: Australia, Canada, France, Greece,)Tj
0 -1.04 TD
0.0045 Tc
0.1091 Tw
[(Italy)95.9(, Spain and the USA. In the modelling exercise, the seasonality)]TJ
0 -1.0514 TD
0.2767 Tw
(of the data is treated using the deterministic seasonal dummies,)Tj
T*
0.1583 Tw
(seasonal unit root test techniques and the unobservable component)Tj
0 -1.04 TD
0.0043 Tc
0.2809 Tw
(method. The empirical results suggest that no single forecasting)Tj
0 -1.0514 TD
0.0869 Tw
(technique is superior to the others in all situations. As far as overall)Tj
T*
0.2619 Tw
(forecast accuracy is concerned, the Johansen maximum likelihood)Tj
0 -1.04 TD
0.0042 Tc
0.0369 Tw
(error correction model outperforms the other models. The time-series)Tj
0 -1.0514 TD
0.1602 Tw
[(models also show superior performance in dealing with seasonality)95.6(.)]TJ
T*
-0.001 Tc
0 Tw
[(However)56.1(, the time-varying parameter model performs relatively poorly)]TJ
0 -1.04 TD
0.0042 Tc
0.3183 Tw
(in forecasting seasonal tourism demand. This empirical evidence)Tj
0 -1.0514 TD
0.0044 Tc
0.1862 Tw
(suggests that the methods of seasonality treatment affect the fore-)Tj
T*
0.1029 Tw
(casting performance of the models and that the pre-test for seasonal)Tj
0 -1.04 TD
0.1875 Tw
[(unit roots is necessary and can improve forecast accuracy)84.5(.)]TJ
/F1 1 Tf
0 -2.5371 TD
-0.0032 Tc
0 Tw
[(Keywor)19.7(ds)]TJ
/F2 1 Tf
3.2343 0 TD
-0.0034 Tc
[(:)-496.3(seasonality; tourism demand; forecasting; seasonal unit roots;)]TJ
-3.2343 -1.0514 TD
0.004 Tc
0.1953 Tw
(econometric model; time-series model)Tj
9 0 0 9 58.08 69.655 Tm
0.008 Tc
0.0819 Tw
(The work described in this paper was substantially supported by a grant from the Research Grants)Tj
0 -1.12 TD
0.1102 Tw
(Council of the Hong Kong Special Administrative Region, China \(Project No. PloyU 5473/06H\).)Tj
ET
endstream
endobj
5 0 obj
<<
/ProcSet [ /PDF /Text ]
/Font << /F1 6 0 R /F2 7 0 R /F3 8 0 R >>
/ExtGState << /GS2 9 0 R >>
>>
endobj
6 0 obj
<<
/Type /Font
/Subtype /Type1
/FirstChar 32
/LastChar 239
/Widths [ 250 389 333 500 500 833 667 222 333 333 500 600 250 333 250 278 500
500 500 500 500 500 500 500 500 500 250 250 600 600 600 500 800
667 611 611 722 556 556 722 722 333 389 611 556 833 722 556 556
611 611 444 611 722 611 889 722 667 611 389 278 389 600 500 278
444 389 278 444 278 278 389 444 278 278 444 278 611 444 333 389
444 333 278 278 444 389 611 444 389 389 389 222 389 600 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 500 0 0 0 0 0 0 0 0 0 250
0 500 500 0 0 0 0 278 800 0 0 0 333 0 0 0 600 0 0 0 444 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 278 ]
/Encoding /WinAnsiEncoding
/BaseFont /HPBNLD+GaramondThree-Italic
/FontDescriptor 61 0 R
>>
endobj
7 0 obj
<<
/Type /Font
/Subtype /Type1
/FirstChar 32
/LastChar 252
/Widths [ 250 278 333 500 500 833 667 222 278 278 500 600 250 333 250 278 500
500 500 500 500 500 500 500 500 500 250 250 600 600 600 389 800
667 611 611 722 556 500 722 778 333 389 722 500 833 778 722 556
722 667 444 611 722 667 1000 722 667 611 333 278 333 600 500 278
389 500 389 500 389 278 500 500 278 278 500 278 778 500 444 500
500 333 333 333 500 444 667 444 444 389 333 222 333 600 500 0 500
0 0 0 0 0 0 0 0 0 0 0 500 0 500 500 222 222 0 0 500 500 0 0 0 0
0 0 500 0 0 250 0 500 500 0 0 0 0 278 800 0 0 0 333 0 0 0 600 0
0 278 500 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 389 0 0 0 0 0 0 0 0 0 0 0 0 0 278
0 0 0 0 0 0 0 0 0 0 0 0 500 ]
/Encoding /WinAnsiEncoding
/BaseFont /HPBOKJ+GaramondThree
/FontDescriptor 63 0 R
>>
endobj
8 0 obj
<<
/Type /Font
/Subtype /Type1
/FirstChar 32
/LastChar 207
/Widths [ 250 278 333 500 500 833 722 222 278 278 500 600 250 333 250 278 500
500 500 500 500 500 500 500 500 500 250 250 600 600 600 389 800
722 667 611 778 556 556 722 778 389 389 778 556 833 778 722 611
722 667 500 667 778 722 1000 778 722 667 278 278 278 600 500 278
444 556 444 556 444 333 500 556 278 278 556 278 833 556 500 556
556 389 389 333 556 500 778 500 444 444 278 222 278 600 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 222 0 0 0 0 0 0 0 0 0 0 0 0 0 250
0 500 500 0 0 0 0 278 800 0 0 0 333 0 0 0 600 0 0 0 556 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 389 ]
/Encoding /WinAnsiEncoding
/BaseFont /HPBOOL+GaramondThree-Bold
/FontDescriptor 65 0 R
>>
endobj
9 0 obj
<<
/Type /ExtGState
/SA false
/SM 0.02
/OP true
/op true
/OPM 1
/BG2 /Default
/UCR2 /Default
/TR2 /Default
>>
endobj
10 0 obj
<<
/Type /Pages
/Kids [ 3 0 R 11 0 R 14 0 R 18 0 R 21 0 R 24 0 R 27 0 R 30 0 R 35 0 R 38 0 R
]
/Count 10
/Parent 44 0 R
>>
endobj
11 0 obj
<<
/Type /Page
/Parent 10 0 R
/Resources 13 0 R
/Contents 12 0 R
>>
endobj
12 0 obj
<< /Length 5874 >>
stream
BT
/F2 1 Tf
9.5 0 0 9.2 203.38 654.895 Tm
0 0 0 rg
/GS2 gs
0 Tc
0 Tw
(T)Tj
8.1 0 0 7.8 209.26 654.895 Tm
0.0025 Tc
(OURISM)Tj
9.5 0 0 9.2 239.5 654.895 Tm
0.1666 Tc
( E)Tj
8.1 0 0 7.8 248.86 654.895 Tm
0.0018 Tc
(CONOMICS)Tj
9 0 0 9 68.02 655.135 Tm
0.01 Tc
(694)Tj
11.5 0 0 11.5 68.02 635.455 Tm
0.0013 Tc
0.1449 Tw
(Seasonality is one of the most important features of tourism demand and has)Tj
0 -1.0435 TD
0.0913 Tw
[(important impacts on many aspects of the tourism industry)84.8(. Accurate forecasts)]TJ
T*
0.0151 Tc
0.3278 Tw
(of seasonal tourism demand are crucial for the formulation of effective)Tj
T*
0.0023 Tc
0.328 Tw
[(marketing)-248.1(strategies and tourism policies for both the private and public)]TJ
T*
0.0013 Tc
0.0242 Tw
(sectors. In addition, seasonality has also been recognized as one of the important)Tj
T*
0.1877 Tw
(tourism research areas \(Rodrigues and Gouveia, 2004\).)Tj
1.0435 -1.0435 TD
0.0011 Tc
0.0512 Tw
(Modelling seasonal variation in international tourism demand has become an)Tj
-1.0435 -1.0435 TD
0.0014 Tc
0.1394 Tw
[(important issue in tourism forecasting in recent years \(Kulendran and W)64(ong,)]TJ
T*
0.0015 Tc
0.0148 Tw
[(2005\). However)53.7(, most previous studies focused only on the time-series methods,)]TJ
T*
0.0011 Tc
0.3058 Tw
(such as the seasonal nave model, the traditional autoregressive integrated)Tj
T*
0.0178 Tw
(moving average \(ARIMA\) model, the seasonal ARIMA \(or SARIMA\) model and)Tj
T*
0.0013 Tc
0.1344 Tw
[(the basic structural time-series model \(BSM\) \(see, for example, Goh and Law)74.3(,)]TJ
T*
0.1379 Tw
[(2002; Lim and McAleer)53.6(, 2002; Kulendran and W)74.4(ong, 2005; V)43.1(u, 2006\). The)]TJ
T*
0.2343 Tw
(augmented BSM with explanatory variables, that is, the multivariate causal)Tj
T*
0.0091 Tw
(structure time-series model \(STSM\), has also appeared in some published studies)Tj
T*
0.0011 Tc
0.1613 Tw
[(that compare the forecasting performance of the above mentioned time-series)]TJ
T*
0.0013 Tc
0.22 Tw
[(models \(see, for instance, Gonzlez and Moral, 1995 and 1996; T)95.2(urner and)]TJ
T*
0.0208 Tc
0.3277 Tw
[(W)50.4(itt, 2001\). Other econometric techniques, such as the autoregressive)]TJ
T*
0.0014 Tc
0.2551 Tw
[(distributed)-259.5(lag \(ADL\) model, the error correction model \(ECM\), the vector)]TJ
T*
0.0012 Tc
0.0999 Tw
[(autoregressive \(V)84.7(AR\) model and the time-varying parameter \(TVP\) model, are)]TJ
T*
0.2217 Tw
(often omitted from the forecasting competition of seasonal tourism demand)Tj
T*
0.0011 Tc
0.3111 Tw
(models, with only a few exceptions, such as Gonzlez and Moral \(1995\),)Tj
T*
0.0012 Tc
0.0447 Tw
[(Kulendran and King \(1997\), Kulendran and W)63.8(i)0.1(lson \(2000\) and Kulendran and)]TJ
T*
0.1589 Tw
[(W)51.7(itt \(2001, 2003a,b\), who compared the forecast accuracy of the ECM with)]TJ
T*
0.1722 Tw
[(that of the time-series models; V)53.4(eloce \(2004\) included both the V)84.7(AR model)]TJ
T*
0.1996 Tw
(and ECM in their forecasting competition, while Smeral and Wger \(2005\))Tj
T*
0.001 Tc
0.0754 Tw
(examined the forecasting performance of the seasonal nave, ARIMA, SARIMA)Tj
T*
0.0014 Tc
0.1482 Tw
(and ADL models. It can be seen from these studies that, in seasonal tourism)Tj
T*
0.0011 Tc
0.1454 Tw
(demand forecasting comparisons, only a small number of econometric models)Tj
T*
0.1281 Tw
(\(very often one or two only\) are selected in the forecasting competition and a)Tj
T*
0.0014 Tc
0.1837 Tw
(predominant focus is often given to the time-series models.)Tj
1.0435 -1.0435 TD
0.0042 Tc
0.328 Tw
(Meanwhile, some efforts have been made in evaluating the forecasting)Tj
-1.0435 -1.0435 TD
0.0012 Tc
0.1001 Tw
[(performance)-259.7(of various econometric models as far as annual tourism demand is)]TJ
T*
0.05 Tw
(concerned, such as Song )Tj
/F1 1 Tf
9.6417 0 TD
0.0026 Tc
0.0478 Tw
(et al)Tj
/F2 1 Tf
1.5861 0 TD
0.0015 Tc
0.051 Tw
[( \(2003\) and Li )]TJ
/F1 1 Tf
5.9478 0 TD
0.0026 Tc
0.0478 Tw
(et al)Tj
/F2 1 Tf
1.5861 0 TD
0.0014 Tc
0.0511 Tw
[( \(2006\). However)53.6(, these studies)]TJ
-18.7617 -1.0435 TD
0.0009 Tc
0.1822 Tw
(provided no empirical evidence on the ability of the econometric approaches)Tj
T*
0.0012 Tc
0.1478 Tw
(to deal with seasonality in tourism demand forecasting. There are only a few)Tj
T*
0.0338 Tw
[(exceptions, such as Kulendran and W)53.4(itt \(2001\) and W)74.2(ong )]TJ
/F1 1 Tf
23.0296 0 TD
0.0006 Tc
0.0348 Tw
(et al)Tj
/F2 1 Tf
1.5652 0 TD
0.002 Tc
[( \(2007\). In these)]TJ
-24.5948 -1.0435 TD
0.0012 Tc
0.0491 Tw
[(comparisons, no more than five forecasting models were included. In particular)53.4(,)]TJ
T*
0.0011 Tc
0.1743 Tw
(one of the most recently developed econometric models the TVP model )Tj
T*
0.0013 Tc
0.1332 Tw
(has not been considered. The TVP model has shown its superior performance)Tj
T*
0.1304 Tw
(over the other econometric and time-series models in annual tourism demand)Tj
T*
0.0941 Tw
[(forecasting, especially in the short run \(see Riddington, 1999; Song and W)53.5(itt,)]TJ
T*
0.0016 Tc
0.1396 Tw
[(2000; Song and W)74.6(ong, 2003; Song )]TJ
/F1 1 Tf
14.6713 0 TD
0.0006 Tc
0.1391 Tw
(et al)Tj
/F2 1 Tf
1.68 0 TD
0.0016 Tc
0.1406 Tw
(, 2003; Li )Tj
/F1 1 Tf
4.4661 0 TD
0.0006 Tc
0.1391 Tw
(et al)Tj
/F2 1 Tf
1.68 0 TD
0.0017 Tc
0.1407 Tw
[(, 2006\). However)53.9(, its)]TJ
-22.4974 -1.0435 TD
0.0012 Tc
0.2222 Tw
(performance in seasonal tourism demand forecasting in comparison to other)Tj
T*
0.001 Tc
0.0538 Tw
(econometric and time-series models has not been examined. An overview of the)Tj
T*
0.0205 Tc
0.328 Tw
(tourism forecasting literature shows that there has been no systematic)Tj
T*
0.0012 Tc
0.027 Tw
[(comparison)-259.7(across all of the above models regarding their abilities in forecasting)]TJ
T*
0.1891 Tw
(seasonal tourism demand.)Tj
1.0435 -1.0435 TD
0.3204 Tw
(Seasonality is an important feature of tourism demand time series and)Tj
ET
endstream
endobj
13 0 obj
<<
/ProcSet [ /PDF /Text ]
/Font << /F1 6 0 R /F2 7 0 R >>
/ExtGState << /GS2 9 0 R >>
>>
endobj
14 0 obj
<<
/Type /Page
/Parent 10 0 R
/Resources 16 0 R
/Contents 15 0 R
>>
endobj
15 0 obj
<< /Length 10063 >>
stream
BT
/F2 1 Tf
9 0 0 9 400.2 655.135 Tm
0 0 0 rg
/GS2 gs
0.01 Tc
0 Tw
(695)Tj
/F1 1 Tf
-25.64 0 TD
0.0083 Tc
0.1666 Tw
(Seasonality and tourism demand models)Tj
/F2 1 Tf
11.5 0 0 11.5 56.64 635.455 Tm
0.0012 Tc
0.1599 Tw
(requires careful treatment in modelling and forecasting seasonal \(quarterly or)Tj
0 -1.0435 TD
0.0157 Tw
(monthly\) tourism demand \(Kulendran and King, 1997\). Kim and Moosa \(2001,)Tj
T*
0.0011 Tc
0.0749 Tw
(p 382\) noted that, no conclusive evidence was found as to whether one should)Tj
T*
0.0014 Tc
0.3227 Tw
(treat seasonality as stochastic or deterministic, though the assumption of)Tj
T*
0.0012 Tc
0.0877 Tw
(stochastic seasonality has been popular in recent studies \(for example, Lim and)Tj
T*
0.0797 Tw
[(McAleer)42.9(, 2002\). Empirical evidence in both tourism \(for example, Kim, 1999\))]TJ
T*
0.0011 Tc
0.0137 Tw
(and general economic literatures \(for example, Osborn )Tj
/F1 1 Tf
21.1617 0 TD
0.0023 Tc
0.0174 Tw
[(et al)]TJ
/F2 1 Tf
1.5443 0 TD
0.0014 Tc
0.0145 Tw
(, 1999\) has suggested)Tj
-22.7061 -1.0435 TD
0.0019 Tc
0.328 Tw
(that deterministic seasonality may be more appropriate in modelling and)Tj
T*
0.0014 Tc
0.1114 Tw
(forecasting the seasonal time series. This study will examine this issue further)Tj
T*
0.0012 Tc
0.1837 Tw
(in the tourism demand context.)Tj
1.0435 -1.0435 TD
0.0014 Tc
0.192 Tw
(The primary objective of this study is therefore to bridge the gap in the)Tj
-1.0435 -1.0435 TD
0.001 Tc
0.0591 Tw
(literature by conducting a comprehensive comparison of the performance of the)Tj
T*
0.0012 Tc
0.2152 Tw
(econometric and time-series models in forecasting seasonal tourism demand.)Tj
T*
0.0011 Tc
0.1695 Tw
(Nine econometric and time-series forecasting models will be included in the)Tj
T*
0.0012 Tc
0.1435 Tw
(forecasting comparison. Particular attention is paid to the performance of the)Tj
T*
0.0011 Tc
0.1419 Tw
(modern econometric models such as the TVP model in dealing with seasonal)Tj
T*
0.0014 Tc
0.187 Tw
(tourism demand.)Tj
/F3 1 Tf
12 0 0 12 100.2 407.335 Tm
0 Tc
0.1808 Tw
[(T)87(echniques of forecasting seasonal tourism demand)]TJ
/F1 1 Tf
4.46 -2.11 TD
0.0011 Tc
0.1736 Tw
(Seasonal time-series forecasting models)Tj
/F2 1 Tf
11.5 0 0 11.5 56.64 363.415 Tm
0.2782 Tw
(Three univariate time-series models are often included in seasonal demand)Tj
0 -1.0435 TD
0.0587 Tw
(forecasting comparisons: the seasonal nave model, the SARIMA model and the)Tj
T*
0.0004 Tc
0 Tw
(BSM.)Tj
/F1 1 Tf
0 -2.0765 TD
0.0015 Tc
0.226 Tw
(Seasonal nave model.)Tj
/F2 1 Tf
8.4522 0 TD
0.0011 Tc
0.2282 Tw
(The seasonal nave model is normally used to generate)Tj
-8.4522 -1.0435 TD
0.0755 Tw
(baseline forecasts. As far as quarterly data are concerned, the forecast generated)Tj
T*
0.0685 Tw
(by the seasonal nave model for the period )Tj
/F1 1 Tf
16.9357 0 TD
0 Tc
0 Tw
(t)Tj
/F2 1 Tf
0.2817 0 TD
0.0014 Tc
0.0686 Tw
[( + 4 is equal to the value of period)]TJ
/F1 1 Tf
-17.2174 -1.0435 TD
0 Tc
0 Tw
(t)Tj
/F2 1 Tf
0.2817 0 TD
0.0026 Tc
0.1536 Tw
(, that is, )Tj
6.7 0 0 6.7 104.76 285.895 Tm
0 Tc
0 Tw
(^)Tj
/F1 1 Tf
11.5 0 0 11.5 102.84 279.535 Tm
(F)Tj
6.7 0 0 6.7 109.2 276.895 Tm
(t)Tj
/F2 1 Tf
0.2687 0 TD
-0.0075 Tc
(+4)Tj
11.5 0 0 11.5 118.44 279.535 Tm
0 Tc
0.1543 Tw
[( = )]TJ
/F1 1 Tf
1.4087 0 TD
0 Tw
(F)Tj
6.7 0 0 6.7 141.12 276.895 Tm
(t)Tj
/F2 1 Tf
11.5 0 0 11.5 142.92 279.535 Tm
0.0011 Tc
0.1548 Tw
(. For example, forecasts of one and two quarters ahead are)Tj
-7.5026 -1.0435 TD
0.0013 Tc
0.0894 Tw
(obtained by using the values of the corresponding quarter in the previous year)Tj
T*
0.0009 Tc
0.1847 Tw
[(\(see Kulendran and W)42.6(itt, 2001\).)]TJ
/F1 1 Tf
0 -2.0765 TD
0.0138 Tc
0.3261 Tw
(SARIMA model.)Tj
/F2 1 Tf
7.0017 0 TD
0.013 Tc
0.3279 Tw
(As tourism demand series measured at regular calendar)Tj
-7.0017 -1.0435 TD
0.001 Tc
0.2415 Tw
[(intervals)-259.9(in a year may exhibit periodic behaviour)53.2(, the general BoxJenkins)]TJ
T*
0.0008 Tc
0.0163 Tw
(model with seasonal difference \()Tj
/F1 1 Tf
12.3235 0 TD
0 Tc
0 Tw
(D)Tj
/F2 1 Tf
0.72 0 TD
0.0013 Tc
0.0163 Tw
(\), seasonal autoregressive term \()Tj
/F1 1 Tf
12.1878 0 TD
0 Tc
0 Tw
(P)Tj
/F2 1 Tf
0.5635 0 TD
0.001 Tc
0.0174 Tw
(\) and seasonal)Tj
-25.7948 -1.0435 TD
0.1464 Tw
(moving average term \()Tj
/F1 1 Tf
9.1096 0 TD
0 Tc
0 Tw
(Q)Tj
/F2 1 Tf
0.6052 0 TD
0.0012 Tc
0.1464 Tw
(\), can be represented by ARIMA \()Tj
/F1 1 Tf
13.8887 0 TD
0 Tc
0 Tw
(p)Tj
/F2 1 Tf
0.3861 0 TD
-0.0009 Tc
(, )Tj
/F1 1 Tf
0.6574 0 TD
0 Tc
(d)Tj
/F2 1 Tf
0.4383 0 TD
0.0035 Tc
(, )Tj
/F1 1 Tf
0.647 0 TD
0 Tc
(q)Tj
/F2 1 Tf
0.4487 0 TD
0.0031 Tc
0.1478 Tw
(\) \()Tj
/F1 1 Tf
0.96 0 TD
0 Tc
0 Tw
(P)Tj
/F2 1 Tf
0.553 0 TD
-0.0009 Tc
(, )Tj
/F1 1 Tf
0.6574 0 TD
0 Tc
(D)Tj
/F2 1 Tf
0.72 0 TD
-0.0009 Tc
(, )Tj
/F1 1 Tf
0.647 0 TD
0 Tc
(Q)Tj
/F2 1 Tf
0.6157 0 TD
(\))Tj
6.7 0 0 6.7 408.72 193.015 Tm
(s)Tj
11.5 0 0 11.5 410.88 195.655 Tm
(:)Tj
/F4 1 Tf
10 0 2.1 10 68.64 175.135 Tm
()Tj
/F1 1 Tf
6.7 0 0 6.7 73.92 172.495 Tm
(p)Tj
/F2 1 Tf
11.5 0 0 11.5 76.56 175.135 Tm
(\()Tj
/F1 1 Tf
0.2817 0 TD
(B)Tj
/F2 1 Tf
0.6052 0 TD
(\))Tj
/F4 1 Tf
10 0 2.1 10 90 175.135 Tm
()Tj
/F1 1 Tf
6.7 0 0 6.7 97.68 172.495 Tm
(P)Tj
/F2 1 Tf
11.5 0 0 11.5 101.4 175.135 Tm
(\()Tj
/F1 1 Tf
0.2817 0 TD
(B)Tj
6.7 0 0 6.7 111.72 178.975 Tm
(s)Tj
/F2 1 Tf
11.5 0 0 11.5 113.52 175.135 Tm
(\))Tj
/F4 1 Tf
10 0 0 10 116.76 175.135 Tm
()Tj
/F1 1 Tf
6.7 0 0 6.7 123.96 178.975 Tm
(d)Tj
/F4 1 Tf
10 0 0 10 126.84 175.135 Tm
()Tj
/F1 1 Tf
6.7 0 0 6.7 134.04 178.975 Tm
(D)Tj
0.2328 -0.9672 TD
(s)Tj
11.5 0 0 11.5 139.08 175.135 Tm
(Z)Tj
6.7 0 0 6.7 146.04 172.495 Tm
(t)Tj
/F2 1 Tf
11.5 0 0 11.5 147.96 175.135 Tm
0.0022 Tc
0.1782 Tw
[( = )]TJ
/F4 1 Tf
10 0 2.1 10 164.76 175.135 Tm
0 Tc
0 Tw
()Tj
/F1 1 Tf
6.7 0 0 6.7 170.04 172.495 Tm
(q)Tj
/F2 1 Tf
11.5 0 0 11.5 173.04 175.135 Tm
(\()Tj
/F1 1 Tf
0.2713 0 TD
(B)Tj
/F2 1 Tf
0.6157 0 TD
(\))Tj
/F4 1 Tf
10 0 0 10 186.48 175.135 Tm
()Tj
/F1 1 Tf
6.7 0 0 6.7 193.92 172.495 Tm
(Q)Tj
/F2 1 Tf
11.5 0 0 11.5 198 175.135 Tm
(\()Tj
/F1 1 Tf
0.2817 0 TD
(B)Tj
6.7 0 0 6.7 208.2 178.975 Tm
(s)Tj
/F2 1 Tf
11.5 0 0 11.5 210.12 175.135 Tm
(\))Tj
/F4 1 Tf
10 0 2.1 10 213.24 175.135 Tm
()Tj
/F1 1 Tf
6.7 0 0 6.7 217.8 172.495 Tm
(t)Tj
/F2 1 Tf
11.5 0 0 11.5 401.64 175.135 Tm
-0.001 Tc
(\(1\))Tj
-30 -1.68 TD
0.0008 Tc
(where )Tj
/F1 1 Tf
2.6713 0 TD
0 Tc
(Z)Tj
6.7 0 0 6.7 94.44 153.175 Tm
(t)Tj
/F2 1 Tf
11.5 0 0 11.5 96.24 155.815 Tm
0.0014 Tc
0.1397 Tw
[( is a stationary data point at time )]TJ
/F1 1 Tf
14.1183 0 TD
0 Tc
0 Tw
(t)Tj
/F2 1 Tf
0.2817 0 TD
0.0078 Tc
(, )Tj
/F1 1 Tf
0.6365 0 TD
0 Tc
(B)Tj
/F2 1 Tf
0.6157 0 TD
0.0014 Tc
0.1392 Tw
[( is the backshift operator)53.6(,)-6.4( )]TJ
/F1 1 Tf
10.6748 0 TD
0 Tc
0 Tw
(s)Tj
/F2 1 Tf
0.2817 0 TD
0.0039 Tc
0.1434 Tw
[( is)]TJ
-30.0522 -1.0435 TD
0.0011 Tc
0.0551 Tw
[(the seasonal periodicity)84.6(,)-7.3( )]TJ
/F4 1 Tf
10 0 2.1 10 166.44 143.815 Tm
0 Tc
0 Tw
()Tj
/F1 1 Tf
6.7 0 0 6.7 171 141.175 Tm
(t)Tj
/F2 1 Tf
11.5 0 0 11.5 172.8 143.815 Tm
0.0014 Tc
0.0565 Tw
[( is the disturbance at time )]TJ
/F1 1 Tf
10.6539 0 TD
0 Tc
0 Tw
(t)Tj
/F2 1 Tf
0.2817 0 TD
0.007 Tc
(, )Tj
/F4 1 Tf
10 0 2.1 10 305.04 143.815 Tm
0 Tc
()Tj
/F1 1 Tf
6.7 0 0 6.7 310.2 141.175 Tm
(p)Tj
/F2 1 Tf
11.5 0 0 11.5 312.84 143.815 Tm
(\()Tj
/F1 1 Tf
0.2817 0 TD
(B)Tj
/F2 1 Tf
0.6052 0 TD
0.0013 Tc
0.0565 Tw
(\) is the non-seasonal)Tj
-23.1652 -1.0435 TD
0.0005 Tc
0.1087 Tw
[(AR operator)42.2(,)-6.5( )]TJ
/F4 1 Tf
10 0 2.1 10 119.04 131.815 Tm
0 Tc
0 Tw
()Tj
/F1 1 Tf
6.7 0 0 6.7 126.84 129.175 Tm
(P)Tj
/F2 1 Tf
11.5 0 0 11.5 130.56 131.815 Tm
(\()Tj
/F1 1 Tf
0.2817 0 TD
(B)Tj
6.7 0 0 6.7 140.76 135.655 Tm
(s)Tj
/F2 1 Tf
11.5 0 0 11.5 142.68 131.815 Tm
0.0015 Tc
0.1098 Tw
[(\) is the seasonal AR operator)53.7(, )]TJ
/F4 1 Tf
10 0 2.1 10 281.4 131.815 Tm
0 Tc
0 Tw
()Tj
/F1 1 Tf
6.7 0 0 6.7 286.56 129.175 Tm
(q)Tj
/F2 1 Tf
11.5 0 0 11.5 289.56 131.815 Tm
(\()Tj
/F1 1 Tf
0.2817 0 TD
(B)Tj
/F2 1 Tf
0.6052 0 TD
0.0012 Tc
0.1108 Tw
(\) is the non-seasonal MA)Tj
-21.1409 -1.0435 TD
0.001 Tc
0 Tw
[(operator)53.2(, )]TJ
/F4 1 Tf
10 0 0 10 99 119.815 Tm
0 Tc
()Tj
/F1 1 Tf
6.7 0 0 6.7 106.44 117.175 Tm
(Q)Tj
/F2 1 Tf
11.5 0 0 11.5 110.52 119.815 Tm
(\()Tj
/F1 1 Tf
0.2713 0 TD
(B)Tj
6.7 0 0 6.7 120.72 123.655 Tm
(s)Tj
/F2 1 Tf
11.5 0 0 11.5 122.64 119.815 Tm
0.0012 Tc
0.0558 Tw
[(\) is the seasonal MA operator)53.4(,)-6.5( )]TJ
/F4 1 Tf
10 0 0 10 259.44 119.815 Tm
0 Tc
0 Tw
()Tj
/F1 1 Tf
6.7 0 0 6.7 266.64 123.655 Tm
(d)Tj
/F2 1 Tf
11.5 0 0 11.5 269.64 119.815 Tm
0.0012 Tc
0.0566 Tw
[( is the non-seasonal differencing)]TJ
-18.5217 -1.0435 TD
0.0011 Tc
0.1848 Tw
(operator and )Tj
/F4 1 Tf
10 0 0 10 119.28 107.815 Tm
0 Tc
0 Tw
()Tj
/F1 1 Tf
6.7 0 0 6.7 126.48 111.655 Tm
(D)Tj
0.2149 -0.9672 TD
(s)Tj
/F2 1 Tf
11.5 0 0 11.5 129.84 107.815 Tm
0.0012 Tc
0.1861 Tw
[( is the seasonal differencing operator)63.8(.)]TJ
-5.3217 -1.0435 TD
0.1106 Tw
(Identification is a critical step in estimating an ARIMA \()Tj
/F1 1 Tf
22.9878 0 TD
0 Tc
0 Tw
(p)Tj
/F2 1 Tf
0.3861 0 TD
0.0026 Tc
(, )Tj
/F1 1 Tf
0.6261 0 TD
0 Tc
(d)Tj
/F2 1 Tf
0.4383 0 TD
0.007 Tc
(, )Tj
/F1 1 Tf
0.6157 0 TD
0 Tc
(q)Tj
/F2 1 Tf
0.4383 0 TD
0.0048 Tc
0.1131 Tw
(\) \()Tj
/F1 1 Tf
0.9287 0 TD
0 Tc
0 Tw
(P)Tj
/F2 1 Tf
0.553 0 TD
0.007 Tc
(, )Tj
/F1 1 Tf
0.6157 0 TD
0 Tc
(D)Tj
/F2 1 Tf
0.72 0 TD
0.0026 Tc
(, )Tj
/F1 1 Tf
0.6261 0 TD
0 Tc
(Q)Tj
/F2 1 Tf
0.6052 0 TD
(\))Tj
6.7 0 0 6.7 411.6 93.175 Tm
(s)Tj
11.5 0 0 11.5 56.64 83.815 Tm
0.001 Tc
0.1131 Tw
(model, where )Tj
/F1 1 Tf
5.6661 0 TD
0 Tc
0 Tw
(p)Tj
/F2 1 Tf
0.3861 0 TD
0.0012 Tc
0.1154 Tw
[( is the AR order)53.4(, which indicates the number of parameters of)]TJ
/F4 1 Tf
10 0 2.1 10 56.64 71.815 Tm
0 Tc
0 Tw
()Tj
/F2 1 Tf
11.5 0 0 11.5 61.92 71.815 Tm
0.007 Tc
(, )Tj
/F1 1 Tf
0.6157 0 TD
0 Tc
(d)Tj
/F2 1 Tf
0.4383 0 TD
0.0014 Tc
0.109 Tw
[( is the number of times that the series needs to be differenced in order to)]TJ
-1.513 -1.0435 TD
0.1487 Tw
(achieve a stationary series Z, )Tj
/F1 1 Tf
11.8957 0 TD
0 Tc
0 Tw
(q)Tj
/F2 1 Tf
0.4487 0 TD
0.0013 Tc
0.1489 Tw
[( is the MA order)53.5(, which indicates the number)]TJ
ET
endstream
endobj
16 0 obj
<<
/ProcSet [ /PDF /Text ]
/Font << /F1 6 0 R /F2 7 0 R /F3 8 0 R /F4 17 0 R >>
/ExtGState << /GS2 9 0 R >>
>>
endobj
17 0 obj
<<
/Type /Font
/Subtype /Type1
/FirstChar 1
/LastChar 181
/Widths [ 521 763 713 521 741 439 411 795 549 549 549 250 250 250 250 250 250
250 250 250 250 250 250 250 250 250 250 250 250 250 250 250 250
250 250 250 250 250 250 250 250 250 250 250 250 250 250 250 250
250 250 250 250 250 250 250 250 250 250 250 250 250 250 250 250
250 250 250 250 250 250 250 250 250 250 250 250 250 250 250 250
250 250 250 250 250 250 250 250 250 250 250 250 250 250 250 250
250 250 250 250 250 250 250 250 250 250 250 250 250 250 250 250
250 250 250 250 250 250 250 250 250 250 250 250 250 250 250 250
250 250 250 250 250 250 250 250 250 250 250 250 250 250 250 250
250 250 250 250 250 250 250 250 250 250 250 250 250 250 250 250
250 250 250 250 250 250 250 250 250 250 250 250 250 250 250 250
250 250 250 576 ]
/Encoding 71 0 R
/BaseFont /HPCAKJ+Symbol
/FontDescriptor 67 0 R
/ToUnicode 72 0 R
>>
endobj
18 0 obj
<<
/Type /Page
/Parent 10 0 R
/Resources 20 0 R
/Contents 19 0 R
>>
endobj
19 0 obj
<< /Length 8315 >>
stream
BT
/F2 1 Tf
9.5 0 0 9.2 203.38 654.895 Tm
0 0 0 rg
/GS2 gs
0 Tc
0 Tw
(T)Tj
8.1 0 0 7.8 209.26 654.895 Tm
0.0025 Tc
(OURISM)Tj
9.5 0 0 9.2 239.5 654.895 Tm
0.1666 Tc
( E)Tj
8.1 0 0 7.8 248.86 654.895 Tm
0.0018 Tc
(CONOMICS)Tj
9 0 0 9 68.02 655.135 Tm
0.01 Tc
(696)Tj
11.5 0 0 11.5 68.02 635.455 Tm
0.0013 Tc
0.287 Tw
(of parameters of )Tj
/F4 1 Tf
10 0 2.1 10 151.3 635.455 Tm
0 Tc
0 Tw
()Tj
/F2 1 Tf
11.5 0 0 11.5 156.58 635.455 Tm
0.0061 Tc
(, )Tj
/F1 1 Tf
0.793 0 TD
0 Tc
(P)Tj
/F2 1 Tf
0.553 0 TD
0.0012 Tc
0.2875 Tw
[( is the seasonal AR order indicating the number of)]TJ
-9.047 -1.0435 TD
0.0057 Tc
0.3261 Tw
(parameters of )Tj
/F4 1 Tf
10 0 2.1 10 138.46 623.455 Tm
0 Tc
0 Tw
()Tj
/F2 1 Tf
11.5 0 0 11.5 146.26 623.455 Tm
0.0044 Tc
(, )Tj
/F1 1 Tf
0.8452 0 TD
0 Tc
(Q)Tj
/F2 1 Tf
0.6157 0 TD
0.0066 Tc
0.328 Tw
[( is the seasonal MA order indicating the number of)]TJ
-8.2643 -1.0435 TD
0.0016 Tc
0 Tw
[(parameters)-248.8(of )]TJ
/F4 1 Tf
10 0 0 10 132.82 611.455 Tm
0 Tc
()Tj
/F2 1 Tf
11.5 0 0 11.5 140.26 611.455 Tm
0.0025 Tc
0.2348 Tw
[( and )]TJ
/F1 1 Tf
2.3687 0 TD
0 Tc
0 Tw
(D)Tj
/F2 1 Tf
0.72 0 TD
0.0015 Tc
0.2348 Tw
[( is the number of times that the series needs to be)]TJ
-9.3704 -1.0435 TD
0.0012 Tc
0.1304 Tw
[(seasonally differenced to arrive at a seasonally stationary series. However)53.4(, it is)]TJ
T*
0.0221 Tw
(important that the data are properly processed before the estimation takes place,)Tj
T*
0.0014 Tc
0.0418 Tw
[(as the ARIMA model requires the time series to be stationary)84.9(. As most seasonal)]TJ
T*
0.0012 Tc
0.0265 Tw
(time series exhibit increasing trend and/or seasonal variations, both seasonal and)Tj
T*
0.0011 Tc
0.187 Tw
(non-seasonal differencing are often used to achieve a stationary time series.)Tj
/F1 1 Tf
0 -2.0765 TD
0.0039 Tc
0 Tw
(BSM.)Tj
/F2 1 Tf
2.6504 0 TD
0.0013 Tc
0.3273 Tw
[(A BSM is formulated by decomposing the time series into several)]TJ
-2.6504 -1.0435 TD
0.1884 Tw
(unobservable components as follows:)Tj
/F1 1 Tf
1.0435 -1.5652 TD
0 Tc
0 Tw
(Y)Tj
6.7 0 0 6.7 87.7 494.935 Tm
(t)Tj
/F2 1 Tf
11.5 0 0 11.5 89.5 497.575 Tm
0.0022 Tc
0.1782 Tw
[( = )]TJ
/F4 1 Tf
10 0 2.1 10 106.42 497.575 Tm
0 Tc
0 Tw
()Tj
/F1 1 Tf
6.7 0 0 6.7 112.18 494.935 Tm
(t)Tj
/F2 1 Tf
11.5 0 0 11.5 114.1 497.575 Tm
0.1804 Tw
[( + )]TJ
/F4 1 Tf
10 0 2.1 10 130.9 497.575 Tm
0 Tw
()Tj
/F1 1 Tf
6.7 0 0 6.7 135.1 494.935 Tm
(t)Tj
/F2 1 Tf
11.5 0 0 11.5 136.9 497.575 Tm
0.1804 Tw
[( + )]TJ
/F4 1 Tf
10 0 2.1 10 153.82 497.575 Tm
0 Tw
()Tj
/F1 1 Tf
6.7 0 0 6.7 161.86 494.935 Tm
(t)Tj
/F2 1 Tf
11.5 0 0 11.5 163.66 497.575 Tm
0.1804 Tw
[( + )]TJ
/F4 1 Tf
10 0 2.1 10 180.58 497.575 Tm
0 Tw
()Tj
/F1 1 Tf
6.7 0 0 6.7 185.02 494.935 Tm
(t)Tj
/F2 1 Tf
11.5 0 0 11.5 413.02 497.575 Tm
-0.001 Tc
(\(2\))Tj
-30 -1.5652 TD
0.0001 Tc
(where )Tj
/F1 1 Tf
2.7548 0 TD
0 Tc
(Y)Tj
6.7 0 0 6.7 107.38 476.935 Tm
(t)Tj
/F2 1 Tf
11.5 0 0 11.5 109.3 479.575 Tm
0.0015 Tc
0.2298 Tw
[( is the actual tourism demand and )]TJ
/F4 1 Tf
10 0 2.1 10 281.26 479.575 Tm
0 Tc
0 Tw
()Tj
/F1 1 Tf
6.7 0 0 6.7 287.14 476.935 Tm
(t)Tj
/F2 1 Tf
11.5 0 0 11.5 288.94 479.575 Tm
(, )Tj
/F4 1 Tf
10 0 2.1 10 297.46 479.575 Tm
()Tj
/F1 1 Tf
6.7 0 0 6.7 301.54 476.935 Tm
(t)Tj
/F2 1 Tf
11.5 0 0 11.5 303.46 479.575 Tm
(, )Tj
/F4 1 Tf
10 0 2.1 10 311.86 479.575 Tm
()Tj
/F1 1 Tf
6.7 0 0 6.7 319.9 476.935 Tm
(t)Tj
/F2 1 Tf
11.5 0 0 11.5 321.7 479.575 Tm
0.0006 Tc
0.2282 Tw
[( and )]TJ
/F4 1 Tf
10 0 2.1 10 348.7 479.575 Tm
0 Tc
0 Tw
()Tj
/F1 1 Tf
6.7 0 0 6.7 353.26 476.935 Tm
(t)Tj
/F2 1 Tf
11.5 0 0 11.5 355.06 479.575 Tm
0.0018 Tc
0.2304 Tw
[( are the trend,)]TJ
-24.96 -1.0435 TD
0.001 Tc
0.0324 Tw
[(seasonal, cyclical and irregular components, respectively)84.5(. Each component of the)]TJ
T*
0.0012 Tc
0.1369 Tw
[(series can be modelled in several ways \(see Gonzlez and Moral, 1996\). W)53.4(ith)]TJ
T*
0.0135 Tc
0.3279 Tw
(respect to the seasonal component, the trigonometric form is the most)Tj
T*
0.0012 Tc
0.0918 Tw
[(commonly)-259.7(used in the literature and will be applied in this empirical study as)]TJ
T*
0.0014 Tc
0.1618 Tw
(well. The irregular component represents the transitory variations in tourism)Tj
T*
0.001 Tc
0.257 Tw
(demand which cannot be explained by the other components. A particular)Tj
T*
0.0013 Tc
0.1455 Tw
(feature of the BSM is that stochastic movements are permitted. For example,)Tj
T*
0.001 Tc
0.1831 Tw
(a slowly changing seasonal component may indicate seasonality is stochastic.)Tj
T*
0.0013 Tc
0.0124 Tw
[(More details of the BSM specifications can be found in Harvey and T)84.8(odd \(1983\).)]TJ
/F1 1 Tf
12 0 0 12 206.74 341.815 Tm
0.0007 Tc
0.1666 Tw
(Econometric models)Tj
/F2 1 Tf
11.5 0 0 11.5 68.02 323.215 Tm
0.001 Tc
0.2153 Tw
(Five econometric models have been commonly used in the tourism demand)Tj
T*
0.0012 Tc
0.1971 Tw
(forecasting literature and they are the reduced autoregressive distributed lag)Tj
T*
0.0234 Tw
[(\(RE-ADL\) model, the W)53.4(ickensBreusch error correction model \(WB-ECM\), the)]TJ
T*
0.2065 Tw
[(Johansen maximum likelihood error correction model \(JML-ECM\), the V)84.7(A)10(R)]TJ
T*
0.1065 Tw
(model and the TVP model. The specifications of these models are available in)Tj
T*
0.002 Tc
0 Tw
(Song )Tj
/F1 1 Tf
2.233 0 TD
0.0006 Tc
0.087 Tw
(et al)Tj
/F2 1 Tf
1.6174 0 TD
0.0013 Tc
0.0893 Tw
[( \(2003\). In addition, the STSM has appeared recently in the seasonal)]TJ
-3.8504 -1.0435 TD
0.0261 Tw
(tourism demand forecasting literature. This model has shown relatively superior)Tj
T*
0.001 Tc
0.115 Tw
(forecasting performance, especially when it is compared with ECMs \(Gonzlez)Tj
T*
0 Tw
(and Moral, 1995\). Three different techniques are applied to deal with seasonality)Tj
T*
0.0015 Tc
0.1439 Tw
(in this study: the deterministic seasonal dummies, the seasonal unit root test)Tj
T*
0.001 Tc
0.1884 Tw
(and the unobservable components.)Tj
/F1 1 Tf
0 -2.0765 TD
0.0016 Tc
0.0282 Tw
(Deterministic seasonal dummies.)Tj
/F2 1 Tf
11.687 0 TD
0.0013 Tc
0.0303 Tw
[(For the RE-ADL, V)84.8(AR and TVP models, seasonal)]TJ
-11.687 -1.0435 TD
0.1179 Tw
[(dummies are incorporated into the model specifications to capture seasonality)95.1(.)]TJ
T*
0.0011 Tc
0.2834 Tw
(The process generated by the seasonal dummies is normally called a pure)Tj
T*
0.0013 Tc
0.2548 Tw
(deterministic seasonal process. The parameters of the dummies are used to)Tj
T*
0.0011 Tc
0.15 Tw
(describe the seasonal fluctuations and their effects on the dependent variable.)Tj
T*
0.0081 Tc
0 Tw
[(Normally)91.6(, )]TJ
/F1 1 Tf
4.5496 0 TD
0 Tc
(s)Tj
/F2 1 Tf
0.2817 0 TD
0.0081 Tc
0.326 Tw
(1 \()Tj
/F1 1 Tf
1.8887 0 TD
0 Tc
0 Tw
(s)Tj
/F2 1 Tf
0.2922 0 TD
0.0089 Tc
0.3276 Tw
[( is the number of seasons in a one-year cycle, that is,)]TJ
/F1 1 Tf
-7.0122 -1.0435 TD
0 Tc
0 Tw
(s)Tj
/F2 1 Tf
0.2817 0 TD
0.0012 Tc
( = 4 for quarterly time series and )Tj
/F1 1 Tf
12.9183 0 TD
0 Tc
(s)Tj
/F2 1 Tf
0.2817 0 TD
0.0008 Tc
0.0003 Tw
[( = 12 for monthly data\) seasonal dummies)-239.2(are)]TJ
-13.4817 -1.0435 TD
0.0012 Tc
0.1521 Tw
(included in a forecasting model. Seasonal dummies )Tj
/F1 1 Tf
20.9009 0 TD
0 Tc
0 Tw
(D)Tj
6.7 0 0 6.7 316.66 92.695 Tm
-0.0008 Tc
(it)Tj
/F2 1 Tf
0.5552 0 TD
0 Tc
( )Tj
11.5 0 0 11.5 323.02 95.335 Tm
(\()Tj
/F1 1 Tf
0.2817 0 TD
-0.0019 Tc
(i )Tj
/F2 1 Tf
0.6783 0 TD
0.0026 Tc
0.1521 Tw
[(= 1,2, .)-258.3(.)-248.3(., )]TJ
/F1 1 Tf
4.8522 0 TD
0 Tc
0 Tw
(s)Tj
/F2 1 Tf
0.2713 0 TD
0.0002 Tc
0.1565 Tw
(1\) are)Tj
-28.2574 -1.0435 TD
0.0007 Tc
0.0869 Tw
(defined as )Tj
/F1 1 Tf
4.247 0 TD
0 Tc
0 Tw
(D)Tj
6.7 0 0 6.7 125.14 80.695 Tm
-0.0008 Tc
(it)Tj
/F2 1 Tf
11.5 0 0 11.5 128.86 83.335 Tm
0.0023 Tc
0.0887 Tw
[( = 1 if time )]TJ
/F1 1 Tf
5.1548 0 TD
0 Tc
0 Tw
(t)Tj
/F2 1 Tf
0.2713 0 TD
0.0014 Tc
0.0881 Tw
[( corresponds to season )]TJ
/F1 1 Tf
9.047 0 TD
0 Tc
0 Tw
(s)Tj
/F2 1 Tf
0.2817 0 TD
0.0008 Tc
0.0869 Tw
[( and )]TJ
/F1 1 Tf
2.0661 0 TD
0 Tc
0 Tw
(D)Tj
6.7 0 0 6.7 330.7 80.695 Tm
-0.0008 Tc
(it)Tj
/F2 1 Tf
11.5 0 0 11.5 334.3 83.335 Tm
0.0015 Tc
0.0891 Tw
[( = 0 otherwise. The)]TJ
-23.1548 -1.0435 TD
0.0013 Tc
0.1448 Tw
(use of seasonal dummies implies that the seasonal pattern in a time series )Tj
/F1 1 Tf
30.2296 0 TD
0 Tc
0 Tw
(Y)Tj
6.7 0 0 6.7 423.34 68.695 Tm
(t)Tj
/F2 1 Tf
11.5 0 0 11.5 68.02 59.335 Tm
0.0014 Tc
0.1839 Tw
(\(for example, tourism demand\) is constant over time.)Tj
ET
endstream
endobj
20 0 obj
<<
/ProcSet [ /PDF /Text ]
/Font << /F1 6 0 R /F2 7 0 R /F4 17 0 R >>
/ExtGState << /GS2 9 0 R >>
>>
endobj
21 0 obj
<<
/Type /Page
/Parent 10 0 R
/Resources 23 0 R
/Contents 22 0 R
>>
endobj
22 0 obj
<< /Length 11884 >>
stream
BT
/F2 1 Tf
9 0 0 9 400.2 655.135 Tm
0 0 0 rg
/GS2 gs
0.01 Tc
0 Tw
(697)Tj
/F1 1 Tf
-25.64 0 TD
0.0083 Tc
0.1666 Tw
(Seasonality and tourism demand models)Tj
/F2 1 Tf
11.5 0 0 11.5 68.64 635.455 Tm
0.0012 Tc
0.2381 Tw
[(However)53.4(, Abeysinghe \(1994\) shows that using the deterministic seasonal)]TJ
-1.0435 -1.0435 TD
0.0014 Tc
0.2924 Tw
(dummies in removing seasonality in the time series is likely to result in)Tj
T*
0.1539 Tw
(spurious regressions, as deterministic dummy variables do not reflect the dy-)Tj
T*
0.0012 Tc
0.1243 Tw
(namic nature of the seasonality inherent in the actual tourism demand. If the)Tj
T*
0.001 Tc
0.3009 Tw
(seasonal effects changed gradually over time, this approach would lead to)Tj
T*
0.0913 Tw
(misspecifications of the dynamic structure of the model because the estimated)Tj
T*
0.1022 Tw
(coefficients on the dummies reflect initial conditions plus the accumulation of)Tj
T*
0.0014 Tc
0.0923 Tw
(random shocks \(Miron, 1994, p 217\). As a result, the forecasts may be biased)Tj
T*
0.001 Tc
0.1887 Tw
(and can cause inappropriate decision making.)Tj
/F1 1 Tf
0 -2.0765 TD
0.004 Tc
0.3271 Tw
[(T)82.8(esting for seasonal unit roots.)]TJ
/F2 1 Tf
12.3757 0 TD
0.0037 Tc
0.3283 Tw
(Although the patterns of seasonality can be)Tj
-12.3757 -1.0435 TD
0.001 Tc
0.0251 Tw
[(deterministic)-259.9(because of the calendar and weather effects, some fluctuations may)]TJ
T*
0.0013 Tc
0.2281 Tw
(be caused by the behaviour of tourists and are unlikely to be constant. As)Tj
T*
0.1599 Tw
(Franses \(1996, p 299\) noted, non-durable consumption patterns may change)Tj
T*
0.0011 Tc
0.0988 Tw
[(when preferences for certain holiday seasons change .)-249.3(.)-260.2(. sales can depend upon)]TJ
T*
0.0013 Tc
0.117 Tw
[(the state of the economy. Miron \(1994, p 219\) ar)11.7(gues that it does not make)]TJ
T*
0.0011 Tc
0.198 Tw
(sense to examine estimated seasonal dummy coefficients unless seasonal unit)Tj
T*
0.0013 Tc
0.1809 Tw
(roots can be treated as absent. A model with seasonal dummies is likely to)Tj
T*
0.0011 Tc
0.1483 Tw
(misrepresent the seasonality as compared with the seasonal integrated process)Tj
T*
0.0007 Tc
0.0048 Tw
[(proposed by Hylleber)11.1(g )]TJ
/F1 1 Tf
9.0574 0 TD
0.0004 Tc
0.0043 Tw
(et al)Tj
/F2 1 Tf
1.5339 0 TD
0.0016 Tc
( \(1990\). A time-series model with seasonal unit roots)Tj
-10.5913 -1.0435 TD
0.0013 Tc
0.0669 Tw
(is an approximation that allows for changes in the seasonal pattern; that is, the)Tj
T*
0.1251 Tw
(series is integrated at the seasonal frequencies. Such a model is subject to the)Tj
T*
0.0016 Tc
0.15 Tw
(seasonal unit root test, which involves two steps. The first step is to test for)Tj
T*
0.0013 Tc
0.1806 Tw
(seasonal and non-seasonal unit roots in a time series; and the second step is)Tj
T*
0.034 Tw
(to apply different seasonal differencing filters to obtain a stationary series before)Tj
T*
0.0011 Tc
0.1644 Tw
(the normal cointegration procedure can be applied to estimate the model, as)Tj
T*
0.0015 Tc
0.094 Tw
(suggested by the results of the seasonal unit root test. The construction of the)Tj
T*
0.1021 Tw
(WB-ECM and JML-ECM depends on the results of the seasonal unit root test)Tj
T*
0.0011 Tc
0.1859 Tw
(when seasonal data are used in estimating the model.)Tj
1.0435 -1.0435 TD
0.0014 Tc
0.0837 Tw
[(The HEGY test \(Hylleber)11.8(g )]TJ
/F1 1 Tf
11.1339 0 TD
0.0006 Tc
0.087 Tw
(et al)Tj
/F2 1 Tf
1.6278 0 TD
0.0014 Tc
0.0847 Tw
(, 1990\) has been commonly used to test for)Tj
-13.8052 -1.0435 TD
0.0036 Tc
0.3276 Tw
(seasonal and non-seasonal unit roots in a univariate series of a quarterly)Tj
T*
0.0013 Tc
0.1852 Tw
[(frequency)84.8(,)-249.6(and the test is based on the following auxiliary regression:)]TJ
/F1 1 Tf
1.0435 -1.5652 TD
0 Tc
0 Tw
(y)Tj
/F2 1 Tf
6.7 0 0 6.7 73.2 242.935 Tm
(4)Tj
/F1 1 Tf
0.4836 0 TD
(t)Tj
/F2 1 Tf
11.5 0 0 11.5 78.36 245.575 Tm
0.1804 Tw
[( = )]TJ
/F4 1 Tf
10 0 2.1 10 95.28 245.575 Tm
0 Tw
( )Tj
/F2 1 Tf
6.7 0 0 6.7 100.8 242.935 Tm
(1)Tj
/F1 1 Tf
11.5 0 0 11.5 104.16 245.575 Tm
(y)Tj
/F2 1 Tf
6.7 0 0 6.7 108.6 242.935 Tm
(1)Tj
/F1 1 Tf
0.5015 0 TD
(t)Tj
/F2 1 Tf
0.2687 0 TD
0.003 Tc
(1)Tj
11.5 0 0 11.5 120.48 245.575 Tm
0 Tc
0.1804 Tw
[( + )]TJ
/F4 1 Tf
10 0 2.1 10 137.4 245.575 Tm
0 Tw
( )Tj
/F2 1 Tf
6.7 0 0 6.7 142.92 242.935 Tm
(2)Tj
/F1 1 Tf
11.5 0 0 11.5 146.28 245.575 Tm
(y)Tj
/F2 1 Tf
6.7 0 0 6.7 150.72 242.935 Tm
(2)Tj
/F1 1 Tf
0.5015 0 TD
(t)Tj
/F2 1 Tf
0.2866 0 TD
0.003 Tc
(1)Tj
11.5 0 0 11.5 162.6 245.575 Tm
0 Tc
0.1804 Tw
[( + )]TJ
/F4 1 Tf
10 0 2.1 10 179.52 245.575 Tm
0 Tw
( )Tj
/F2 1 Tf
6.7 0 0 6.7 185.04 242.935 Tm
(3)Tj
/F1 1 Tf
11.5 0 0 11.5 188.4 245.575 Tm
(y)Tj
/F2 1 Tf
6.7 0 0 6.7 192.84 242.935 Tm
(3)Tj
/F1 1 Tf
0.5015 0 TD
(t)Tj
/F2 1 Tf
0.2866 0 TD
0.003 Tc
(2)Tj
11.5 0 0 11.5 204.84 245.575 Tm
0 Tc
0.1804 Tw
[( + )]TJ
/F4 1 Tf
10 0 2.1 10 221.64 245.575 Tm
0 Tw
( )Tj
/F2 1 Tf
6.7 0 0 6.7 227.28 242.935 Tm
(4)Tj
/F1 1 Tf
11.5 0 0 11.5 230.64 245.575 Tm
(y)Tj
/F2 1 Tf
6.7 0 0 6.7 235.08 242.935 Tm
(3)Tj
/F1 1 Tf
0.5015 0 TD
(t)Tj
/F2 1 Tf
0.2687 0 TD
0.003 Tc
(1)Tj
11.5 0 0 11.5 246.96 245.575 Tm
0 Tc
0.1804 Tw
[( + )]TJ
/F4 1 Tf
10 0 2.1 10 263.88 245.575 Tm
0 Tw
()Tj
/F1 1 Tf
6.7 0 0 6.7 268.32 242.935 Tm
(t)Tj
/F2 1 Tf
11.5 0 0 11.5 401.64 245.575 Tm
-0.001 Tc
(\(3\))Tj
-30 -1.5652 TD
0.0044 Tc
(where)Tj
/F1 1 Tf
1.0435 -2.0765 TD
0 Tc
(y)Tj
/F2 1 Tf
6.7 0 0 6.7 73.2 201.055 Tm
(4)Tj
/F1 1 Tf
0.4836 0 TD
(t)Tj
/F2 1 Tf
11.5 0 0 11.5 78.36 203.695 Tm
0.002 Tc
0.1782 Tw
[( = \(1)]TJ
/F1 1 Tf
2.7443 0 TD
0 Tc
0 Tw
(B)Tj
/F2 1 Tf
6.7 0 0 6.7 117 207.535 Tm
(4)Tj
11.5 0 0 11.5 120.24 203.695 Tm
(\))Tj
/F1 1 Tf
0.2817 0 TD
(y)Tj
6.7 0 0 6.7 127.92 201.055 Tm
(t)Tj
/F2 1 Tf
11.5 0 0 11.5 129.84 203.695 Tm
0.1804 Tw
[( = )]TJ
/F1 1 Tf
1.4609 0 TD
0 Tw
(y)Tj
6.7 0 0 6.7 151.2 201.055 Tm
(t)Tj
11.5 0 0 11.5 153 203.695 Tm
0.0011 Tc
0.1804 Tw
[( y)]TJ
6.7 0 0 6.7 173.16 201.055 Tm
0 Tc
0 Tw
(t)Tj
/F2 1 Tf
0.2866 0 TD
0.003 Tc
(4)Tj
11.5 0 0 11.5 181.8 203.695 Tm
0.1887 Tc
( ;)Tj
/F1 1 Tf
-9.84 -2.0765 TD
0 Tc
(y)Tj
/F2 1 Tf
6.7 0 0 6.7 73.2 177.175 Tm
(1)Tj
/F1 1 Tf
0.4836 0 TD
(t)Tj
/F2 1 Tf
0.2866 0 TD
0.003 Tc
(1)Tj
11.5 0 0 11.5 85.08 179.815 Tm
0.0007 Tc
0.1783 Tw
[( = \(1+)]TJ
/F1 1 Tf
2.8383 0 TD
0 Tc
0 Tw
(B)Tj
/F2 1 Tf
0.6157 0 TD
(+)Tj
/F1 1 Tf
0.5948 0 TD
(B)Tj
/F2 1 Tf
6.7 0 0 6.7 138.72 183.655 Tm
(2)Tj
11.5 0 0 11.5 142.08 179.815 Tm
(+)Tj
/F1 1 Tf
0.6052 0 TD
(B)Tj
/F2 1 Tf
6.7 0 0 6.7 156 183.655 Tm
(3)Tj
11.5 0 0 11.5 159.36 179.815 Tm
(\))Tj
/F1 1 Tf
0.2817 0 TD
(y)Tj
6.7 0 0 6.7 167.04 177.175 Tm
(t)Tj
/F2 1 Tf
0.2687 0 TD
0.003 Tc
(1)Tj
11.5 0 0 11.5 175.56 179.815 Tm
0 Tc
0.1804 Tw
[( = )]TJ
/F1 1 Tf
1.4713 0 TD
0 Tw
(y)Tj
6.7 0 0 6.7 196.92 177.175 Tm
(t)Tj
/F2 1 Tf
0.2866 0 TD
0.003 Tc
(1)Tj
11.5 0 0 11.5 205.44 179.815 Tm
0.0022 Tc
0.1782 Tw
[( + )]TJ
/F1 1 Tf
1.4713 0 TD
0 Tc
0 Tw
(y)Tj
6.7 0 0 6.7 226.8 177.175 Tm
(t)Tj
/F2 1 Tf
0.2687 0 TD
0.003 Tc
(2)Tj
11.5 0 0 11.5 235.32 179.815 Tm
0 Tc
0.1804 Tw
[( + )]TJ
/F1 1 Tf
1.4713 0 TD
0 Tw
(y)Tj
6.7 0 0 6.7 256.68 177.175 Tm
(t)Tj
/F2 1 Tf
0.2866 0 TD
0.003 Tc
(3)Tj
11.5 0 0 11.5 265.2 179.815 Tm
0.0022 Tc
0.1782 Tw
[( + )]TJ
/F1 1 Tf
1.4713 0 TD
0 Tc
0 Tw
(y)Tj
6.7 0 0 6.7 286.56 177.175 Tm
(t)Tj
/F2 1 Tf
0.2687 0 TD
0.003 Tc
(4)Tj
11.5 0 0 11.5 295.08 179.815 Tm
0.1887 Tc
( ;)Tj
/F1 1 Tf
-19.6904 -2.0765 TD
0 Tc
(y)Tj
/F2 1 Tf
6.7 0 0 6.7 73.2 153.295 Tm
(2)Tj
/F1 1 Tf
0.4836 0 TD
(t)Tj
/F2 1 Tf
0.2866 0 TD
0.003 Tc
(1)Tj
11.5 0 0 11.5 85.08 155.935 Tm
0.0018 Tc
0.1782 Tw
[( = \(1)]TJ
/F1 1 Tf
3.2452 0 TD
0 Tc
0 Tw
(B)Tj
/F2 1 Tf
0.6157 0 TD
(+)Tj
/F1 1 Tf
0.5948 0 TD
(B)Tj
/F2 1 Tf
6.7 0 0 6.7 143.4 159.775 Tm
(2)Tj
11.5 0 0 11.5 146.76 155.935 Tm
()Tj
/F1 1 Tf
0.5009 0 TD
(B)Tj
/F2 1 Tf
6.7 0 0 6.7 159.6 159.775 Tm
(3)Tj
11.5 0 0 11.5 162.84 155.935 Tm
(\))Tj
/F1 1 Tf
0.2817 0 TD
(y)Tj
6.7 0 0 6.7 170.52 153.295 Tm
(t)Tj
/F2 1 Tf
0.2866 0 TD
0.003 Tc
(1)Tj
11.5 0 0 11.5 179.04 155.935 Tm
0.0022 Tc
0.1782 Tw
[( = )]TJ
/F1 1 Tf
1.4713 0 TD
0 Tc
0 Tw
()Tj
/F2 1 Tf
0.5009 0 TD
0.0046 Tc
(\(1)Tj
/F1 1 Tf
0.7722 0 TD
0.0055 Tc
(B)Tj
/F2 1 Tf
1.1165 0 TD
0.001 Tc
(\)\(1+)Tj
/F1 1 Tf
1.6591 0 TD
0 Tc
(B)Tj
/F2 1 Tf
6.7 0 0 6.7 249.6 159.775 Tm
(2)Tj
11.5 0 0 11.5 252.96 155.935 Tm
(\))Tj
/F1 1 Tf
0.2817 0 TD
(y)Tj
6.7 0 0 6.7 260.64 153.295 Tm
(t)Tj
/F2 1 Tf
0.2687 0 TD
0.003 Tc
(1)Tj
11.5 0 0 11.5 269.16 155.935 Tm
0.0022 Tc
0.1782 Tw
[( = )]TJ
/F1 1 Tf
1.4609 0 TD
-0.002 Tc
0 Tw
(y)Tj
6.7 0 0 6.7 296.16 153.295 Tm
0 Tc
(t)Tj
/F2 1 Tf
0.2866 0 TD
0.003 Tc
(1)Tj
11.5 0 0 11.5 304.8 155.935 Tm
0 Tc
0.1804 Tw
[( + )]TJ
/F1 1 Tf
1.4609 0 TD
0 Tw
(y)Tj
6.7 0 0 6.7 326.04 153.295 Tm
(t)Tj
/F2 1 Tf
0.2866 0 TD
0.003 Tc
(2)Tj
11.5 0 0 11.5 334.68 155.935 Tm
0 Tc
0.1783 Tw
[( )]TJ
/F1 1 Tf
1.367 0 TD
0 Tw
(y)Tj
6.7 0 0 6.7 354.84 153.295 Tm
(t)Tj
/F2 1 Tf
0.2687 0 TD
0.003 Tc
(3)Tj
11.5 0 0 11.5 363.36 155.935 Tm
0.0022 Tc
0.1782 Tw
[( + )]TJ
/F1 1 Tf
1.4609 0 TD
0 Tc
0 Tw
(y)Tj
6.7 0 0 6.7 384.6 153.295 Tm
(t)Tj
/F2 1 Tf
0.2866 0 TD
0.003 Tc
(4)Tj
11.5 0 0 11.5 393.24 155.935 Tm
0.1887 Tc
( ;)Tj
/F1 1 Tf
-28.2261 -2.0765 TD
0 Tc
(y)Tj
/F2 1 Tf
6.7 0 0 6.7 73.2 129.415 Tm
(3)Tj
/F1 1 Tf
0.4836 0 TD
(t)Tj
/F2 1 Tf
0.2866 0 TD
0.003 Tc
(2)Tj
11.5 0 0 11.5 85.08 132.055 Tm
0.0018 Tc
0.1782 Tw
[( = \(1)]TJ
/F1 1 Tf
3.2452 0 TD
0 Tc
0 Tw
(B)Tj
/F2 1 Tf
6.7 0 0 6.7 129.48 135.895 Tm
(2)Tj
11.5 0 0 11.5 132.84 132.055 Tm
(\))Tj
/F1 1 Tf
0.2713 0 TD
(y)Tj
6.7 0 0 6.7 140.4 129.415 Tm
(t)Tj
/F2 1 Tf
0.2866 0 TD
0.003 Tc
(2)Tj
11.5 0 0 11.5 149.04 132.055 Tm
0.0018 Tc
0.1782 Tw
[( = \(1)]TJ
/F1 1 Tf
3.2452 0 TD
0 Tc
0 Tw
(B)Tj
/F2 1 Tf
0.6157 0 TD
0.001 Tc
(\)\(1+)Tj
/F1 1 Tf
1.6591 0 TD
0 Tc
(B)Tj
/F2 1 Tf
0.6157 0 TD
(\))Tj
/F1 1 Tf
0.2713 0 TD
(y)Tj
6.7 0 0 6.7 227.16 129.415 Tm
(t)Tj
/F2 1 Tf
0.2866 0 TD
0.003 Tc
(2)Tj
11.5 0 0 11.5 235.8 132.055 Tm
0 Tc
0.1804 Tw
[( = )]TJ
/F1 1 Tf
1.4609 0 TD
-0.002 Tc
0 Tw
(y)Tj
6.7 0 0 6.7 262.8 129.415 Tm
0 Tc
(t)Tj
/F2 1 Tf
0.2866 0 TD
0.003 Tc
(2)Tj
11.5 0 0 11.5 271.44 132.055 Tm
0.0022 Tc
0.1782 Tw
[( + )]TJ
/F1 1 Tf
1.4609 0 TD
0 Tc
0 Tw
(y)Tj
6.7 0 0 6.7 292.68 129.415 Tm
(t)Tj
/F2 1 Tf
0.2866 0 TD
0.003 Tc
(4)Tj
11.5 0 0 11.5 301.2 132.055 Tm
0.1887 Tc
( ;)Tj
/F1 1 Tf
-20.2226 -2.0765 TD
0 Tc
(y)Tj
/F2 1 Tf
6.7 0 0 6.7 73.2 105.535 Tm
(3)Tj
/F1 1 Tf
0.4836 0 TD
(t)Tj
/F2 1 Tf
0.2866 0 TD
0.003 Tc
(1)Tj
11.5 0 0 11.5 85.08 108.175 Tm
0.0018 Tc
0.1782 Tw
[( = \(1)]TJ
/F1 1 Tf
3.2452 0 TD
0 Tc
0 Tw
(B)Tj
/F2 1 Tf
6.7 0 0 6.7 129.48 112.015 Tm
(2)Tj
11.5 0 0 11.5 132.84 108.175 Tm
(\))Tj
/F1 1 Tf
0.2713 0 TD
(y)Tj
6.7 0 0 6.7 140.4 105.535 Tm
(t)Tj
/F2 1 Tf
0.2866 0 TD
0.003 Tc
(1)Tj
11.5 0 0 11.5 149.04 108.175 Tm
0.0038 Tc
0.1804 Tw
[( = )]TJ
/F1 1 Tf
1.9617 0 TD
0 Tc
0 Tw
(y)Tj
6.7 0 0 6.7 176.04 105.535 Tm
(t)Tj
/F2 1 Tf
0.2866 0 TD
0.003 Tc
(1)Tj
11.5 0 0 11.5 184.68 108.175 Tm
0 Tc
0.1804 Tw
[( + )]TJ
/F1 1 Tf
1.4713 0 TD
0 Tw
(y)Tj
6.7 0 0 6.7 206.04 105.535 Tm
(t)Tj
/F2 1 Tf
0.2687 0 TD
0.003 Tc
(3)Tj
11.5 0 0 11.5 214.56 108.175 Tm
0.1887 Tc
( .)Tj
/F1 1 Tf
-13.7322 -2.087 TD
0 Tc
(B)Tj
/F2 1 Tf
0.6157 0 TD
0.002 Tc
0.1097 Tw
[( is the backward shift operator)64.6(, that is, )]TJ
/F1 1 Tf
15.9443 0 TD
0 Tc
0 Tw
(B)Tj
/F2 1 Tf
0.6157 0 TD
(\()Tj
/F1 1 Tf
0.2817 0 TD
(x)Tj
6.7 0 0 6.7 262.44 81.535 Tm
(t)Tj
/F2 1 Tf
11.5 0 0 11.5 264.24 84.175 Tm
0.0004 Tc
0.1087 Tw
(\) = )Tj
/F1 1 Tf
1.607 0 TD
0 Tc
0 Tw
(x)Tj
6.7 0 0 6.7 287.88 81.535 Tm
(t)Tj
/F2 1 Tf
0.2687 0 TD
0.003 Tc
(1)Tj
11.5 0 0 11.5 296.4 84.175 Tm
0.0016 Tc
0.1109 Tw
(, and )Tj
/F4 1 Tf
10 0 2.1 10 323.64 84.175 Tm
0 Tc
0 Tw
()Tj
/F1 1 Tf
6.7 0 0 6.7 328.08 81.535 Tm
(t)Tj
/F2 1 Tf
11.5 0 0 11.5 330 84.175 Tm
0.0013 Tc
0.1098 Tw
[( is a normally and)]TJ
-23.7704 -1.0435 TD
0.213 Tw
(independently distributed error term with zero mean and constant variance.)Tj
T*
0.0011 Tc
0.0907 Tw
(Deterministic components which include an intercept, three seasonal dummies)Tj
ET
endstream
endobj
23 0 obj
<<
/ProcSet [ /PDF /Text ]
/Font << /F1 6 0 R /F2 7 0 R /F4 17 0 R >>
/ExtGState << /GS2 9 0 R >>
>>
endobj
24 0 obj
<<
/Type /Page
/Parent 10 0 R
/Resources 26 0 R
/Contents 25 0 R
>>
endobj
25 0 obj
<< /Length 9947 >>
stream
BT
/F2 1 Tf
9.5 0 0 9.2 203.38 654.895 Tm
0 0 0 rg
/GS2 gs
0 Tc
0 Tw
(T)Tj
8.1 0 0 7.8 209.26 654.895 Tm
0.0025 Tc
(OURISM)Tj
9.5 0 0 9.2 239.5 654.895 Tm
0.1666 Tc
( E)Tj
8.1 0 0 7.8 248.86 654.895 Tm
0.0018 Tc
(CONOMICS)Tj
9 0 0 9 68.02 655.135 Tm
0.01 Tc
(698)Tj
11.5 0 0 11.5 68.02 635.455 Tm
0.0011 Tc
0.087 Tw
(and a time trend are also included in Equation \(3\), which can be estimated by)Tj
0 -1.0435 TD
0.0012 Tc
0.1585 Tw
(ordinary least squares \(OLS\). The three null and alternative hypotheses to be)Tj
T*
0.187 Tw
(tested are as follows:)Tj
/F1 1 Tf
1.0435 -2.0765 TD
0 Tc
0 Tw
(H)Tj
/F2 1 Tf
6.7 0 0 6.7 88.3 584.815 Tm
-0.0082 Tc
(0 )Tj
11.5 0 0 11.5 94.54 587.575 Tm
0 Tc
(: )Tj
/F4 1 Tf
10 0 2.1 10 102.34 587.575 Tm
( )Tj
/F2 1 Tf
6.7 0 0 6.7 107.86 584.815 Tm
(1)Tj
11.5 0 0 11.5 111.22 587.575 Tm
0.0035 Tc
0.1811 Tw
[( = 0, )]TJ
/F1 1 Tf
2.6609 0 TD
0 Tc
0 Tw
(H)Tj
/F2 1 Tf
6.7 0 0 6.7 150.1 584.815 Tm
(1)Tj
11.5 0 0 11.5 153.46 587.575 Tm
0.005 Tc
0.1783 Tw
[( : )]TJ
/F4 1 Tf
10 0 2.1 10 166.3 587.575 Tm
0 Tc
0 Tw
( )Tj
/F2 1 Tf
6.7 0 0 6.7 171.82 584.815 Tm
(1)Tj
11.5 0 0 11.5 175.18 587.575 Tm
0.0035 Tc
0.1811 Tw
[( < 0 ;)]TJ
/F1 1 Tf
-8.2748 -2.087 TD
0 Tc
0 Tw
(H)Tj
/F2 1 Tf
6.7 0 0 6.7 88.3 560.935 Tm
-0.0082 Tc
(0 )Tj
11.5 0 0 11.5 94.54 563.575 Tm
0 Tc
(: )Tj
/F4 1 Tf
10 0 2.1 10 102.34 563.575 Tm
( )Tj
/F2 1 Tf
6.7 0 0 6.7 107.86 560.935 Tm
(2)Tj
11.5 0 0 11.5 111.22 563.575 Tm
0.0035 Tc
0.1811 Tw
[( = 0, )]TJ
/F1 1 Tf
2.6609 0 TD
0 Tc
0 Tw
(H)Tj
/F2 1 Tf
6.7 0 0 6.7 150.1 560.935 Tm
(1)Tj
11.5 0 0 11.5 153.46 563.575 Tm
0.005 Tc
0.1783 Tw
[( : )]TJ
/F4 1 Tf
10 0 2.1 10 166.3 563.575 Tm
0 Tc
0 Tw
( )Tj
/F2 1 Tf
6.7 0 0 6.7 171.82 560.935 Tm
(2)Tj
11.5 0 0 11.5 175.18 563.575 Tm
0.0035 Tc
0.1811 Tw
[( < 0 ;)]TJ
/F1 1 Tf
-8.2748 -2.0765 TD
0 Tc
0 Tw
(H)Tj
/F2 1 Tf
6.7 0 0 6.7 88.3 537.055 Tm
-0.0082 Tc
(0 )Tj
11.5 0 0 11.5 94.54 539.695 Tm
0 Tc
(: )Tj
/F4 1 Tf
10 0 2.1 10 102.34 539.695 Tm
( )Tj
/F2 1 Tf
6.7 0 0 6.7 107.86 537.055 Tm
(3)Tj
11.5 0 0 11.5 111.22 539.695 Tm
0.1804 Tw
[( = )]TJ
/F4 1 Tf
10 0 2.1 10 128.14 539.695 Tm
0 Tw
( )Tj
/F2 1 Tf
6.7 0 0 6.7 133.66 537.055 Tm
(4)Tj
11.5 0 0 11.5 137.02 539.695 Tm
0.0023 Tc
0.1797 Tw
[( = 0, )]TJ
/F1 1 Tf
2.6504 0 TD
0 Tc
0 Tw
(H)Tj
/F2 1 Tf
6.7 0 0 6.7 175.9 537.055 Tm
(1)Tj
11.5 0 0 11.5 179.26 539.695 Tm
0.005 Tc
0.1783 Tw
[( : )]TJ
/F4 1 Tf
10 0 2.1 10 192.1 539.695 Tm
0 Tc
0 Tw
( )Tj
/F2 1 Tf
6.7 0 0 6.7 197.62 537.055 Tm
(3)Tj
11.5 0 0 11.5 200.98 539.695 Tm
( )Tj
/F4 1 Tf
10 0 0 10 205.9 539.695 Tm
(\012)Tj
/F2 1 Tf
11.5 0 0 11.5 211.54 539.695 Tm
0.0015 Tc
0.1797 Tw
[( 0 and/or )]TJ
/F4 1 Tf
10 0 2.1 10 260.38 539.695 Tm
0 Tc
0 Tw
( )Tj
/F2 1 Tf
6.7 0 0 6.7 265.9 537.055 Tm
(4)Tj
11.5 0 0 11.5 269.26 539.695 Tm
( )Tj
/F4 1 Tf
10 0 0 10 274.18 539.695 Tm
(\012)Tj
/F2 1 Tf
11.5 0 0 11.5 279.7 539.695 Tm
0.0022 Tc
0.1804 Tw
[( 0 .)]TJ
-18.407 -2.0765 TD
0.0014 Tc
0.1071 Tw
(The HEGY test involves the use of the )Tj
/F1 1 Tf
16.0278 0 TD
0 Tc
0 Tw
(t)Tj
/F2 1 Tf
0.2713 0 TD
0.0016 Tc
0.1073 Tw
(-test for the first two hypotheses and)Tj
-16.2991 -1.0435 TD
0.0022 Tc
0 Tw
(an )Tj
/F1 1 Tf
1.2522 0 TD
0 Tc
(F)Tj
/F2 1 Tf
0.5635 0 TD
0.0017 Tc
0.1123 Tw
(-test for the third hypothesis. If the first hypothesis is not rejected, there)Tj
-1.8157 -1.0435 TD
0.0015 Tc
0.1452 Tw
[(is a unit root at the zero frequency)95.4(, or a non-seasonal unit root in the series.)]TJ
T*
0.0013 Tc
0.038 Tw
(Non-rejection of the second hypothesis implies that there is a seasonal unit root)Tj
T*
0.179 Tw
[(at the semi-annual frequency)84.8(. Finally)95.2(, if the third hypothesis is not rejected,)]TJ
T*
0.0207 Tw
[(there is a seasonal unit root at the annual frequency)95.2(. These three null hypotheses)]TJ
T*
0.0441 Tw
[(are tested separately)84.8(. If none of the three null hypotheses is rejected, a quarterly)]TJ
T*
0.13 Tw
(time series may have non-seasonal, semi-annual and/or annual unit roots. The)Tj
T*
0.0887 Tw
(order of integration of the series would be I\(1, 1, 1\). The rejection of all three)Tj
T*
0.1232 Tw
(null hypotheses implies that there is no non-seasonal or seasonal unit root, in)Tj
T*
0.0246 Tw
(which case the series is a stationary one and the order of integration of the series)Tj
T*
0.0017 Tc
0.1804 Tw
(would be I\(0, 0, 0\).)Tj
1.0435 -1.0435 TD
0.0015 Tc
0.2966 Tw
(According to the results of the seasonal unit root test, if the order of)Tj
-1.0435 -1.0435 TD
0.0017 Tc
0.1671 Tw
(integration of the series is I\(1, 1, 1\), it requires the filters \(1 )Tj
/F1 1 Tf
26.3791 0 TD
0 Tc
0 Tw
(B)Tj
/F2 1 Tf
0.6157 0 TD
0.0017 Tc
0.1681 Tw
(\), \(1 + )Tj
/F1 1 Tf
3.1722 0 TD
0 Tc
0 Tw
(B)Tj
/F2 1 Tf
0.6157 0 TD
(\))Tj
-30.7826 -1.0435 TD
0.0019 Tc
0.113 Tw
(and \(1 + )Tj
/F1 1 Tf
3.8713 0 TD
0 Tc
0 Tw
(B)Tj
/F2 1 Tf
6.7 0 0 6.7 119.62 351.655 Tm
(2)Tj
11.5 0 0 11.5 122.98 347.815 Tm
0.0014 Tc
0.113 Tw
[(\), respectively)95.3(, to achieve stationarity)84.9(. In other words, the seasonal)]TJ
-4.7791 -1.0435 TD
0.0015 Tc
0.1652 Tw
(differencing filter \(1 )Tj
/F1 1 Tf
9.4748 0 TD
0 Tc
0 Tw
(B)Tj
/F2 1 Tf
6.7 0 0 6.7 184.06 339.655 Tm
(4)Tj
11.5 0 0 11.5 187.42 335.815 Tm
0.0015 Tc
0.1662 Tw
(\) should be applied to obtain a stationary series. If)Tj
-10.3826 -1.0435 TD
0.0602 Tw
(the order of integration is I\(0, 0, 0\), it implies that the series has deterministic)Tj
T*
0.0014 Tc
0.1161 Tw
[(or constant seasonality)95.3(. In this case, it is sufficient to use dummy variables to)]TJ
T*
0.1864 Tw
(capture the seasonal variations in the time series.)Tj
/F1 1 Tf
0 -2.0765 TD
0.0012 Tc
0.1192 Tw
(Unobservable component.)Tj
/F2 1 Tf
9.0887 0 TD
(STSM treats seasonality as an unobservable component.)Tj
-9.0887 -1.0435 TD
0.3059 Tw
(Based on the traditional econometric regression model, STSM additionally)Tj
T*
0.1881 Tw
[(includes the trend, seasonal, cyclical and irregular components:)]TJ
/F1 1 Tf
1.0435 -1.5652 TD
0 Tc
0 Tw
(Y)Tj
6.7 0 0 6.7 87.7 231.295 Tm
(t)Tj
/F2 1 Tf
11.5 0 0 11.5 89.5 233.935 Tm
0.0022 Tc
0.1782 Tw
[( = )]TJ
/F4 1 Tf
10 0 2.1 10 106.42 233.935 Tm
0 Tc
0 Tw
()Tj
/F2 1 Tf
6.7 0 0 6.7 111.94 231.295 Tm
(1)Tj
/F1 1 Tf
11.5 0 0 11.5 115.3 233.935 Tm
(x)Tj
/F2 1 Tf
6.7 0 0 6.7 120.46 231.295 Tm
(1)Tj
11.5 0 0 11.5 123.7 233.935 Tm
0.1804 Tw
[( + )]TJ
/F4 1 Tf
10 0 2.1 10 140.62 233.935 Tm
0 Tw
()Tj
/F2 1 Tf
6.7 0 0 6.7 146.26 231.295 Tm
(2)Tj
/F1 1 Tf
11.5 0 0 11.5 149.5 233.935 Tm
(x)Tj
/F2 1 Tf
6.7 0 0 6.7 154.66 231.295 Tm
(2)Tj
11.5 0 0 11.5 158.02 233.935 Tm
0.0036 Tc
0.1805 Tw
[( + .)-246.8(.)-247.3(. + )]TJ
/F4 1 Tf
10 0 2.1 10 206.26 233.935 Tm
0 Tc
0 Tw
()Tj
/F1 1 Tf
6.7 0 0 6.7 211.9 231.295 Tm
(k)Tj
11.5 0 0 11.5 214.78 233.935 Tm
(x)Tj
6.7 0 0 6.7 219.94 231.295 Tm
(k)Tj
/F2 1 Tf
11.5 0 0 11.5 222.82 233.935 Tm
0.0022 Tc
0.1782 Tw
[( + )]TJ
/F4 1 Tf
10 0 2.1 10 239.74 233.935 Tm
0 Tc
0 Tw
()Tj
/F1 1 Tf
6.7 0 0 6.7 245.5 231.295 Tm
(t)Tj
/F2 1 Tf
11.5 0 0 11.5 247.3 233.935 Tm
0.1804 Tw
[( + )]TJ
/F4 1 Tf
10 0 2.1 10 264.22 233.935 Tm
0 Tw
()Tj
/F1 1 Tf
6.7 0 0 6.7 268.42 231.295 Tm
(t)Tj
/F2 1 Tf
11.5 0 0 11.5 270.22 233.935 Tm
0.1804 Tw
[( + )]TJ
/F4 1 Tf
10 0 2.1 10 287.14 233.935 Tm
0 Tw
()Tj
/F1 1 Tf
6.7 0 0 6.7 295.18 231.295 Tm
(t)Tj
/F2 1 Tf
11.5 0 0 11.5 296.98 233.935 Tm
0.1804 Tw
[( + )]TJ
/F4 1 Tf
10 0 2.1 10 313.9 233.935 Tm
0 Tw
()Tj
/F1 1 Tf
6.7 0 0 6.7 318.34 231.295 Tm
(t)Tj
/F2 1 Tf
11.5 0 0 11.5 413.02 233.935 Tm
-0.001 Tc
(\(4\))Tj
-30 -1.5548 TD
0.0007 Tc
(where )Tj
/F1 1 Tf
2.6922 0 TD
0 Tc
(Y)Tj
6.7 0 0 6.7 106.66 213.415 Tm
(t)Tj
/F2 1 Tf
11.5 0 0 11.5 108.46 216.055 Tm
0.0014 Tc
0.1637 Tw
[( is the actual tourism demand, )]TJ
/F1 1 Tf
12.9496 0 TD
0 Tc
0 Tw
(x)Tj
/F2 1 Tf
6.7 0 0 6.7 262.42 213.415 Tm
(1)Tj
11.5 0 0 11.5 265.78 216.055 Tm
(,)Tj
/F1 1 Tf
0.2504 0 TD
(x)Tj
/F2 1 Tf
6.7 0 0 6.7 273.82 213.415 Tm
(2)Tj
11.5 0 0 11.5 277.18 216.055 Tm
0.0009 Tc
[(,.)-249.5(.)-260.4(.)0(,)]TJ
/F1 1 Tf
1.7635 0 TD
0 Tc
(x)Tj
6.7 0 0 6.7 302.62 213.415 Tm
(k)Tj
/F2 1 Tf
11.5 0 0 11.5 305.5 216.055 Tm
0.001 Tc
0.1637 Tw
[( are explanatory variables,)]TJ
/F4 1 Tf
10 0 2.1 10 68.02 204.055 Tm
0 Tc
0 Tw
()Tj
/F2 1 Tf
6.7 0 0 6.7 73.66 201.415 Tm
(1)Tj
11.5 0 0 11.5 76.9 204.055 Tm
(,)Tj
/F4 1 Tf
10 0 2.1 10 79.9 204.055 Tm
()Tj
/F2 1 Tf
6.7 0 0 6.7 85.42 201.415 Tm
(2)Tj
11.5 0 0 11.5 88.78 204.055 Tm
0.0009 Tc
[(,.)-249.5(.)-260.4(.)0(,)]TJ
/F4 1 Tf
10 0 2.1 10 109.06 204.055 Tm
0 Tc
()Tj
/F1 1 Tf
6.7 0 0 6.7 114.7 201.415 Tm
(k )Tj
/F2 1 Tf
11.5 0 0 11.5 120.58 204.055 Tm
0.0009 Tc
0.2087 Tw
(are unknown parameters and )Tj
/F4 1 Tf
10 0 2.1 10 260.26 204.055 Tm
0 Tc
0 Tw
()Tj
/F1 1 Tf
6.7 0 0 6.7 266.02 201.415 Tm
(t)Tj
/F2 1 Tf
11.5 0 0 11.5 267.82 204.055 Tm
0.007 Tc
(, )Tj
/F4 1 Tf
10 0 2.1 10 276.1 204.055 Tm
0 Tc
()Tj
/F1 1 Tf
6.7 0 0 6.7 280.3 201.415 Tm
(t)Tj
/F2 1 Tf
11.5 0 0 11.5 282.1 204.055 Tm
0.0009 Tc
(, )Tj
/F4 1 Tf
10 0 2.1 10 290.26 204.055 Tm
0 Tc
()Tj
/F1 1 Tf
6.7 0 0 6.7 298.3 201.415 Tm
(t)Tj
/F2 1 Tf
11.5 0 0 11.5 300.1 204.055 Tm
0.0009 Tc
(, )Tj
/F4 1 Tf
10 0 2.1 10 308.38 204.055 Tm
0 Tc
()Tj
/F1 1 Tf
6.7 0 0 6.7 312.82 201.415 Tm
-0.0131 Tc
(t )Tj
/F2 1 Tf
11.5 0 0 11.5 317.62 204.055 Tm
0.0011 Tc
0.2102 Tw
(are the trend, seasonal,)Tj
-21.7043 -1.0435 TD
0.301 Tw
[(cyclical and irregular components, respectively)95(. Such a treatment generally)]TJ
T*
0.15 Tw
(assumes that seasonality evolves gradually over time, while the fixed seasonal)Tj
T*
0.1817 Tw
(effect can be embodied in the specifications of the unobservable components)Tj
T*
0.0009 Tc
0.1898 Tw
(as a special case.)Tj
/F3 1 Tf
12 0 0 12 223.42 119.935 Tm
0.0001 Tc
0.175 Tw
(The data)Tj
/F2 1 Tf
11.5 0 0 11.5 68.02 95.935 Tm
0.0013 Tc
0.1708 Tw
(The empirical study focuses on the demand for outbound leisure tourism by)Tj
T*
0.1754 Tw
(UK residents to seven major destinations: Australia, Canada, France, Greece,)Tj
T*
0.0916 Tw
[(Italy)95.2(, Spain and the USA. The tourism demand function can be written in the)]TJ
T*
0.0009 Tc
0.187 Tw
(following general form:)Tj
ET
endstream
endobj
26 0 obj
<<
/ProcSet [ /PDF /Text ]
/Font << /F1 6 0 R /F2 7 0 R /F3 8 0 R /F4 17 0 R >>
/ExtGState << /GS2 9 0 R >>
>>
endobj
27 0 obj
<<
/Type /Page
/Parent 10 0 R
/Resources 29 0 R
/Contents 28 0 R
>>
endobj
28 0 obj
<< /Length 8624 >>
stream
BT
/F2 1 Tf
9 0 0 9 400.2 655.135 Tm
0 0 0 rg
/GS2 gs
0.01 Tc
0 Tw
(699)Tj
/F1 1 Tf
-25.64 0 TD
0.0083 Tc
0.1666 Tw
(Seasonality and tourism demand models)Tj
11.5 0 0 11.5 68.64 635.455 Tm
0.0051 Tc
0 Tw
(TOU)Tj
6.7 0 0 6.7 90.48 632.8151 Tm
-0.0008 Tc
(it)Tj
/F2 1 Tf
11.5 0 0 11.5 94.2 635.455 Tm
0.0013 Tc
0.1739 Tw
[( = )]TJ
/F1 1 Tf
1.4609 0 TD
0 Tc
0 Tw
(f)Tj
/F2 1 Tf
0.2713 0 TD
(\()Tj
/F1 1 Tf
0.2817 0 TD
(Y)Tj
6.7 0 0 6.7 125.04 632.8151 Tm
(t)Tj
/F2 1 Tf
11.5 0 0 11.5 126.84 635.455 Tm
0.0044 Tc
(, )Tj
/F1 1 Tf
0.6783 0 TD
0.0037 Tc
(RRCP)Tj
6.7 0 0 6.7 162.24 632.8151 Tm
-0.0008 Tc
(it)Tj
/F2 1 Tf
11.5 0 0 11.5 165.96 635.455 Tm
0.0044 Tc
(, )Tj
/F1 1 Tf
0.6783 0 TD
0.004 Tc
(RSUB)Tj
6.7 0 0 6.7 201.24 632.8151 Tm
-0.0008 Tc
(it)Tj
/F2 1 Tf
11.5 0 0 11.5 204.96 635.455 Tm
0.0044 Tc
(, )Tj
/F1 1 Tf
0.6783 0 TD
0.0015 Tc
(dummies)Tj
/F2 1 Tf
2.9635 0 TD
-0.001 Tc
[(\))-13183.9(\(5\))]TJ
-16.5391 -2.0765 TD
-0.0002 Tc
(where )Tj
/F1 1 Tf
2.5461 0 TD
0.0051 Tc
(TOU)Tj
6.7 0 0 6.7 107.64 608.8151 Tm
-0.0008 Tc
(it)Tj
/F2 1 Tf
11.5 0 0 11.5 111.36 611.5751 Tm
0.0013 Tc
0.0108 Tw
[( is the UK outbound leisure tourism demand measured by quarterly)]TJ
-4.7583 -1.0435 TD
0.0015 Tc
0.1282 Tw
(tourist arrivals to the destination country )Tj
/F1 1 Tf
16.8939 0 TD
0 Tc
0 Tw
(i)Tj
/F2 1 Tf
0.2713 0 TD
(; )Tj
/F1 1 Tf
0.6365 0 TD
(Y)Tj
6.7 0 0 6.7 269.04 596.8151 Tm
(t)Tj
/F2 1 Tf
11.5 0 0 11.5 270.84 599.575 Tm
0.0016 Tc
0.1287 Tw
[( is tourist income measured by)]TJ
-18.6261 -1.0435 TD
0.0014 Tc
0.0656 Tw
(real gross domestic product \(GDP\) of the UK in constant prices \(1995 = 100\);)Tj
/F1 1 Tf
T*
0.0037 Tc
0 Tw
(RRCP)Tj
6.7 0 0 6.7 84.24 572.815 Tm
-0.0008 Tc
(it)Tj
/F2 1 Tf
11.5 0 0 11.5 87.96 575.575 Tm
0.0013 Tc
0.037 Tw
[( represents the relative tourism price of destination )]TJ
/F1 1 Tf
20.1496 0 TD
0 Tc
0 Tw
(i)Tj
/F2 1 Tf
0.2713 0 TD
0.0009 Tc
0.0377 Tw
(, which is calculated)Tj
-23.1443 -1.0435 TD
0.0013 Tc
0.1015 Tw
(by dividing the cost of tourism \(measured by the consumer price index, )Tj
/F1 1 Tf
28.9565 0 TD
0.0065 Tc
0 Tw
(CPI)Tj
6.7 0 0 6.7 406.92 560.815 Tm
-0.0008 Tc
(it)Tj
/F2 1 Tf
11.5 0 0 11.5 410.64 563.575 Tm
0 Tc
(\))Tj
-30.7826 -1.0435 TD
0.0014 Tc
0.3193 Tw
(in each destination by the UK CPI \()Tj
/F1 1 Tf
16.153 0 TD
0.0065 Tc
0 Tw
(CPI)Tj
6.7 0 0 6.7 259.68 548.815 Tm
0.0103 Tc
(UK)Tj
/F2 1 Tf
11.5 0 0 11.5 268.68 551.575 Tm
0.0014 Tc
0.3206 Tw
(\), adjusted by the appropriate)Tj
-18.4383 -1.0435 TD
0.0006 Tc
0.187 Tw
(exchange rates \()Tj
/F1 1 Tf
6.4487 0 TD
-0.005 Tc
0 Tw
(EX)Tj
6.7 0 0 6.7 145.44 536.815 Tm
0 Tc
(i)Tj
/F2 1 Tf
11.5 0 0 11.5 147.36 539.575 Tm
0.0019 Tc
0.1891 Tw
[( and )]TJ
/F1 1 Tf
2.2748 0 TD
-0.005 Tc
0 Tw
(EX)Tj
6.7 0 0 6.7 188.16 536.815 Tm
0.0103 Tc
(UK)Tj
/F2 1 Tf
11.5 0 0 11.5 197.16 539.575 Tm
0.0042 Tc
(\):)Tj
/F1 1 Tf
-11.1757 -1.5652 TD
0.1813 Tc
( )Tj
4.32 -0.3444 TD
0.0065 Tc
(CPI)Tj
6.7 0 0 6.7 135.6 514.975 Tm
-0.0008 Tc
(it)Tj
/F2 1 Tf
11.5 0 0 11.5 139.32 517.615 Tm
0 Tc
(/)Tj
/F1 1 Tf
0.2817 0 TD
-0.005 Tc
(EX)Tj
6.7 0 0 6.7 157.2 514.975 Tm
0 Tc
(i)Tj
11.5 0 0 11.5 68.64 509.575 Tm
0.0037 Tc
(RRCP)Tj
6.7 0 0 6.7 96.24 506.935 Tm
-0.0008 Tc
(it)Tj
/F2 1 Tf
11.5 0 0 11.5 99.96 509.575 Tm
0.0008 Tc
0.1723 Tw
[( = )-20774.9(\(6\))]TJ
/F1 1 Tf
-2.7235 -1.0435 TD
0.1834 Tc
0 Tw
( )Tj
4.3304 0.3443 TD
0.0065 Tc
(CPI)Tj
6.7 0 0 6.7 135.84 498.895 Tm
-0.0104 Tc
(uk)Tj
/F2 1 Tf
11.5 0 0 11.5 141.72 501.535 Tm
0 Tc
(/)Tj
/F1 1 Tf
0.2817 0 TD
-0.005 Tc
(EX)Tj
6.7 0 0 6.7 159.6 498.895 Tm
-0.0104 Tc
(uk)Tj
11.5 0 0 11.5 56.64 479.575 Tm
0.004 Tc
(RSUB)Tj
6.7 0 0 6.7 84.24 476.935 Tm
-0.0008 Tc
(it)Tj
/F2 1 Tf
11.5 0 0 11.5 87.96 479.575 Tm
0.0014 Tc
0.1773 Tw
[( represents the relative substitute price of the destination )]TJ
/F1 1 Tf
23.76 0 TD
0 Tc
0 Tw
(i)Tj
/F2 1 Tf
0.2817 0 TD
0.0011 Tc
0.1783 Tw
(, measured)Tj
-26.7652 -1.0435 TD
0.1238 Tw
(by a weighted price index of the main alternative destinations for each of the)Tj
T*
0.0015 Tc
0.0853 Tw
(destinations relative to the tourism price in the UK, with the shares of tourist)Tj
T*
0.086 Tw
[(arrivals in these potential substitute destinations being the weights. For short-)]TJ
T*
0.0013 Tc
0.1222 Tw
[(haul destinations, the four W)74.3(estern European countries \(France, Spain, Greece)]TJ
T*
0.1884 Tw
(and Italy\) are treated as substitute countries to each other; in the long-haul)Tj
T*
0.0779 Tw
[(cases, Australia, Canada and the USA are regarded as substitutes to each other)53.4(.)]TJ
T*
0.0873 Tw
(New Zealand is also added to the long-haul substitution set as it is one of the)Tj
T*
0.0098 Tc
0.3277 Tw
(most popular long-haul destinations for UK tourists. Therefore, Canada,)Tj
T*
0.0013 Tc
0.1214 Tw
[(Australia)-259.6(and New Zealand are substitute destinations for the USA; the USA,)]TJ
T*
0.001 Tc
0.0308 Tw
(Australia and New Zealand for Canada; and the USA, Canada and New Zealand)Tj
T*
0.0014 Tc
0.0814 Tw
(for Australia. The reason for choosing these countries as substitutions has been)Tj
T*
0.1996 Tw
(justified by Divisekera \(2003\). In the case of France, the substitute price is)Tj
T*
0.001 Tc
0.1913 Tw
(defined as:)Tj
/F1 1 Tf
1.0435 -1.5548 TD
0.1782 Tc
0 Tw
( )Tj
4.2783 -0.3443 TD
0.0037 Tc
(RRCP)Tj
6.7 0 0 6.7 145.44 298.975 Tm
-0.0043 Tc
(sp)Tj
/F2 1 Tf
8 0 0 8 150 301.735 Tm
0 Tc
()Tj
/F1 1 Tf
11.5 0 0 11.5 154.08 301.735 Tm
0.0051 Tc
(TOU)Tj
6.7 0 0 6.7 175.8 298.975 Tm
-0.0043 Tc
(sp)Tj
/F2 1 Tf
11.5 0 0 11.5 180.24 301.735 Tm
0.0044 Tc
0.176 Tw
[( + )]TJ
/F1 1 Tf
1.4609 0 TD
0.0037 Tc
0 Tw
(RRCP)Tj
6.7 0 0 6.7 224.64 298.975 Tm
-0.0008 Tc
(it)Tj
/F2 1 Tf
8 0 0 8 228.36 301.735 Tm
0 Tc
()Tj
/F1 1 Tf
11.5 0 0 11.5 232.44 301.735 Tm
0.0051 Tc
(TOU)Tj
6.7 0 0 6.7 254.28 298.975 Tm
-0.0008 Tc
(it)Tj
/F2 1 Tf
11.5 0 0 11.5 258 301.735 Tm
0.0044 Tc
0.176 Tw
[( + )]TJ
/F1 1 Tf
1.4609 0 TD
0.0037 Tc
0 Tw
(RRCP)Tj
6.7 0 0 6.7 302.4 298.975 Tm
-0.0056 Tc
(gr)Tj
/F2 1 Tf
8 0 0 8 307.2 301.735 Tm
0 Tc
()Tj
/F1 1 Tf
11.5 0 0 11.5 311.4 301.735 Tm
0.0051 Tc
(TOU)Tj
6.7 0 0 6.7 333.12 298.975 Tm
-0.0056 Tc
(gr)Tj
11.5 0 0 11.5 68.64 293.695 Tm
0.004 Tc
(RSUB)Tj
6.7 0 0 6.7 96.24 291.055 Tm
-0.002 Tc
(fr)Tj
/F2 1 Tf
11.5 0 0 11.5 100.32 293.695 Tm
0.001 Tc
0.1741 Tw
[( = )-5206(\(7\))]TJ
/F1 1 Tf
-2.7548 -1.0435 TD
0.1816 Tc
0 Tw
( )Tj
8.64 0.3443 TD
0.0051 Tc
(TOU)Tj
6.7 0 0 6.7 189.72 283.015 Tm
-0.0043 Tc
(sp)Tj
/F2 1 Tf
11.5 0 0 11.5 194.16 285.655 Tm
0.0022 Tc
0.1782 Tw
[( + )]TJ
/F1 1 Tf
1.4713 0 TD
0.0051 Tc
0 Tw
(TOU)Tj
6.7 0 0 6.7 232.8 283.015 Tm
-0.0008 Tc
(it)Tj
/F2 1 Tf
11.5 0 0 11.5 236.52 285.655 Tm
0 Tc
0.1804 Tw
[( + )]TJ
/F1 1 Tf
1.4713 0 TD
0.0051 Tc
0 Tw
(TOU)Tj
6.7 0 0 6.7 275.28 283.015 Tm
-0.0056 Tc
(gr)Tj
/F2 1 Tf
11.5 0 0 11.5 56.64 263.695 Tm
0.0012 Tc
0.0294 Tw
(Three dummy variables are included in the models to capture the effects of one-)Tj
0 -1.0435 TD
0.3031 Tw
[(off events on the UK outbound tourism demand. Among them, DUM86)]TJ
T*
0.1289 Tw
(represents the severe decline of the world oil prices in 1986 \(DUM86 = 1 in)Tj
T*
0.0945 Tw
(1986Q2 and 1986Q3, 0 otherwise\). The severe decline of world oil prices was)Tj
T*
0.0695 Tw
(due to disagreement in the oil cartel \(OPEC\), which soon led to its breakdown)Tj
T*
0.1241 Tw
[(\(T)63.9(rehan, 1986\). The decline of world oil prices is supposed to have a positive)]TJ
T*
0.1893 Tw
[(effect on UK outbound tourism demand. DUM90 captures the effect of the)]TJ
T*
0.0014 Tc
0.1648 Tw
(invasion of Kuwait by Iraq in 1990 \(DUM90 = 1 in 1990Q3 and 1990Q4,)Tj
T*
0.1627 Tw
[(0 otherwise\). DUM91 is used to detect the effects of the Gulf W)64(ar in 1991)]TJ
T*
0.0185 Tw
(\(DUM91 = 1 in 1991Q1, 1991Q2 and 1991Q3, 0 otherwise\). These two events)Tj
T*
0.0012 Tc
0.1854 Tw
[(may have negative effects on UK outbound tourism demand.)]TJ
1.0435 -1.0435 TD
0.2226 Tw
[(T)69.7(ravel costs between the origin and destination countries may also affect)]TJ
-1.0435 -1.0435 TD
0.0635 Tw
[(tourism demand to some extent. However)42.9(, precise measurements of travel costs)]TJ
T*
0.0009 Tc
0.0486 Tw
[(are rarely available, especially at the aggregate level. The average economy-class)]TJ
T*
0.0012 Tc
0.1206 Tw
(airfares between the origin and destination were used to represent travel costs)Tj
T*
0.2467 Tw
[(in past studies. However)53.5(, due to the complex structure of airfares and the)]TJ
T*
0.001 Tc
0.0808 Tw
[(low)-259.9(relevance of airfares to the overall price of all-inclusive package tours, only)]TJ
T*
0.0012 Tc
0.2475 Tw
(in a few cases did the use of such a proxy result in significant coefficient)Tj
ET
endstream
endobj
29 0 obj
<<
/ProcSet [ /PDF /Text ]
/Font << /F1 6 0 R /F2 7 0 R >>
/ExtGState << /GS2 9 0 R >>
>>
endobj
30 0 obj
<<
/Type /Page
/Parent 10 0 R
/Resources 32 0 R
/Contents 31 0 R
>>
endobj
31 0 obj
<< /Length 5727 >>
stream
BT
/F2 1 Tf
9.5 0 0 9.2 203.38 654.895 Tm
0 0 0 rg
/GS2 gs
0 Tc
0 Tw
(T)Tj
8.1 0 0 7.8 209.26 654.895 Tm
0.0025 Tc
(OURISM)Tj
9.5 0 0 9.2 239.5 654.895 Tm
0.1666 Tc
( E)Tj
8.1 0 0 7.8 248.86 654.895 Tm
0.0018 Tc
(CONOMICS)Tj
9 0 0 9 68.02 655.135 Tm
0.01 Tc
(700)Tj
ET
0 0 0 RG
0 J 0 j 0.96 w 10 M []0 d
/GS1 gs
1 i
68.02 642.895 m
425.26 642.895 l
S
0.48 w
68.02 625.255 m
425.26 625.255 l
S
BT
/F3 1 Tf
9.5 0 0 9.5 68.02 630.655 Tm
/GS2 gs
0.0065 Tc
-0.0074 Tw
[(T)79.8(able 1.)-498.8(Results of unit root tests.)]TJ
0 -1.9579 TD
0.0064 Tc
0 Tw
(Destination country)Tj
/F5 1 Tf
10.4589 0 TD
0.0078 Tc
[(L)71.2(TOU)]TJ
5.5 0 0 5.5 191.74 609.775 Tm
0.0104 Tc
(it)Tj
9.5 0 0 9.5 245.86 612.0551 Tm
-0.076 Tc
(LY)Tj
5.5 0 0 5.5 256.9 609.775 Tm
0 Tc
(t)Tj
9.5 0 0 9.5 306.1 612.0551 Tm
0.0038 Tc
(LRRCP)Tj
5.5 0 0 5.5 336.46 609.775 Tm
0.0104 Tc
(it)Tj
9.5 0 0 9.5 377.26 612.0551 Tm
0.0052 Tc
(LRSUB)Tj
5.5 0 0 5.5 407.14 609.775 Tm
0.0104 Tc
(it)Tj
/F2 1 Tf
9.5 0 0 9.5 68.02 589.975 Tm
0.006 Tc
[(Australia)-6916.1(I\(1,1,0\))-4516.1(I\(1,0,0\))-4528.7(I\(1,0,0\))-4528.7(I\(1,0,0\))]TJ
0 -1.1621 TD
0.0058 Tc
[(France)-7939.5(I\(1,1,1\))-4516.3(I\(1,0,0\))-4528.9(I\(1,0,0\))-4528.9(I\(1,1,0\))]TJ
0 -1.1495 TD
0.0055 Tc
[(Greece)-7838.7(I\(1,1,0\))-4516.6(I\(1,0,0\))-4529.2(I\(1,1,0\))-4529.2(I\(1,1,0\))]TJ
0 -1.1621 TD
0.0061 Tc
[(Italy)-8659.2(I\(1,1,1\))-4516(I\(1,0,0\))-4528.6(I\(1,1,0\))-4528.6(I\(1,1,0\))]TJ
T*
0.0059 Tc
[(Spain)-8330.9(I\(1,1,1\))-4516.2(I\(1,0,0\))-4528.8(I\(1,0,0\))-4528.8(I\(1,1,0\))]TJ
0 -1.1495 TD
0.0054 Tc
[(Canada)-7674.6(I\(1,1,1\))-4516.7(I\(1,0,0\))-4529.3(I\(1,0,0\))-4529.3(I\(1,0,0\))]TJ
ET
0.96 w
/GS1 gs
68.02 518.815 m
425.26 518.815 l
S
BT
9.5 0 0 9.5 68.02 523.975 Tm
/GS2 gs
0.0059 Tc
[(U)6.6(S)6.6(A)-8634.1(I\(1,0,1\))-4516.2(I\(1,0,0\))-4528.8(I\(1,0,0\))-4528.8(I\(1,1,0\))]TJ
11.5 0 0 11.5 68.02 475.135 Tm
0.0011 Tc
0.1087 Tw
(estimates \(Li )Tj
/F1 1 Tf
5.353 0 TD
0.0004 Tc
(et al)Tj
/F2 1 Tf
1.6383 0 TD
0.0013 Tc
0.1096 Tw
(, 2005\). Therefore, the travel costs variable is excluded from)Tj
-6.9913 -1.0435 TD
0.0016 Tc
0.1782 Tw
[(this study)85.1(.)]TJ
1.0435 -1.0435 TD
0.0012 Tc
0.1347 Tw
[(The data cover the period 1984Q12004Q4. The series on GDP)126.4(, exchange)]TJ
-1.0435 -1.0435 TD
0.0016 Tc
0.1732 Tw
(rates and CPI are obtained from the )Tj
/F1 1 Tf
15.1826 0 TD
0.0012 Tc
0.174 Tw
[(International Financial Statistics Y)116(earbooks)]TJ
/F2 1 Tf
-15.1826 -1.0435 TD
0.0014 Tc
0.0926 Tw
(published by the International Monetary Fund \(IMF\). The tourist arrivals data)Tj
T*
0.2293 Tw
(are obtained from the )Tj
/F1 1 Tf
9.4435 0 TD
0.0012 Tc
0.2284 Tw
[(T)90.5(ourism Statistical Y)105.6(earbooks)]TJ
/F2 1 Tf
10.7374 0 TD
0.0015 Tc
0.2304 Tw
(, published by the United)Tj
-20.1809 -1.0435 TD
0.0014 Tc
0.1848 Tw
[(Nations W)64(orld T)84.9(ourism Or)22.3(ganization \(UNWTO\).)]TJ
/F3 1 Tf
12 0 0 12 203.02 367.135 Tm
-0.0004 Tc
0.1917 Tw
(Empirical results)Tj
/F2 1 Tf
11.5 0 0 11.5 68.02 343.135 Tm
0.0011 Tc
0.1809 Tw
(The six econometric models and three time-series models outlined above are)Tj
T*
0.0012 Tc
0.2446 Tw
(used to generate individual )Tj
/F1 1 Tf
11.6765 0 TD
0.0018 Tc
0.2435 Tw
(ex post)Tj
/F2 1 Tf
2.5043 0 TD
0.0012 Tc
0.2449 Tw
[( forecasts. The HEGY test developed by)]TJ
-14.1809 -1.0435 TD
0.0003 Tc
0 Tw
[(Hylleber)10.7(g )]TJ
/F1 1 Tf
4.2574 0 TD
0.0013 Tc
0.126 Tw
(et al)Tj
/F2 1 Tf
1.6591 0 TD
0.0015 Tc
0.1284 Tw
[( \(1990\) is used to test for seasonal and non-seasonal unit roots)]TJ
-5.9165 -1.0435 TD
(in the series. The results of unit root tests for the dependent and explanatory)Tj
T*
0.0013 Tc
0.1405 Tw
(variables related to UK tourists to the seven destinations under consideration)Tj
T*
0.0955 Tw
[(are presented in T)84.8(able 1. The tests were carried out for each of the time series)]TJ
T*
0.1819 Tw
(for the period 1984Q12004Q4 using EVIEWS 5.0.)Tj
/F1 1 Tf
12 0 0 12 138.94 245.935 Tm
0.001 Tc
0.175 Tw
(Performance of the individual forecasting methods)Tj
/F2 1 Tf
11.5 0 0 11.5 68.02 227.335 Tm
0.0016 Tc
0.0274 Tw
(The results of the HEGY test show that the UK outbound tourist arrivals series)Tj
T*
0.0012 Tc
0.0326 Tw
[(and some of the price variables exhibit trend and seasonality)84.7(. The reduced ADL,)]TJ
T*
0.001 Tc
0.163 Tw
[(V)94.1(AR and TVP models include seasonal dummy variables in the specification)]TJ
T*
0.0756 Tw
[(of the models to account for deterministic seasonality)84.6(, while seasonal difference)]TJ
T*
0.0013 Tc
0.0652 Tw
(has been used in the WB-ECM and JML-ECM approaches in which seasonality)Tj
T*
0.0971 Tw
[(is regarded as stochastic. BSMs and STSMs are estimated using ST)64(AMP 7 and)]TJ
T*
0.0016 Tc
0.1838 Tw
(the other models are estimated using EVIEW 5.0.)Tj
1.0435 -1.0435 TD
0.0012 Tc
0.3051 Tw
(The demand models are estimated based on the data from 1984Q1 to)Tj
-1.0435 -1.0435 TD
0.0009 Tc
0.0102 Tw
(1996Q4 and the )Tj
/F1 1 Tf
6.6365 0 TD
0.002 Tc
0.013 Tw
(ex post)Tj
/F2 1 Tf
2.2748 0 TD
0.0012 Tc
0.0112 Tw
[( forecasts are generated for the period 1997Q12004Q4.)]TJ
6.7 0 0 6.7 421.9 135.175 Tm
0 Tc
0 Tw
(1)Tj
11.5 0 0 11.5 68.02 119.335 Tm
0.0013 Tc
0.1755 Tw
(The recursive forecasting technique is used to generate forecasts; that is, the)Tj
0 -1.0435 TD
0.08 Tw
(models are estimated over the period 1984Q11996Q4 first and the estimated)Tj
T*
0.1435 Tw
(models are used to forecast tourist arrivals over the period 1997Q12004Q4.)Tj
T*
0.2587 Tw
[(Subsequently)95.3(, the models are re-estimated using the data from 1984Q1 to)]TJ
T*
0.0011 Tc
0.1674 Tw
(1997Q1 and forecasts are generated for the period 1997Q22004Q4. Such a)Tj
T*
0.0012 Tc
0.0986 Tw
(procedure is repeated until all observations are exhausted. As a result, 32 one-)Tj
ET
endstream
endobj
32 0 obj
<<
/ProcSet [ /PDF /Text ]
/Font << /F1 6 0 R /F2 7 0 R /F3 8 0 R /F5 33 0 R >>
/ExtGState << /GS1 34 0 R /GS2 9 0 R >>
>>
endobj
33 0 obj
<<
/Type /Font
/Subtype /Type1
/FirstChar 66
/LastChar 116
/Widths [ 667 667 778 611 611 778 778 389 444 667 556 833 778 667 611 722 667
500 667 722 667 889 778 667 611 389 278 389 600 500 333 500 500
333 500 389 333 444 500 333 333 500 278 778 556 389 500 500 389
333 333 ]
/Encoding /WinAnsiEncoding
/BaseFont /HPCDCK+GaramondThree-BoldItalic
/FontDescriptor 69 0 R
>>
endobj
34 0 obj
<<
/Type /ExtGState
/SA false
/SM 0.02
/OP false
/op false
/OPM 1
/BG2 /Default
/UCR2 /Default
/TR2 /Default
>>
endobj
35 0 obj
<<
/Type /Page
/Parent 10 0 R
/Resources 37 0 R
/Contents 36 0 R
>>
endobj
36 0 obj
<< /Length 6337 >>
stream
BT
/F2 1 Tf
9 0 0 9 400.2 655.135 Tm
0 0 0 rg
/GS2 gs
0.01 Tc
0 Tw
(701)Tj
/F1 1 Tf
-25.64 0 TD
0.0083 Tc
0.1666 Tw
(Seasonality and tourism demand models)Tj
/F2 1 Tf
11.5 0 0 11.5 56.64 635.455 Tm
0.0012 Tc
0.0014 Tw
[(quarter)32.5(-ahead forecasts, 31 two-quarters-ahead forecasts, 30 three-quarters-ahead)]TJ
0 -1.0435 TD
0.0721 Tw
[(forecasts, 29 four)42.9(-quarters-ahead forecasts and 25 eight-quarters-ahead forecasts)]TJ
T*
0.0011 Tc
0.2786 Tw
[(are generated. W)42.8(ith regard to the SARIMA model, a variety of SARIMA)]TJ
T*
0.0013 Tc
0.1674 Tw
(models with different combinations of the orders of )Tj
/F1 1 Tf
21.36 0 TD
0 Tc
0 Tw
(p)Tj
/F2 1 Tf
0.3861 0 TD
0.0087 Tc
(, )Tj
/F1 1 Tf
0.6678 0 TD
0 Tc
(q)Tj
/F2 1 Tf
0.4487 0 TD
0.0087 Tc
(, )Tj
/F1 1 Tf
0.6678 0 TD
0 Tc
(P)Tj
/F2 1 Tf
0.5635 0 TD
0.0026 Tc
(, )Tj
/F1 1 Tf
0.6678 0 TD
0 Tc
(Q)Tj
/F2 1 Tf
0.6157 0 TD
0.0011 Tc
0.1696 Tw
[( are estimated)]TJ
-25.3774 -1.0435 TD
0.0012 Tc
0.1682 Tw
(first and the optimal model is selected based on such information criteria as)Tj
T*
0.1048 Tw
(the Akaike information criterion \(AIC\) and Schwarz criterion \(SC\). The orders)Tj
T*
-0.0008 Tc
0 Tw
(of )Tj
/F1 1 Tf
1.1374 0 TD
0 Tc
(p)Tj
/F2 1 Tf
0.3861 0 TD
0.0009 Tc
(, )Tj
/F1 1 Tf
0.6678 0 TD
0 Tc
(q)Tj
/F2 1 Tf
0.4383 0 TD
0.0009 Tc
(, )Tj
/F1 1 Tf
0.6574 0 TD
0 Tc
(P)Tj
/F2 1 Tf
0.5635 0 TD
0.0025 Tc
0.1566 Tw
[( and )]TJ
/F1 1 Tf
2.2017 0 TD
0 Tc
0 Tw
(Q)Tj
/F2 1 Tf
0.6157 0 TD
0.0012 Tc
0.1577 Tw
[( are chosen from 0 to 2, according to Pankratz \(1983\), who)]TJ
-6.6678 -1.0435 TD
0.1106 Tw
(states that in practice, all the orders \()Tj
/F1 1 Tf
15.0574 0 TD
0 Tc
0 Tw
(p)Tj
/F2 1 Tf
0.3861 0 TD
0.0026 Tc
(, )Tj
/F1 1 Tf
0.6157 0 TD
0 Tc
(d)Tj
/F2 1 Tf
0.4487 0 TD
0.007 Tc
(, )Tj
/F1 1 Tf
0.6052 0 TD
0 Tc
(q)Tj
/F2 1 Tf
0.4487 0 TD
0.0026 Tc
(, )Tj
/F1 1 Tf
0.6157 0 TD
0 Tc
(P)Tj
/F2 1 Tf
0.553 0 TD
0.007 Tc
(, )Tj
/F1 1 Tf
0.6157 0 TD
0 Tc
(D)Tj
/F2 1 Tf
0.72 0 TD
0.0026 Tc
(, )Tj
/F1 1 Tf
0.6261 0 TD
0 Tc
(Q)Tj
/F2 1 Tf
0.6052 0 TD
0.0014 Tc
0.1113 Tw
(\) tend to be small, often)Tj
-21.2974 -1.0435 TD
0.1813 Tw
[(no more than 1 or 2 \(for SARIMA\).)]TJ
1.0435 -1.0435 TD
0.0243 Tc
0.328 Tw
[(T)82.3(o)0( be consistent with the previous tourism forecasting studies, the)]TJ
-1.0435 -1.0435 TD
0.0011 Tc
0.1449 Tw
[(forecasting)-259.8(accuracy comparison is carried out based on two measures of error)]TJ
T*
0.2283 Tw
(magnitude: the mean absolute percentage error \(MAPE\) and the root mean)Tj
T*
0.0014 Tc
0.093 Tw
[(square percentage error \(RMSPE\) \(for detailed justification see W)53.6(itt and W)64(itt,)]TJ
T*
0.0012 Tc
0.18 Tw
(1992\). According to the two error measures, the forecasting performances of)Tj
T*
0.0959 Tw
[(the alternative models are ranked and the results are shown in T)84.8(ables 2 and)-249.1(3.)]TJ
/F1 1 Tf
0 -2.0765 TD
0.0054 Tw
[(Forecasting accuracy comparison across different destinations.)]TJ
/F2 1 Tf
21.5061 0 TD
0.0016 Tc
0.0053 Tw
(The forecast errors of the)Tj
-21.5061 -1.0435 TD
0.0011 Tc
0.141 Tw
[(alternative models for each destination are presented in T)84.6(a)-0.2(ble 2. It should be)]TJ
T*
0.0013 Tc
0.119 Tw
(noted that STSM collapses to BSM in four out of seven cases \(France, Greece,)Tj
T*
0.0233 Tc
0.3278 Tw
(Italy and the USA\), which suggests that after allowing for stochastic)Tj
T*
0.0074 Tc
[(determination)-253.5(of the various components of the time series, none of the)]TJ
T*
0.0012 Tc
0.0319 Tw
[(explanatory)-259.7(variables is statistically significant. The results of T)84.7(able 2 show that)]TJ
T*
0.0014 Tc
0.0304 Tw
(no single model is superior to other techniques across different destinations. For)Tj
T*
0.2308 Tw
(example, the JML-ECM is ranked top overall in terms of both MAPE and)Tj
T*
0.0011 Tc
0.036 Tw
[(RMSPE in the Australia and Greece models. However)53.3(, the seasonal nave model)]TJ
T*
0.0013 Tc
0.213 Tw
(outperforms all the other models based on the two measures in the case of)Tj
T*
0.1276 Tw
(France. The WB-ECM is the best performer among all competing models for)Tj
T*
0.0601 Tw
[(Italy)84.9(. W)53.6(ith respect to Spain and the USA, the JML-ECM and SARIMA models)]TJ
T*
0.185 Tw
(take alternative turns to be the best based on different error measures.)Tj
1.0435 -1.0435 TD
0.0015 Tc
0.1391 Tw
[(W)51.9(ith respect to the least accurate forecasts, the WB-ECM is outperformed)]TJ
-1.0435 -1.0435 TD
0.0012 Tc
0.0971 Tw
(by its competitors in the Australia and Greece cases. The reduced ADL model)Tj
T*
0.2672 Tw
(performs worst in the French and Spanish models according to both error)Tj
T*
0.1218 Tw
[(measures. In the USA case, the V)84.7(A)10(R)-9.2( )7.8(model exhibits the poorest performance,)]TJ
T*
0.1861 Tw
(while for Italy the TVP model produces the least accurate forecasts.)Tj
1.0435 -1.0435 TD
0.001 Tc
0.1941 Tw
(The information above can help forecasters to decide which model to use)Tj
-1.0435 -1.0435 TD
0.0013 Tc
0.2551 Tw
(when forecasting UK tourist arrivals to specific destinations. Although the)Tj
T*
0.0011 Tc
0.0706 Tw
(forecasting performance of the alternative models varies across destinations, the)Tj
T*
0.0057 Tc
0.328 Tw
(best performing models always treat seasonality as stochastic rather than)Tj
T*
0.0011 Tc
0.1636 Tw
[(deterministic)-259.8(with respect to each destination except France. The implication)]TJ
T*
0.0014 Tc
0.2648 Tw
(is that, although the demand for tourism to different destinations features)Tj
T*
0.0012 Tc
0.06 Tw
(different seasonal patterns, to treat seasonality as stochastic is always favourable)Tj
T*
0.0014 Tc
0.1843 Tw
(as far as forecasting the future demand for tourism is concerned.)Tj
/F1 1 Tf
0 -2.087 TD
0.0015 Tc
0.3149 Tw
(Forecasting accuracy comparison over different forecast horizons. )Tj
/F2 1 Tf
24.3652 0 TD
0.0006 Tc
0.3174 Tw
(At each forecast)Tj
-24.3652 -1.0435 TD
0.0011 Tc
0.0556 Tw
(horizon, aggregated error measures are calculated across seven destinations. The)Tj
T*
0.0015 Tc
0.2912 Tw
[(results are shown in T)85(able 3. It can be observed from the table that for)]TJ
T*
0.0013 Tc
0.1354 Tw
[(1-quarter)32.6(-ahead and 2-quarters-ahead forecasts, JML-ECM outperforms all the)]TJ
T*
0.0612 Tw
(other models in terms of both error measures, followed by the nave model and)Tj
T*
0.2739 Tw
(BSM. For 3-quarters-ahead and 4-quarters-ahead forecasts, the nave model)Tj
ET
endstream
endobj
37 0 obj
<<
/ProcSet [ /PDF /Text ]
/Font << /F1 6 0 R /F2 7 0 R >>
/ExtGState << /GS2 9 0 R >>
>>
endobj
38 0 obj
<<
/Type /Page
/Parent 10 0 R
/Resources 40 0 R
/Contents 39 0 R
>>
endobj
39 0 obj
<< /Length 4161 >>
stream
BT
/F2 1 Tf
9.5 0 0 9.2 203.38 654.895 Tm
0 0 0 rg
/GS2 gs
0 Tc
0 Tw
(T)Tj
8.1 0 0 7.8 209.26 654.895 Tm
0.0025 Tc
(OURISM)Tj
9.5 0 0 9.2 239.5 654.895 Tm
0.1666 Tc
( E)Tj
8.1 0 0 7.8 248.86 654.895 Tm
0.0018 Tc
(CONOMICS)Tj
9 0 0 9 68.02 655.135 Tm
0.01 Tc
(702)Tj
ET
0 0 0 RG
0 J 0 j 0.96 w 10 M []0 d
/GS1 gs
1 i
140.855 59.545 m
140.855 644.896 l
S
0.48 w
158.615 59.545 m
158.615 644.896 l
S
BT
/F3 1 Tf
0 9.4999 -9.4999 0 153.2152 59.5448 Tm
/GS2 gs
0.0061 Tc
-0.0093 Tw
[(T)92(able 2.)-499.2(Comparison of models overall forecasting performance across dif)18.7(ferent destinations.)]TJ
0 -1.9579 TD
0.0053 Tc
0 Tw
[(Destination)-2495.8(Measure)-2041(NAVE)-2445.2(SARIMA)-1914.7(RE-ADL)-1826.3(WB-ECM)-1573.7(JML-ECM)-2584.2(TVP)-3657.9(V)108.3(A)4.8(R)-3948.4(BSM)-2369.4(STSM)]TJ
/F2 1 Tf
0 -2.3242 TD
0.0062 Tc
-0.0099 Tw
[(Australia)-3897(MAPE)-2861.2(4.516\(2\))-2608.5(5.048\(5\))-2318(5.839\(6\))-2318(7.251\(9\))-2191.7(4.130 \(1\))-2052.8(6.307 \(8\))-2065.4(5.035 \(4\))-2494.9(4.710\(3\))-1269.6(6.175\(7\))]TJ
7.4526 -1.1495 TD
0.0059 Tc
0 Tw
[(RMSPE)-2406.7(6.351\(2\))-2608.8(7.089\(5\))-2318.3(7.711\(6\))-2318.3(9.170\(9\))-2318.3(5.266\(1\))-2305.7(7.839\(8\))-2318.3(6.717\(4\))-2621.5(6.478\(3\))-1269.9(7.819\(7\))]TJ
-7.4526 -1.1621 TD
0.0056 Tc
[(Canada)-4655.5(MAPE)-2861.8(3.783\(1\))-2609.1(4.147\(3\))-2318.6(9.365\(7\))-2318.6(4.356\(4\))-2318.6(5.322\(6\))-2053.4(10.032\(8\))-1813.4(10.591\(9\))-2369.1(5.237\(5\))-1270.2(3.909\(2\))]TJ
7.4526 -1.1621 TD
0.0059 Tc
[(RMSPE)-2406.7(4.654\(1\))-2608.8(5.011\(3\))-2065.7(13.112\(9\))-2065.7(5.540\(4\))-2318.3(6.524\(5\))-2053.1(12.011\(7\))-1813.1(12.665\(8\))-2368.8(7.175\(6\))-1269.9(4.745\(2\))]TJ
-7.4526 -1.1495 TD
0.0057 Tc
[(France)-4920.6(MAPE)-2861.7(0.932\(1\))-2609(0.935\(2\))-2318.5(2.152\(8\))-2318.5(1.067\(3\))-2318.5(1.070\(4\))-2305.9(1.498\(6\))-2318.5(1.559\(7\))-2621.7(1.080\(5\))-2078.5(N/A)]TJ
7.4526 -1.1621 TD
0.0059 Tc
[(RMSPE)-2406.7(1.251\(1\))-2608.8(1.544\(3\))-2318.3(3.540\(8\))-2318.3(1.586\(4\))-2318.3(1.599\(5\))-2305.7(2.023\(7\))-2318.3(1.853\(6\))-2621.5(1.534\(2\))-2078.3(N/A)]TJ
-7.4526 -1.1621 TD
0.0056 Tc
[(Greece)-4819.7(MAPE)-2861.8(4.734\(2\))-2609.1(5.136\(3\))-2318.6(7.506\(7\))-2318.6(9.334\(8\))-2318.6(4.581\(1\))-2306(5.252\(4\))-2318.6(5.647\(5\))-2621.8(5.685\(6\))-2078.6(N/A)]TJ
7.4526 -1.1495 TD
0.0058 Tc
[(RMSPE)-2406.8(7.173\(2\))-2608.9(7.756\(4\))-2065.8(11.702\(7\))-1813.2(12.024\(8\))-2065.8(5.727\(1\))-2305.8(7.510\(3\))-2318.4(8.303\(6\))-2621.6(8.081\(5\))-2078.4(N/A)]TJ
-7.4526 -1.1621 TD
0.0059 Tc
[(Italy)-5640.4(MAPE)-2861.5(2.191\(4\))-2608.8(1.963\(3\))-2318.3(2.902\(7\))-2318.3(1.448\(1\))-2318.3(1.818\(2\))-2305.7(3.556\(8\))-2318.3(2.777\(6\))-2621.5(2.221\(5\))-2078.3(N/A)]TJ
7.4526 -1.1621 TD
[(RMSPE)-2406.7(2.783\(5\))-2608.8(2.470\(3\))-2318.3(3.477\(7\))-2318.3(1.925\(1\))-2318.3(2.264\(2\))-2305.7(4.265\(8\))-2318.3(3.460\(6\))-2621.5(2.781\(4\))-2078.3(N/A)]TJ
-7.4526 -1.1495 TD
[(Spain)-5312(MAPE)-2861.5(1.193\(6\))-2608.8(0.833\(1\))-2318.3(1.747\(9\))-2318.3(1.537\(8\))-2318.3(0.923\(3\))-2305.7(1.117\(5\))-2318.3(0.918\(2\))-2621.5(1.087\(4\))-1269.9(1.231\(7\))]TJ
7.4526 -1.1621 TD
[(RMSPE)-2406.7(1.450\(5\))-2608.8(1.213\(2\))-2318.3(2.212\(9\))-2318.3(2.188\(8\))-2318.3(1.156\(1\))-2305.7(1.468\(6\))-2318.3(1.218\(3\))-2621.5(1.362\(4\))-1269.9(1.629\(7\))]TJ
-7.4526 -1.1621 TD
0.0058 Tc
[(U)6.5(S)6.5(A)-5615.3(MAPE)-2861.6(2.261\(4\))-2608.9(1.766\(2\))-2318.4(2.485\(6\))-2318.4(2.010\(3\))-2318.4(1.655\(1\))-2305.8(2.795\(7\))-2318.4(3.939\(8\))-2621.6(2.280\(5\))-2078.4(N/A)]TJ
ET
0.96 w
/GS1 gs
341.972 59.545 m
341.972 644.896 l
S
BT
0 9.4999 -9.4999 0 336.8124 130.3437 Tm
/GS2 gs
0.0059 Tc
[(RMSPE)-2406.7(2.652\(3\))-2608.8(2.241\(1\))-2318.3(3.052\(5\))-2318.3(2.686\(4\))-2318.3(2.262\(2\))-2305.7(3.379\(7\))-2318.3(4.773\(8\))-2621.5(3.080\(6\))-2078.3(N/A)]TJ
/F1 1 Tf
0 8.9999 -8.9999 0 354.8121 59.5448 Tm
0.0052 Tc
(Note)Tj
/F2 1 Tf
1.64 0 TD
0.0081 Tc
-0.0061 Tw
[(:)-501.9(Figures in parentheses indicate the ranks of individual forecasting methods. N/A means the STSM collapses to the BSM in that i)]TJ
50.847 0 TD
0 Tw
(nstance.)Tj
ET
endstream
endobj
40 0 obj
<<
/ProcSet [ /PDF /Text ]
/Font << /F1 6 0 R /F2 7 0 R /F3 8 0 R >>
/ExtGState << /GS1 34 0 R /GS2 9 0 R >>
>>
endobj
41 0 obj
<<
/Type /Page
/Parent 45 0 R
/Resources 43 0 R
/Contents 42 0 R
>>
endobj
42 0 obj
<< /Length 3629 >>
stream
BT
/F2 1 Tf
9 0 0 9 400.2 655.135 Tm
0 0 0 rg
/GS2 gs
0.01 Tc
0 Tw
(703)Tj
/F1 1 Tf
-25.64 0 TD
0.0083 Tc
0.1666 Tw
(Seasonality and tourism demand models)Tj
ET
0 0 0 RG
0 J 0 j 0.96 w 10 M []0 d
/GS1 gs
1 i
131.624 59.425 m
131.624 642.616 l
S
0.48 w
149.264 59.425 m
149.264 642.616 l
S
BT
/F3 1 Tf
0 9.4999 -9.4999 0 143.8641 59.4248 Tm
/GS2 gs
0.0061 Tc
-0.0094 Tw
[(T)79.4(able 3.)-499.2(Comparison of models overall forecasting performance across dif)18.7(ferent forecast horizons.)]TJ
0 -1.9579 TD
0.005 Tc
7.2103 Tw
[( Measure)-2041.3(NAVE)-2445.5(SARIMA)-1915(RE-ADL)-1814(WB-ECM)-1574(JML-ECM)-2597.1(TVP)-3658.2(V)120.7(A)4.5(R)-3961.3(BSM)-2357.1(STSM)]TJ
/F2 1 Tf
0 -2.3242 TD
0.0057 Tc
0 Tw
[(1-quarter)31(-ahead)-1308(MAPE)-2849(2.632\(2\))-2621.7(2.890\(4\))-2318.5(4.195\(8\))-2305.9(3.370\(6\))-2318.5(2.572\(1\))-2318.5(4.484\(9\))-2318.5(3.983\(7\))-2609(2.768\(3\))-1270.1(3.332\(5\))]TJ
7.4653 -1.1621 TD
0.0059 Tc
[(RMSPE)-2394.1(4.138\(2\))-2621.5(4.773\(4\))-2318.3(7.353\(9\))-2305.7(5.664\(4\))-2318.3(3.767\(1\))-2318.3(6.660\(8\))-2318.3(6.211\(7\))-2608.8(4.599\(3\))-1269.9(4.820\(5\))]TJ
-7.4653 -1.1495 TD
0.0058 Tc
[(2-quarters-ahead)-928.9(MAPE)-2848.9(2.608\(2\))-2621.6(2.857\(4\))-2318.4(3.840\(7\))-2305.8(3.642\(6\))-2318.4(2.588\(1\))-2318.4(4.941\(9\))-2318.4(4.142\(8\))-2608.9(2.799\(3\))-1270(3.227\(5\))]TJ
7.4653 -1.1621 TD
0.0059 Tc
[(RMSPE)-2394.1(4.135\(2\))-2621.5(4.724\(4\))-2318.3(6.655\(8\))-2305.7(6.158\(4\))-2318.3(3.776\(1\))-2318.3(7.640\(9\))-2318.3(6.415\(6\))-2608.8(4.456\(3\))-1269.9(4.825\(5\))]TJ
-7.4653 -1.1621 TD
0.0058 Tc
[(3-quarters-ahead)-928.9(MAPE)-2848.9(2.629\(1\))-2621.6(2.898\(4\))-2318.4(4.136\(7\))-2305.8(3.791\(6\))-2318.4(2.804\(2\))-2318.4(4.554\(9\))-2318.4(4.137\(8\))-2608.9(2.886\(3\))-1270(3.574\(5\))]TJ
7.4653 -1.1495 TD
0.0059 Tc
[(RMSPE)-2394.1(4.167\(1\))-2621.5(4.774\(4\))-2318.3(6.538\(8\))-2305.7(6.256\(6\))-2318.3(4.207\(2\))-2318.3(6.666\(9\))-2318.3(6.398\(7\))-2608.8(4.545\(3\))-1269.9(5.114\(5\))]TJ
-7.4653 -1.1621 TD
0.0058 Tc
[(4-quarters-ahead)-928.9(MAPE)-2848.9(2.623\(1\))-2621.6(2.837\(2\))-2318.4(4.511\(9\))-2305.8(3.953\(7\))-2318.4(2.862\(3\))-2318.4(3.217\(5\))-2318.4(4.136\(8\))-2608.9(3.068\(4\))-1270(3.583\(6\))]TJ
7.4653 -1.1621 TD
0.0059 Tc
[(RMSPE)-2394.1(4.152\(1\))-2621.5(4.601\(3\))-2318.3(6.940\(9\))-2305.7(6.350\(8\))-2318.3(4.245\(2\))-2318.3(4.852\(5\))-2318.3(6.207\(7\))-2608.8(4.779\(4\))-1269.9(4.871\(6\))]TJ
-7.4653 -1.1495 TD
0.0058 Tc
[(8-quarters-ahead)-928.9(MAPE)-2848.9(3.515\(3\))-2621.6(2.680\(1\))-2318.4(6.172\(9\))-2305.8(4.531\(5\))-2318.4(3.101\(2\))-2318.4(4.630\(6\))-2318.4(5.362\(8\))-2608.9(4.407\(4\))-954.2(5.1427\(7\))]TJ
7.4653 -1.1621 TD
0.0059 Tc
[(RMSPE)-2394.1(5.039\(3\))-2621.5(4.294\(1\))-2065.7(10.078\(9\))-2053.1(6.987\(7\))-2318.3(4.424\(2\))-2318.3(6.491\(4\))-2318.3(8.073\(8\))-2608.8(6.647\(5\))-1269.9(6.888\(6\))]TJ
-7.4653 -1.1621 TD
0.0058 Tc
[(Overall)-4592.1(MAPE)-2848.9(2.801\(2\))-2621.6(2.833\(3\))-2318.4(4.571\(9\))-2305.8(3.857\(6\))-2318.4(2.786\(1\))-2318.4(4.365\(8\))-2318.4(4.352\(7\))-2608.9(3.186\(4\))-1270(3.772\(5\))]TJ
ET
0.96 w
/GS1 gs
310.662 59.425 m
310.662 642.616 l
S
BT
0 9.4999 -9.4999 0 305.5016 130.3437 Tm
/GS2 gs
0.0059 Tc
[(RMSPE)-2394.1(4.326\(2\))-2621.5(4.633\(3\))-2318.3(7.513\(9\))-2305.7(6.283\(6\))-2318.3(4.084\(1\))-2318.3(6.462\(7\))-2318.3(6.661\(8\))-2608.8(5.005\(4\))-1269.9(5.304\(5\))]TJ
/F1 1 Tf
0 8.9999 -8.9999 0 323.5014 59.4248 Tm
0.0052 Tc
(Note)Tj
/F2 1 Tf
1.6533 0 TD
0.0081 Tc
-0.0056 Tw
[(:)-488.6(Figures in brackets indicate the ranks of individual forecasting methods.)]TJ
ET
endstream
endobj
43 0 obj
<<
/ProcSet [ /PDF /Text ]
/Font << /F1 6 0 R /F2 7 0 R /F3 8 0 R >>
/ExtGState << /GS1 34 0 R /GS2 9 0 R >>
>>
endobj
44 0 obj
<<
/Type /Pages
/Kids [ 10 0 R 45 0 R ]
/Count 16
/MediaBox [ 0 0 482 692 ]
>>
endobj
45 0 obj
<<
/Type /Pages
/Kids [ 41 0 R 46 0 R 49 0 R 52 0 R 55 0 R 58 0 R ]
/Count 6
/Parent 44 0 R
>>
endobj
46 0 obj
<<
/Type /Page
/Parent 45 0 R
/Resources 48 0 R
/Contents 47 0 R
>>
endobj
47 0 obj
<< /Length 5276 >>
stream
BT
/F2 1 Tf
9.5 0 0 9.2 203.38 654.895 Tm
0 0 0 rg
/GS2 gs
0 Tc
0 Tw
(T)Tj
8.1 0 0 7.8 209.26 654.895 Tm
0.0025 Tc
(OURISM)Tj
9.5 0 0 9.2 239.5 654.895 Tm
0.1666 Tc
( E)Tj
8.1 0 0 7.8 248.86 654.895 Tm
0.0018 Tc
(CONOMICS)Tj
9 0 0 9 68.02 655.135 Tm
0.01 Tc
(704)Tj
11.5 0 0 11.5 68.02 635.455 Tm
0.0013 Tc
0.1913 Tw
(generates the most accurate forecasts, followed by JML-ECM. The BSM and)Tj
0 -1.0435 TD
0.1035 Tw
(SARIMA models generate more accurate forecasts than the rest of the models.)Tj
T*
0.183 Tw
(For 8-quarters-ahead forecasts, the SARIMA model turns out to be the top-)Tj
T*
0.0221 Tw
(performing model, followed by JML-ECM and the nave model in terms of both)Tj
T*
0.0164 Tc
0.3278 Tw
[(error measures. The reduced ADL model and the V)89.4(AR model perform)]TJ
T*
0.001 Tc
0.1701 Tw
[(consistently)-259.9(poorly over different forecast horizons and are ranked among the)]TJ
T*
0.0013 Tc
0.0265 Tw
(bottom three. The TVP model produces the least accurate forecasts for the short)Tj
T*
0.0165 Tc
0.3277 Tw
[(forecast horizons \(1-, 2- and 3-quarters-ahead\); however)58.2(, slightly better)]TJ
T*
0.0013 Tc
0.3018 Tw
[(performance)-249.1(is observed for the longer forecast horizons. BSM outperforms)]TJ
T*
0.1991 Tw
(STSM for all forecast horizons, which suggests that inclusion of explanatory)Tj
T*
0.0619 Tw
(variables does not seem to improve the forecast accuracy of the structural time-)Tj
T*
0.0016 Tc
0.1167 Tw
[(series model. This result is in line with the findings of previous studies, such)]TJ
T*
0.0018 Tc
0.1807 Tw
[(as T)95.7(urner and W)54(itt \(2001\).)]TJ
1.0435 -1.0435 TD
0.0015 Tc
0.1865 Tw
(The results highlighted above show that for short-term \(one to two steps)Tj
-1.0435 -1.0435 TD
0.0043 Tc
0.328 Tw
(ahead\) forecasting, JML-ECM performs the best, which indicates that by)Tj
T*
0.0352 Tc
0.3276 Tw
[(incorporating)-257(the short-run dynamics into the econometric modelling)]TJ
T*
0.001 Tc
0.0575 Tw
[(procedure,)-249.4(the forecasting accuracy can be improved. However)63.6(, the nave model)]TJ
T*
0.0015 Tc
0.0862 Tw
(and the BSM still outperform the other models, although they do not perform)Tj
T*
0.1573 Tw
[(as well as the JML-ECM. For medium-term forecasting \(three and four steps)]TJ
T*
0.0011 Tc
0.0355 Tw
(ahead\), the nave model generates the most accurate forecasts, followed by JML-)Tj
T*
0.0013 Tc
0.2723 Tw
(ECM. As far as the long-run forecasts are concerned, the SARIMA model)Tj
T*
0.1122 Tw
(outperforms all of its counterparts, closely followed by the JML-ECM and the)Tj
T*
0.1619 Tw
(nave model. It is clear from the above analysis that the econometric models)Tj
T*
0.0015 Tc
0.2978 Tw
(outperform the time-series models in the short run, while the time-series)Tj
T*
0.0012 Tc
0.185 Tw
(models are more accurate in the medium- to long-run forecasting.)Tj
/F1 1 Tf
0 -2.0765 TD
0.106 Tw
[(Overall forecasting accuracy)32.5(.)]TJ
/F2 1 Tf
10.5183 0 TD
0.1061 Tw
[(The bottom rows of T)84.8(able 3 present the aggregated)]TJ
-10.5183 -1.0435 TD
0.0011 Tc
0.2226 Tw
(accuracy measures across all destinations and over all forecast horizons. The)Tj
T*
0.0015 Tc
0.1597 Tw
(results show that JML-ECM is superior to all other models when forecasting)Tj
T*
0.0014 Tc
0.2152 Tw
(UK outbound tourist arrivals to the seven major destination countries. The)Tj
T*
0.1198 Tw
(results are contradictory to the conclusions drawn in previous studies \(such as)Tj
T*
0.0163 Tc
0.3278 Tw
[(Kulendran and King, 1997; Kulendran and W)47.6(itt, 2001\) in which the)]TJ
T*
0.001 Tc
0.1114 Tw
[(econometric)-259.9(models are outperformed by simple univariate time-series models.)]TJ
T*
0.0012 Tc
0.1766 Tw
(The time-series models perform well and rank second \(the nave model\) and)Tj
T*
0.001 Tc
0.0287 Tw
[(third \(the SARIMA model\), respectively)84.5(. The reduced ADL model generates the)]TJ
T*
0.054 Tw
[(least accurate forecasts, while the TVP model \(or the V)84.6(AR model\) produces the)]TJ
T*
0.0012 Tc
0.187 Tw
(second least accurate forecasts based on either MAPE or RMSPE.)Tj
1.0435 -1.0435 TD
0.0749 Tw
(The superior performances of JML-ECM and the seasonal time-series models)Tj
-1.0435 -1.0435 TD
0.1524 Tw
(might be explained by the way in which the seasonality in the data series is)Tj
T*
0.2953 Tw
(treated. The nature of seasonality in the time series is regarded as either)Tj
T*
0.034 Tw
(deterministic or stochastic. When seasonality is stochastic, the time series needs)Tj
T*
0.1145 Tw
(to be seasonally differenced to account for seasonal unit roots. If seasonality is)Tj
T*
0.0001 Tc
0 Tw
[(found to be deterministic, seasonal dummies co)5.8(uld be used to capture seasonality)]TJ
T*
0.0014 Tc
0.1007 Tw
(in the model. The specifications of JML-ECM and WB-ECM are based on the)Tj
T*
0.0728 Tw
(results of the seasonal unit root test \(HEGY hereby\). The seasonal nave model)Tj
T*
0.0232 Tc
0.3276 Tw
[(and the SARIMA model assume that there are seasonal unit roots at)]TJ
T*
0.0014 Tc
0.2513 Tw
[(seasonal)-259.5(frequencies, while this assumption is consistent with the results of)]TJ
T*
0.0473 Tc
0.3277 Tw
[(the)-255.3(seasonal HEGY test, which indicate that UK outbound tourist)]TJ
T*
0.0034 Tc
[(arrivals)-257.5(exhibit)-247(stochastic seasonality)86.9(. By including stochastic seasonality as)]TJ
ET
endstream
endobj
48 0 obj
<<
/ProcSet [ /PDF /Text ]
/Font << /F1 6 0 R /F2 7 0 R >>
/ExtGState << /GS2 9 0 R >>
>>
endobj
49 0 obj
<<
/Type /Page
/Parent 45 0 R
/Resources 51 0 R
/Contents 50 0 R
>>
endobj
50 0 obj
<< /Length 5062 >>
stream
BT
/F2 1 Tf
9 0 0 9 400.2 655.135 Tm
0 0 0 rg
/GS2 gs
0.01 Tc
0 Tw
(705)Tj
/F1 1 Tf
-25.64 0 TD
0.0083 Tc
0.1666 Tw
(Seasonality and tourism demand models)Tj
/F2 1 Tf
11.5 0 0 11.5 56.64 635.455 Tm
0.0012 Tc
0.3126 Tw
[(o)5.1(n)5.1(e)-270.1(of)-249.2(the main components, BSM and STSM allow for changing seasonal)]TJ
0 -1.0435 TD
0.0471 Tc
0.328 Tw
[(patterns in tourism demand. However)57.5(, the other three models use)]TJ
T*
0.0012 Tc
0.1233 Tw
[(seasonal)-259.7(dummies to account for seasonality in tourism demand. According to)]TJ
T*
0.0525 Tw
(Abeysinghe \(1994\), the use of seasonal dummies in removing seasonality in the)Tj
T*
0.0015 Tc
0.1343 Tw
[(data is likely to produce spurious regressions. Moreover)53.7(, such a simplification)]TJ
T*
0.0012 Tc
0.283 Tw
[(is incapable of reflecting the dynamic)-259.7(nature of the seasonality in tourism)]TJ
T*
0.0857 Tw
(demand. Moosa and Kennedy \(1998\) draw a similar conclusion. That may also)Tj
T*
0.001 Tc
0.0991 Tw
[(explain the relatively poor forecasting performances of the reduced ADL, V)84.5(A)9.8(R)]TJ
T*
0.1826 Tw
(and TVP models.)Tj
1.0435 -1.0435 TD
0.0155 Tw
(It is necessary to examine the statistical significance of the forecast differences)Tj
-1.0435 -1.0435 TD
0.0014 Tc
0.3269 Tw
(between the above two groups of forecasting models, that is, JML-ECM,)Tj
T*
0.177 Tw
(WB-ECM, STSM, BSM and the SARIMA model, which all treat seasonality)Tj
T*
0.0011 Tc
0.2387 Tw
[(as stochastic, and the reduced ADL, V)84.6(AR and TVP models, which assume)]TJ
T*
0.0301 Tc
0.3277 Tw
[(deterministic seasonality)82.3(. The HarveyLeybourneNewbold \(HLN\) test)]TJ
T*
0.0013 Tc
0.1127 Tw
[(proposed)-249.1(by Harvey )]TJ
/F1 1 Tf
8.1704 0 TD
0.0019 Tc
0.1131 Tw
(et al)Tj
/F2 1 Tf
1.6383 0 TD
0.0017 Tc
0.1135 Tw
[( \(1997\) is adopted for this purpose. The HLN test is)]TJ
-9.8087 -1.0435 TD
0.0013 Tc
0.1187 Tw
(a modified version of the Diebold and Mariano \(1995\) test and is suitable for)Tj
T*
0.1723 Tw
(small samples and multiple-step-ahead forecasts. The technical illustration of)Tj
T*
0.0015 Tc
0.0043 Tw
(the statistic is omitted here as it can be found in Harvey )Tj
/F1 1 Tf
21.913 0 TD
0.0004 Tc
(et al)Tj
/F2 1 Tf
1.5443 0 TD
0.0018 Tc
0.0058 Tw
[( \(1997\). The results)]TJ
-23.4574 -1.0435 TD
0.0014 Tc
0.2475 Tw
[(of the tests are reported in T)84.9(ables 46. The evidence clearly suggests that)]TJ
T*
0.0079 Tc
0.3281 Tw
(treating seasonality as a stochastic process can improve forecast accuracy)Tj
T*
0.0016 Tc
0.3244 Tw
[(significantly)85.1(.)-249.3(The JML-ECM, WB-ECM, BSM, STSM and ARIMA models)]TJ
T*
0.0011 Tc
0.1419 Tw
[(outperform the reduced ADL, V)84.6(AR and TVP models significantly in 70% of)]TJ
T*
0.0067 Tc
0.3278 Tw
(the comparisons, with only WB-ECM showing relatively less satisfactory)Tj
T*
0.0024 Tc
0 Tw
(performance.)Tj
1.0435 -1.0435 TD
0.0015 Tc
0.0286 Tw
(A significant contribution of this study is that it has confirmed that different)Tj
-1.0435 -1.0435 TD
0.0012 Tc
0.1134 Tw
(treatments of seasonality can affect the forecasting performance of the tourism)Tj
T*
0.0183 Tw
(demand forecasting models significantly and pre-tests for seasonal unit roots can)Tj
T*
0.1366 Tw
[(improve forecast accuracy)84.7(. This finding is consistent with Alleyne \(2006\) and)]TJ
T*
0.001 Tc
0.1855 Tw
(Diebold and Kilian \(2000\).)Tj
/F3 1 Tf
12 0 0 12 202.44 263.335 Tm
-0.0007 Tc
0.1807 Tw
[( Conclusions)]TJ
/F2 1 Tf
11.5 0 0 11.5 56.64 239.335 Tm
0.0126 Tc
0.328 Tw
(This study has provided a comprehensive comparison of the forecasting)Tj
T*
0.0011 Tc
0.1083 Tw
[(performance)-259.8(of six econometric models and three time-series models related to)]TJ
T*
0.1706 Tw
[(UK outbound tourism demand. The econometric models comprise a reduced)]TJ
T*
0.0016 Tc
0.3116 Tw
[(ADL model, two ECMs \(WB-ECM and JML-ECM\), an unrestricted V)85.1(A)10.4(R)]TJ
T*
0.0013 Tc
0.1402 Tw
(model, a TVP model and an STSM. Five of these models represent the latest)Tj
T*
0.0011 Tc
0.0309 Tw
(developments of econometric modelling and forecasting methods in the tourism)Tj
T*
0.0012 Tc
0.2075 Tw
(context and have shown their advantages over the other models in previous)Tj
T*
0.1691 Tw
(empirical studies. The time-series models include a nave model, a SARIMA)Tj
T*
0.0112 Tc
0.3277 Tw
(model and a BSM. These time-series models have been used widely in)Tj
T*
0.0014 Tc
0.1936 Tw
[(modelling)-249(seasonal time series.)]TJ
1.0435 -1.0435 TD
0.0987 Tw
[(As quarterly data were used in this study)84.9(, three methods were used to treat)]TJ
-1.0435 -1.0435 TD
0.0012 Tc
0.2005 Tw
(seasonality in the econometric modelling process using seasonal dummies,)Tj
T*
-0.0091 Tc
0 Tw
(testing for seasonal roots \(seasonal differencing\) and decomposing the unobservable)Tj
T*
0.0014 Tc
0.1034 Tw
(components. The HEGY unit root test was employed to test for the nature of)Tj
T*
0.0135 Tc
0.3278 Tw
(seasonality in the time series. The forecasting accuracy of the different)Tj
T*
0.0011 Tc
0.143 Tw
[(forecasting)-259.8(techniques was compared based on two error magnitude measures,)]TJ
ET
endstream
endobj
51 0 obj
<<
/ProcSet [ /PDF /Text ]
/Font << /F1 6 0 R /F2 7 0 R /F3 8 0 R >>
/ExtGState << /GS2 9 0 R >>
>>
endobj
52 0 obj
<<
/Type /Page
/Parent 45 0 R
/Resources 54 0 R
/Contents 53 0 R
>>
endobj
53 0 obj
<< /Length 5640 >>
stream
BT
/F2 1 Tf
9.5 0 0 9.2 203.38 654.895 Tm
0 0 0 rg
/GS2 gs
0 Tc
0 Tw
(T)Tj
8.1 0 0 7.8 209.26 654.895 Tm
0.0025 Tc
(OURISM)Tj
9.5 0 0 9.2 239.5 654.895 Tm
0.1666 Tc
( E)Tj
8.1 0 0 7.8 248.86 654.895 Tm
0.0018 Tc
(CONOMICS)Tj
9 0 0 9 68.02 655.135 Tm
0.01 Tc
(706)Tj
ET
0 0 0 RG
0 J 0 j 0.96 w 10 M []0 d
/GS1 gs
1 i
68.02 642.895 m
425.26 642.895 l
S
BT
/F3 1 Tf
9.5 0 0 9.5 68.02 630.655 Tm
/GS2 gs
0.0062 Tc
-0.0412 Tw
[(T)79.5(able 4.)-499.1(T)92.1(ests for equal forecast accuracy between RE-ADL model and other models with)]TJ
ET
0.48 w
/GS1 gs
68.02 614.215 m
425.26 614.215 l
S
BT
9.5 0 0 9.5 68.02 619.615 Tm
/GS2 gs
0.006 Tc
-0.0158 Tw
[(stochastic seasonality)43.9(.)]TJ
0 -1.9579 TD
0.0054 Tc
-0.0151 Tw
[(Forecasting horizon)-1447.2(SARIMA)-2104.1(WB-ECM)-1939.9(JML-ECM)-2407.2(BSM)-3354.6(STSM)]TJ
/F2 1 Tf
0 -2.3242 TD
0 Tw
[(1-quarter)30.8(-ahead)-3695.6(2.250**)-2672.4(1.478)-3695.6(2.734***)-2760.8(2.671***)-1573.4(1.939*)]TJ
0 -1.1495 TD
0.0058 Tc
[(2-quarters-ahead)-3316.3(2.038**)-2672.1(0.827)-3695.3(2.543**)-3265.8(2.398**)-2078.4(1.795*)]TJ
0 -1.1621 TD
[(3-quarters-ahead)-3316.4(2.495**)-2672.2(0.404)-3695.4(3.592***)-2760.6(3.458***)-1573.2(2.452**)]TJ
T*
[(4-quarters-ahead)-3316.3(2.102**)-2672.1(0.792)-3695.3(2.481**)-3265.8(2.525**)-2078.4(1.631)]TJ
ET
0.96 w
/GS1 gs
68.02 529.855 m
425.26 529.855 l
S
BT
9.5 0 0 9.5 68.02 535.015 Tm
/GS2 gs
[(8-quarters-ahead)-3316.4(2.552**)-2672.2(1.906*)-3190.1(2.743***)-2760.6(2.041**)-2078.5(1.437)]TJ
/F1 1 Tf
9 0 0 9 68.02 517.015 Tm
0.0052 Tc
(Note)Tj
/F2 1 Tf
1.6533 0 TD
0.0081 Tc
-0.0076 Tw
[(:)-488.6(*, ** and *** represent 10%, 5% and 1% significant levels, respectively)88.1(.)]TJ
ET
/GS1 gs
68.02 489.415 m
425.26 489.415 l
S
BT
/F3 1 Tf
9.5 0 0 9.5 68.02 477.175 Tm
/GS2 gs
0.0061 Tc
0.0951 Tw
[(T)79.4(able 5.)-499.2(T)92(ests for equal forecast accuracy between TVP model and other models with)]TJ
ET
0.48 w
/GS1 gs
68.02 460.735 m
425.26 460.735 l
S
BT
9.5 0 0 9.5 68.02 466.135 Tm
/GS2 gs
-0.0158 Tw
[(stochastic seasonality)43.9(.)]TJ
0 -1.9705 TD
0.0054 Tc
-0.0151 Tw
[(Forecasting horizon)-1447.2(SARIMA)-2104.1(WB-ECM)-1939.9(JML-ECM)-2407.2(BSM)-3354.6(STSM)]TJ
/F2 1 Tf
0 -2.3116 TD
0 Tw
[(1-quarter)30.8(-ahead)-3695.6(3.945***)-2672.4(1.736*)-2685(5.778***)-3266.1(5.976***)-1068.2(4.909***)]TJ
0 -1.1621 TD
0.0057 Tc
[(2-quarters-ahead)-3316.4(4.338***)-2672.2(2.210**)-2179.6(5.932***)-3265.9(5.414***)-1068(4.851***)]TJ
T*
[(3-quarters-ahead)-3316.4(2.758***)-2672.2(0.450)-3190.1(3.920***)-3265.9(3.760***)-1068(2.915***)]TJ
0 -1.1495 TD
[(4-quarters-ahead)-3316.3(0.289)-3682.6(1.955*)-2684.7(0.591)-4781.6(0.214)-2583.7(0.736)]TJ
ET
0.96 w
/GS1 gs
68.02 376.255 m
425.26 376.255 l
S
BT
9.5 0 0 9.5 68.02 381.415 Tm
/GS2 gs
[(8-quarters-ahead)-3316.3(2.592**)-2672.1(0.421)-3190(2.440**)-3265.8(0.118)-2583.7(0.556)]TJ
/F1 1 Tf
9 0 0 9 68.02 363.415 Tm
0.0052 Tc
(Note)Tj
/F2 1 Tf
1.6533 0 TD
0.0079 Tc
-0.0273 Tw
[(:)-488.8(*, ** and *** represent 10%, 5% and 1% significant levels, respectively)87.9(. The negative statistics)]TJ
-1.6533 -1.1067 TD
0.008 Tc
-0.0063 Tw
(suggest that the TVP model outperforms its competitors.)Tj
ET
/GS1 gs
68.02 325.855 m
425.26 325.855 l
S
BT
/F3 1 Tf
9.5 0 0 9.5 68.02 313.615 Tm
/GS2 gs
0.0062 Tc
0.0952 Tw
[(T)79.5(able 6.)-499.1(T)92.1(ests for equal forecast accuracy between V)107.3(AR model and other models with)]TJ
ET
0.48 w
/GS1 gs
68.02 297.175 m
425.26 297.175 l
S
BT
9.5 0 0 9.5 68.02 302.575 Tm
/GS2 gs
0.006 Tc
-0.0158 Tw
[(stochastic seasonality)43.9(.)]TJ
0 -1.9579 TD
0.0054 Tc
-0.0151 Tw
[(Forecasting horizon)-1447.2(SARIMA)-2104.1(WB-ECM)-1939.9(JML-ECM)-2407.2(BSM)-3354.6(STSM)]TJ
/F2 1 Tf
0 -2.3242 TD
0 Tw
[(1-quarter)30.8(-ahead)-3695.6(2.838***)-2167.1(0.836)-3695.6(4.328***)-2760.8(5.116***)-1573.4(3.975***)]TJ
0 -1.1621 TD
0.0057 Tc
[(2-quarters-ahead)-3316.4(2.750***)-2166.9(0.552)-3695.4(4.342***)-2760.6(4.154***)-1573.2(3.110***)]TJ
0 -1.1495 TD
[(3-quarters-ahead)-3316.4(2.536**)-2672.2(0.169)-3695.4(3.594***)-2760.6(3.799***)-1573.2(2.540**)]TJ
0 -1.1621 TD
[(4-quarters-ahead)-3316.4(2.034**)-2672.2(0.072)-3695.4(2.590**)-3265.9(2.624***)-1573.2(1.857*)]TJ
ET
0.96 w
/GS1 gs
68.02 212.695 m
425.26 212.695 l
S
BT
9.5 0 0 9.5 68.02 217.855 Tm
/GS2 gs
[(8-quarters-ahead)-3316.3(1.939*)-3177.4(0.654)-3695.3(1.916*)-3771(1.569)-3088.9(0.931)]TJ
/F1 1 Tf
9 0 0 9 68.02 199.975 Tm
0.0052 Tc
(Note)Tj
/F2 1 Tf
1.6533 0 TD
0.0081 Tc
-0.0076 Tw
[(:)-488.6(*, ** and *** represent 10%, 5% and 1% significant levels, respectively)88.1(.)]TJ
11.5 0 0 11.5 68.02 167.575 Tm
0.0014 Tc
0.2213 Tw
[(that is, the MAPE and RMSPE. Rankings of these forecasting models over)]TJ
0 -1.0435 TD
0.0012 Tc
0.187 Tw
(different time horizons were provided.)Tj
1.0435 -1.0435 TD
0.2317 Tw
(The empirical results have shown that no single forecasting technique is)Tj
-1.0435 -1.0435 TD
0.0014 Tc
0.0417 Tw
(superior to the others in all situations as the performance of tourism forecasting)Tj
T*
0.1166 Tw
(models tends to vary across destinations and forecast horizons. This finding is)Tj
T*
0.0885 Tw
(in line with previous studies \(for example, Fildes and Lusk, 1984; Makridakis,)Tj
T*
0.0024 Tc
0.1785 Tw
[(1986; W)75.4(ong )]TJ
/F1 1 Tf
5.4991 0 TD
0.0013 Tc
0.1782 Tw
(et al)Tj
/F2 1 Tf
1.7113 0 TD
0.0026 Tc
(, 2007\).)Tj
-6.167 -1.0435 TD
0.0012 Tc
0.0668 Tw
[(W)51.6(ith regard to the overall performance of the models in forecasting seasonal)]TJ
-1.0435 -1.0435 TD
0.0016 Tc
0.2882 Tw
(tourism demand, the results show that JML-ECM is superior to all other)Tj
T*
0.0011 Tc
0.1925 Tw
(models, followed by the simple nave model and the SARIMA model. BSM)Tj
ET
endstream
endobj
54 0 obj
<<
/ProcSet [ /PDF /Text ]
/Font << /F1 6 0 R /F2 7 0 R /F3 8 0 R >>
/ExtGState << /GS1 34 0 R /GS2 9 0 R >>
>>
endobj
55 0 obj
<<
/Type /Page
/Parent 45 0 R
/Resources 57 0 R
/Contents 56 0 R
>>
endobj
56 0 obj
<< /Length 6968 >>
stream
BT
/F2 1 Tf
9 0 0 9 400.2 655.135 Tm
0 0 0 rg
/GS2 gs
0.01 Tc
0 Tw
(707)Tj
/F1 1 Tf
-25.64 0 TD
0.0083 Tc
0.1666 Tw
(Seasonality and tourism demand models)Tj
/F2 1 Tf
11.5 0 0 11.5 56.64 635.455 Tm
0.0013 Tc
0.0947 Tw
(outperforms STSM, suggesting that the inclusion of explanatory variables does)Tj
0 -1.0435 TD
0.1865 Tw
(not seem to improve the forecast accuracy of the basic structural time-series)Tj
T*
0.001 Tc
0.1486 Tw
[(model. The ADLM, V)84.5(AR and TVP models, which include seasonal dummies)]TJ
T*
0.204 Tw
[(to account for seasonality)84.6(, performed relatively poorly as compared with the)]TJ
T*
0.0013 Tc
0.286 Tw
(other models. This empirical evidence suggests that different treatment of)Tj
T*
0.001 Tc
0.016 Tw
(seasonality affects the forecasting performance of the econometric models. When)Tj
T*
0.0012 Tc
0.2597 Tw
(seasonality is regarded as stochastic, the time series needs to be seasonally)Tj
T*
0.3113 Tw
(differenced, or seasonality should be treated as a stochastic component to)Tj
T*
0.192 Tw
[(account for seasonal unit roots in the time series. T)84.8(o)0.1( model seasonality as a)]TJ
T*
0.3024 Tw
(deterministic component by using seasonal dummies may result in model)Tj
T*
0.0014 Tc
0.0226 Tw
(misspecification. Therefore, the test for seasonal unit roots before estimating the)Tj
T*
0.0009 Tc
0.1888 Tw
[(model is necessary and can improve forecast accuracy)84.4(.)]TJ
1.0435 -1.0435 TD
0.0013 Tc
0.1268 Tw
(Although the applications of the TVP model have shown its superiority in)Tj
-1.0435 -1.0435 TD
0.0011 Tc
0.0478 Tw
(tourism demand forecasting as far as annual data are concerned, its performance)Tj
T*
0.2031 Tw
(has been unsatisfactory when quarterly data are used and only deterministic)Tj
T*
0.0012 Tc
0.127 Tw
[(seasonal dummies are introduced. As the time-series models employed in this)]TJ
T*
0.0212 Tc
0.3278 Tw
(study can treat seasonality as a stochastic component using BSM, the)Tj
T*
0.0013 Tc
0.1775 Tw
[(combination)-259.6(of the TVP model and the BSM might improve the forecasting)]TJ
T*
0.0124 Tc
0.3277 Tw
(performance with higher frequency data \(monthly or quarterly\). Further)Tj
T*
0.0013 Tc
0.1582 Tw
[(research)-259.6(in this respect would certainly be of interest to both researchers and)]TJ
T*
0.0024 Tc
0 Tw
(practitioners.)Tj
/F3 1 Tf
13.5861 -1.7844 TD
0.0018 Tc
(Endnotes)Tj
/F2 1 Tf
9 0 0 9 56.64 358.855 Tm
0.008 Tc
0.1945 Tw
[(1)11.3(.)-818.7(The results are omitted here due to space constraints but are available from the authors on)]TJ
1.5733 -1.1067 TD
0.0076 Tc
0 Tw
(request.)Tj
/F3 1 Tf
11.5 0 0 11.5 209.04 325.615 Tm
0.0018 Tc
(References)Tj
/F2 1 Tf
9 0 0 9 56.64 309.055 Tm
0.0023 Tc
[(Abeysinghe, T)95.6(.)0( \(1994\), Deterministic seasonal models and spurious regressions, )]TJ
/F1 1 Tf
31.3333 0 TD
0.0018 Tc
(Journal of Econometrics)Tj
/F2 1 Tf
8.1067 0 TD
0 Tc
(,)Tj
-38.1733 -1.12 TD
0.0078 Tc
0.1629 Tw
[(V)61.5(ol 61, pp 259272.)]TJ
-1.2667 -1.1067 TD
0.008 Tc
0.2342 Tw
(Alleyne, D. \(2006\), Can seasonal unit root testing improve the forecasting accuracy of tourist)Tj
1.2667 -1.1067 TD
0.0077 Tc
0 Tw
(arrivals?, )Tj
/F1 1 Tf
4.1467 0 TD
0.0082 Tc
0.1611 Tw
[(T)99.2(ourism Economics)]TJ
/F2 1 Tf
6.72 0 TD
0.0085 Tc
0.1638 Tw
[(, V)48.5(ol 12, pp 4564.)]TJ
-12.1333 -1.12 TD
0.0082 Tc
0.1428 Tw
[(Diebold, F)114.9(.X., and Kilian, L. \(2000\), Unit root tests are useful for selecting forecasting models,)]TJ
/F1 1 Tf
1.2667 -1.1067 TD
0.1645 Tw
(Journal of Business and Economic Statistics)Tj
/F2 1 Tf
16.1333 0 TD
0.165 Tw
[(, V)48.1(ol 18, pp 265273.)]TJ
-17.4 -1.1067 TD
0.0079 Tc
0.1568 Tw
[(Diebold, F)101.2(.X., and Mariano, R.S. \(1995\), Comparing predictive accuracy, )]TJ
/F1 1 Tf
30.7867 0 TD
0.1574 Tw
(Journal of Business and)Tj
-29.52 -1.12 TD
0.0077 Tc
0.1611 Tw
(Economic Statistics)Tj
/F2 1 Tf
6.7867 0 TD
0.0084 Tc
0.1637 Tw
[(, V)61.7(ol 13, pp 253263.)]TJ
-8.0533 -1.1067 TD
0.008 Tc
0.1672 Tw
(Divisekera, S. \(2003\), A model of demand for international tourism, )Tj
/F1 1 Tf
29.0667 0 TD
0.0078 Tc
0.1685 Tw
[(Annals of T)87.8(ourism Research)]TJ
/F2 1 Tf
10.3733 0 TD
0 Tc
0 Tw
(,)Tj
-38.1733 -1.1067 TD
0.0088 Tc
0.1639 Tw
[(V)62.5(o)-2.5(l)-244.5(30, pp 3149.)]TJ
-1.2667 -1.12 TD
0.008 Tc
0.1315 Tw
(Fildes, R., and Lusk, E.J. \(1984\), The choice of a forecasting model, )Tj
/F1 1 Tf
28.8267 0 TD
0.0065 Tc
0 Tw
(OMEGA)Tj
/F2 1 Tf
3.3733 0 TD
0.0087 Tc
0.1321 Tw
[(, V)62(ol 12, pp 427)]TJ
-30.9333 -1.1067 TD
0.0033 Tc
0 Tw
(435.)Tj
-1.2667 -1.1067 TD
0.008 Tc
0.0819 Tw
[(Franses, P)128(.H. \(1996\), Recent advances in modelling seasonality, )]TJ
/F1 1 Tf
26.2533 0 TD
0.0083 Tc
0.0815 Tw
(Journal of Economic Survey)Tj
/F2 1 Tf
9.8667 0 TD
0.0832 Tw
[(, V)61.6(ol 10,)]TJ
-34.8533 -1.12 TD
0.0079 Tc
0.1611 Tw
(pp 299345.)Tj
-1.2667 -1.1067 TD
0.008 Tc
0.0258 Tw
[(Goh, C., and Law)61.3(, R. \(2002\), Modelling and forecasting tourism demand for arrivals with stochastic)]TJ
1.2667 -1.1067 TD
0.1639 Tw
(non-stationary seasonality and intervention, )Tj
/F1 1 Tf
18.2667 0 TD
0.0085 Tc
0.1611 Tw
[(T)86.2(ourism Management)]TJ
/F2 1 Tf
7.8 0 TD
0 Tc
0 Tw
(,)Tj
/F1 1 Tf
0.2533 0 TD
( )Tj
/F2 1 Tf
0.4267 0 TD
0.0082 Tc
0.1649 Tw
[(V)61.9(ol 23, pp 499510.)]TJ
-28.0133 -1.12 TD
0.0079 Tc
0.1941 Tw
[(Gonzlez, P)114.6(., and Moral, P)127.9(.)-0.1( \(1995\), An analysis of the international tourism demand in Spain,)]TJ
/F1 1 Tf
1.2667 -1.1067 TD
0.008 Tc
0.163 Tw
(International Journal of Forecasting)Tj
/F2 1 Tf
13.3067 0 TD
0.165 Tw
[(, V)61.4(ol 11, pp 233251.)]TJ
-14.5733 -1.1067 TD
0.0078 Tc
0.0648 Tw
[(Gonzlez, P)114.5(., and Moral, P)127.8(.)-0.1( \(1996\), Analysis of tourism trends in Spain, )]TJ
/F1 1 Tf
29.3733 0 TD
0.0083 Tc
0.0666 Tw
[(Annals of T)88.3(ourism Research)]TJ
/F2 1 Tf
10.0667 0 TD
0 Tc
0 Tw
(,)Tj
-38.1733 -1.12 TD
0.0078 Tc
0.1629 Tw
[(V)61.5(ol 23, pp 739754.)]TJ
-1.2667 -1.1067 TD
0.008 Tc
0.0777 Tw
[(Harvey)88(, A.C., and T)101.3(odd, P)114.7(.H.J. \(1983\), Forecasting economic time series with structural and Box)]TJ
1.2667 -1.1067 TD
0.0079 Tc
0.1644 Tw
(Jenkins models: a case study, )Tj
/F1 1 Tf
12.6533 0 TD
0.0081 Tc
0.1645 Tw
(Journal of Business and Economic Statistics)Tj
/F2 1 Tf
16.1467 0 TD
0.165 Tw
[(, V)61.5(ol 1, pp 299307.)]TJ
-30.0667 -1.12 TD
0.0214 Tw
[(Harvey)88(, D., Leybourne, S., and Newbold, P)128(.)0( \(1997\), T)88(esting the equality of prediction mean squared)]TJ
1.2667 -1.1067 TD
0.0073 Tc
0 Tw
(errors, )Tj
/F1 1 Tf
3.12 0 TD
0.0082 Tc
0.1648 Tw
(International Journal of Forecasting)Tj
/F2 1 Tf
13.32 0 TD
0.0084 Tc
0.1637 Tw
[(, V)61.7(ol 13, pp 281291.)]TJ
-17.7067 -1.1067 TD
0.008 Tc
0.3216 Tw
[(Hylleber)8(g, S., Engle, R.F)101.3(., Granger)61.3(, C.W)128(.J., and Y)88(oo, B.S. \(1990\), Seasonal integration and)]TJ
1.2667 -1.12 TD
0.0076 Tc
0 Tw
(cointegration, )Tj
/F1 1 Tf
6.12 0 TD
0.0081 Tc
0.1639 Tw
(Journal of Econometrics)Tj
/F2 1 Tf
8.5733 0 TD
0.0084 Tc
0.1637 Tw
[(, V)61.7(ol 44, pp 215238.)]TJ
ET
endstream
endobj
57 0 obj
<<
/ProcSet [ /PDF /Text ]
/Font << /F1 6 0 R /F2 7 0 R /F3 8 0 R >>
/ExtGState << /GS2 9 0 R >>
>>
endobj
58 0 obj
<<
/Type /Page
/Parent 45 0 R
/Resources 60 0 R
/Contents 59 0 R
>>
endobj
59 0 obj
<< /Length 10795 >>
stream
BT
/F2 1 Tf
9.5 0 0 9.2 203.38 654.895 Tm
0 0 0 rg
/GS2 gs
0 Tc
0 Tw
(T)Tj
8.1 0 0 7.8 209.26 654.895 Tm
0.0025 Tc
(OURISM)Tj
9.5 0 0 9.2 239.5 654.895 Tm
0.1666 Tc
( E)Tj
8.1 0 0 7.8 248.86 654.895 Tm
0.0018 Tc
(CONOMICS)Tj
9 0 0 9 68.02 655.135 Tm
0.01 Tc
(708)Tj
0 -2.04 TD
0.008 Tc
0.0574 Tw
(Kim, J.H. \(1999\), Forecasting monthly tourist departures from Australia, )Tj
/F1 1 Tf
30.12 0 TD
0.0556 Tw
[(T)99.1(ourism Economics)]TJ
/F2 1 Tf
6.6267 0 TD
0.0059 Tc
0.0612 Tw
[(, V)59.2(o)-5.4(l)-247.4(5)-4.1(,)]TJ
-35.48 -1.1067 TD
0.0079 Tc
0.1611 Tw
(pp 277291.)Tj
-1.2667 -1.12 TD
0.008 Tc
0.3019 Tw
(Kim, J.H., and Moosa, I. \(2001\), Seasonal behaviour of monthly international tourist flows:)Tj
1.2667 -1.1067 TD
0.0079 Tc
0.1639 Tw
(specification and implications for forecasting models, )Tj
/F1 1 Tf
22.2933 0 TD
0.1667 Tw
[(T)98.8(ourism Economics)]TJ
/F2 1 Tf
6.7333 0 TD
0.0082 Tc
0.165 Tw
[(, V)61.5(ol 7, pp 381396.)]TJ
-30.2933 -1.1067 TD
0.008 Tc
0.0372 Tw
[( Kulendran, N., and King, M. \(1997\), Forecasting international quarterly tourism flows using error)]TJ
1.2667 -1.12 TD
0.1644 Tw
(correction and time series models, )Tj
/F1 1 Tf
14.6133 0 TD
0.0082 Tc
0.1648 Tw
(International Journal of Forecasting)Tj
/F2 1 Tf
13.3067 0 TD
0.165 Tw
[(, V)48.1(ol 13, pp 319327.)]TJ
-29.1867 -1.1067 TD
0.0078 Tc
0.0377 Tw
[(Kulendran, N., and W)47.8(ilson, K. \(2000\), Modelling business travel, )]TJ
/F1 1 Tf
26.9867 0 TD
0.0083 Tc
0.0389 Tw
[(T)99.3(ourism Economics)]TJ
/F2 1 Tf
6.6 0 TD
0 Tc
0 Tw
(,)Tj
/F1 1 Tf
0.2667 0 TD
( )Tj
/F2 1 Tf
0.2933 0 TD
0.0086 Tc
0.0389 Tw
[(V)62.3(ol 6, pp 47)]TJ
-32.88 -1.1067 TD
0.0083 Tc
0 Tw
(59.)Tj
-1.2667 -1.12 TD
0.0078 Tc
0.2293 Tw
[(Kulendran, N., and W)47.8(itt, S.F)114.5(. \(2001\), Cointegration versus least squares regression, )]TJ
/F1 1 Tf
35.9733 0 TD
0.0088 Tc
0.2334 Tw
(Annals of)Tj
-34.7067 -1.1067 TD
0.0082 Tc
0.1611 Tw
[(T)99.2(ourism Research)]TJ
/F2 1 Tf
6.28 0 TD
0 Tc
0 Tw
(,)Tj
/F1 1 Tf
0.2533 0 TD
( )Tj
/F2 1 Tf
0.4133 0 TD
0.0084 Tc
0.1638 Tw
[(V)62.1(ol 28, pp)]TJ
/F1 1 Tf
4.4933 0 TD
0 Tc
0 Tw
( )Tj
/F2 1 Tf
0.4133 0 TD
0.0071 Tc
(291311.)Tj
-13.12 -1.1067 TD
0.0081 Tc
0.0565 Tw
[(Kulendran, N., and W)48.1(itt, S.F)114.8(. \(2003a\), Forecasting the demand for international business tourism,)]TJ
/F1 1 Tf
1.2667 -1.12 TD
0.1629 Tw
[(Journal of T)88.1(ravel Research)]TJ
/F2 1 Tf
9.9067 0 TD
0 Tc
0 Tw
(,)Tj
/F1 1 Tf
0.2533 0 TD
( )Tj
/F2 1 Tf
0.4267 0 TD
0.0082 Tc
0.1649 Tw
[(V)61.9(ol 41, pp 265271.)]TJ
-11.8533 -1.1067 TD
0.0079 Tc
0.1322 Tw
[(Kulendran, N., and W)47.9(itt, S.F)114.6(. \(2003b\), Leading indicator tourism forecasts, )]TJ
/F1 1 Tf
31.6667 0 TD
0.0085 Tc
0.1334 Tw
[(T)99.5(ourism Management)]TJ
/F2 1 Tf
7.7733 0 TD
0 Tc
0 Tw
(,)Tj
-38.1733 -1.1067 TD
0.0078 Tc
0.1629 Tw
[(V)61.5(ol 24, pp 503510.)]TJ
-1.2667 -1.12 TD
0.008 Tc
0.1565 Tw
[(Kulendran, N., and W)61.3(ong, K.K.F)114.7(. \(2005\), Modelling seasonality in tourism forecasting, )]TJ
/F1 1 Tf
36.9733 0 TD
0.0092 Tc
0 Tw
(Journal)Tj
-35.7067 -1.1067 TD
0.0078 Tc
0.1631 Tw
[(of T)87.8(ravel Research)]TJ
/F2 1 Tf
6.76 0 TD
0.0085 Tc
0.1629 Tw
[(, V)61.8(ol 44, pp)]TJ
/F1 1 Tf
5.1733 0 TD
0 Tc
0 Tw
( )Tj
/F2 1 Tf
0.4267 0 TD
0.0071 Tc
(163170.)Tj
-13.6267 -1.1067 TD
0.0079 Tc
0.291 Tw
[(Li, G., Song, H., and W)47.9(itt, S.F)114.6(. \(2005\), Recent development in econometric modelling and)]TJ
1.2667 -1.12 TD
0.0075 Tc
0 Tw
(forecasting, )Tj
/F1 1 Tf
5.16 0 TD
0.0079 Tc
0.1647 Tw
[(Journal of T)87.9(ravel Research)]TJ
/F2 1 Tf
9.9067 0 TD
0.0085 Tc
0.1638 Tw
[(, V)61.8(ol 44, pp 8299.)]TJ
-16.3333 -1.1067 TD
0.008 Tc
0.0947 Tw
[(Li, G., W)61.3(ong, K.F)101.3(., Song, H., and W)61.3(itt, S.F)114.7(. \(2006\), T)88(ourism demand forecasting: a time varying)]TJ
1.2667 -1.1067 TD
0.0078 Tc
0.1639 Tw
(parameter error correction model, )Tj
/F1 1 Tf
14.2933 0 TD
0.1647 Tw
[(Journal of T)87.9(ravel Research)]TJ
/F2 1 Tf
9.9067 0 TD
0.0081 Tc
0.165 Tw
[(, V)48.1(ol 45, pp 175185.)]TJ
-25.4667 -1.12 TD
0.0079 Tc
0.021 Tw
[(Lim, C., and McAleer)47.9(, M. \(2002\), T)47.9(ime series forecasts of international travel demand for Australia,)]TJ
/F1 1 Tf
1.2667 -1.1067 TD
0.0085 Tc
0.1611 Tw
[(T)99.5(ourism Management)]TJ
/F2 1 Tf
7.7867 0 TD
0 Tc
0 Tw
(,)Tj
/F1 1 Tf
0.2667 0 TD
( )Tj
/F2 1 Tf
0.4133 0 TD
0.0084 Tc
0.1638 Tw
[(V)62.1(ol 23, pp)]TJ
/F1 1 Tf
4.4933 0 TD
0 Tc
0 Tw
( )Tj
/F2 1 Tf
0.4133 0 TD
0.0071 Tc
(389396.)Tj
-14.64 -1.1067 TD
0.008 Tc
0.203 Tw
(Makridakis, S. \(1986\), The art and science of forecasting: an assessment and future directions,)Tj
/F1 1 Tf
1.2667 -1.12 TD
0.163 Tw
(International Journal of Forecasting)Tj
/F2 1 Tf
13.3067 0 TD
0.0075 Tc
0.1652 Tw
[(, V)60.8(ol 2, pp 1529.)]TJ
-14.5733 -1.1067 TD
0.0079 Tc
0.125 Tw
[(Miron, J.A. \(1994\), The economics of seasonal cycles, in Sims, C.A., ed, )]TJ
/F1 1 Tf
30.2667 0 TD
0.0082 Tc
(Advances in Econometrics)Tj
/F2 1 Tf
9.1733 0 TD
0 Tc
0 Tw
(,)Tj
/F1 1 Tf
-38.1733 -1.1067 TD
0.0079 Tc
0.1555 Tw
[(Sixth W)87.9(orld Congress of the Econometric Society)]TJ
/F2 1 Tf
17.72 0 TD
0.008 Tc
0.1556 Tw
(, Cambridge University Press, Cambridge.)Tj
-18.9867 -1.12 TD
0.0076 Tc
-0.0002 Tw
[(Moosa, I.A., and Kennedy)87.6(, P)127.6(.)0( \(1998\), Modelling seasonality in the Australian consumption function,)]TJ
/F1 1 Tf
1.2667 -1.1067 TD
0.0081 Tc
0.1639 Tw
(Australian Economic Papers)Tj
/F2 1 Tf
10.3067 0 TD
0.0084 Tc
0.1638 Tw
[(, V)61.7(ol 37, pp 88102.)]TJ
-11.5733 -1.1067 TD
0.008 Tc
0.1223 Tw
(Osborn, D.R., Harevi, S., and Birchenhall, C.R. \(1999\), Seasonal unit roots and forecasts of two-)Tj
1.2667 -1.12 TD
0.0078 Tc
0.1639 Tw
(digit European industrial production, )Tj
/F1 1 Tf
15.88 0 TD
0.0082 Tc
0.1648 Tw
(International Journal of Forecasting)Tj
/F2 1 Tf
13.3067 0 TD
0.165 Tw
[(, V)48.3(ol 15, pp 2747.)]TJ
-30.4533 -1.1067 TD
0.008 Tc
0.1629 Tw
(Pankratz, A. \(1983\), )Tj
/F1 1 Tf
8.7733 0 TD
0.0082 Tc
0.1625 Tw
(Forecasting with Univariate BoxJenkins Models)Tj
/F2 1 Tf
18.2933 0 TD
0.0075 Tc
0.1649 Tw
[(, W)60.8(iley)87.5(, New Y)100.8(ork.)]TJ
-27.0667 -1.1067 TD
0.008 Tc
0.2985 Tw
(Riddington, G. \(1999\), Forecasting ski demand: comparing learning curve and time varying)Tj
1.2667 -1.12 TD
0.0076 Tc
0.1639 Tw
(parameter approaches, )Tj
/F1 1 Tf
9.48 0 TD
0.008 Tc
0.1638 Tw
(Journal of Forecasting)Tj
/F2 1 Tf
8.1733 0 TD
0.0084 Tc
0.1637 Tw
[(, V)61.7(ol 18, pp 205214.)]TJ
-18.92 -1.1067 TD
0.008 Tc
0.1333 Tw
[(Rodrigues, P)128(.M.M., and Gouveia, P)128(.M.D.C.B. \(2004\), An application of P)48(AR models for tourism)]TJ
1.2667 -1.1067 TD
0.0075 Tc
0 Tw
(forecasting, )Tj
/F1 1 Tf
5.16 0 TD
0.0082 Tc
0.1611 Tw
[(T)99.2(ourism Economics)]TJ
/F2 1 Tf
6.7333 0 TD
0 Tc
0 Tw
(,)Tj
/F1 1 Tf
0.2533 0 TD
( )Tj
/F2 1 Tf
0.4133 0 TD
0.0091 Tc
0.1611 Tw
[(V)49.4(ol 10, pp)]TJ
/F1 1 Tf
4.4933 0 TD
0 Tc
0 Tw
( )Tj
/F2 1 Tf
0.4133 0 TD
0.0071 Tc
(281303.)Tj
-18.7333 -1.12 TD
0.0081 Tc
0.1582 Tw
[(Smeral, E., and Wger)61.4(, M. \(2005\), Does complexity matter? Methods for improving forecasting)]TJ
1.2667 -1.1067 TD
0.1642 Tw
(accuracy in tourism: the case of Australia, )Tj
/F1 1 Tf
17.92 0 TD
0.0079 Tc
0.1647 Tw
[(Journal of T)87.9(ravel Research)]TJ
/F2 1 Tf
9.92 0 TD
0.0081 Tc
0.165 Tw
[(, V)61.4(ol 44, pp 100110.)]TJ
-29.1067 -1.1067 TD
0.0078 Tc
0.2851 Tw
[(Song, H., and W)47.8(itt, S.F)101.1(. \(2000\), )]TJ
/F1 1 Tf
14.56 0 TD
0.0081 Tc
0.2861 Tw
[(T)85.8(ourism Demand Modelling and Forecasting: Modern Econometric)]TJ
-13.2933 -1.12 TD
0 Tw
(Approaches)Tj
/F2 1 Tf
3.9067 0 TD
0.0079 Tc
0.1576 Tw
[(, Per)7.9(gamon, Cambridge.)]TJ
-5.1733 -1.1067 TD
0.3238 Tw
[(Song, H., and W)61.2(ong, K.K.F)114.6(. \(2003\), T)101.2(ourism demand modelling: a time-varying parameter)]TJ
1.2667 -1.1067 TD
0 Tw
(approach,)Tj
/F1 1 Tf
4.24 0 TD
0.1647 Tw
[(Journal of T)87.9(ravel Research)]TJ
/F2 1 Tf
9.92 0 TD
0 Tc
0 Tw
(,)Tj
/F1 1 Tf
0.2533 0 TD
( )Tj
/F2 1 Tf
0.4133 0 TD
0.0081 Tc
0.1666 Tw
[(V)61.8(ol 42, pp 5764.)]TJ
-16.0933 -1.12 TD
0.014 Tc
0.325 Tw
[(Song, H., W)54(itt, S.F)120.7(., and Jensen, T)94(.C. \(2003\), T)107.3(ourism forecasting: accuracy of alternative)]TJ
1.2667 -1.1067 TD
0.0079 Tc
0 Tw
[(econometric)-245.4(models, )]TJ
/F1 1 Tf
8.68 0 TD
0.0082 Tc
0.1648 Tw
(International Journal of Forecasting)Tj
/F2 1 Tf
13.3067 0 TD
0 Tc
0 Tw
(,)Tj
/F1 1 Tf
0.2667 0 TD
( )Tj
/F2 1 Tf
0.4133 0 TD
0.0083 Tc
0.1648 Tw
[(V)48.6(o)0.1(l 19, pp 123141.)]TJ
-23.9333 -1.1067 TD
0.008 Tc
0.1906 Tw
[(T)72.3(rehan, B. \(1986\), Oil prices, exchange rates and the US economy: an empirical investigation,)]TJ
/F1 1 Tf
1.2667 -1.12 TD
0.0085 Tc
0.1555 Tw
(Economic Review)Tj
/F2 1 Tf
6.08 0 TD
0 Tc
0 Tw
(,)Tj
/F1 1 Tf
0.2533 0 TD
( )Tj
/F2 1 Tf
0.4133 0 TD
0.008 Tc
0.1586 Tw
[(Federal Reserve Bank of San Francisco, V)48(ol 4, pp 2543.)]TJ
-8.0133 -1.1067 TD
0.0075 Tc
-0.0001 Tw
[(T)98.5(urner)60.8(, L.W)127.5(., and W)47.5(itt, S.F)114.2(. \(2001\), Forecasting tourism using univariate and multivariate structural)]TJ
1.2667 -1.1067 TD
0.008 Tc
0.163 Tw
(time series models, )Tj
/F1 1 Tf
8.4267 0 TD
0.0078 Tc
0.1667 Tw
[(T)85.5(ourism Economics)]TJ
/F2 1 Tf
6.7333 0 TD
0.0093 Tc
0.1611 Tw
[(, V)62.6(o)-2(l)]TJ
/F1 1 Tf
2.04 0 TD
0 Tc
0 Tw
( )Tj
/F2 1 Tf
0.4267 0 TD
0.0076 Tc
0.1639 Tw
(7, pp 135147.)Tj
-18.8933 -1.12 TD
0.0081 Tc
0.0536 Tw
[(V)48.4(eloce, W)141.4(.)0.1( \(2004\), Forecasting inbound Canadian tourism: an evaluation of error corrections model)]TJ
1.2667 -1.1067 TD
0.0075 Tc
0 Tw
(forecasts, )Tj
/F1 1 Tf
4.2 0 TD
0.0082 Tc
0.1611 Tw
[(T)99.2(ourism Economics)]TJ
/F2 1 Tf
6.72 0 TD
0 Tc
0 Tw
(,)Tj
/F1 1 Tf
0.2667 0 TD
( )Tj
/F2 1 Tf
0.4133 0 TD
0.0078 Tc
0.1629 Tw
[(V)48.1(ol 10, pp 263280.)]TJ
-12.8667 -1.1067 TD
0.1328 Tw
[(V)34.9(u, C.J. \(2006\), Effect of demand volume on forecasting accuracy, )]TJ
/F1 1 Tf
27.92 0 TD
0.0083 Tc
0.1333 Tw
[(T)99.3(ourism Economics)]TJ
/F2 1 Tf
6.7067 0 TD
0 Tc
0 Tw
(,)Tj
/F1 1 Tf
0.2533 0 TD
( )Tj
/F2 1 Tf
0.3867 0 TD
0.0077 Tc
0.1333 Tw
[(V)61.4(ol 12, pp)]TJ
-34 -1.12 TD
0.0071 Tc
0 Tw
(263276.)Tj
-1.2667 -1.1067 TD
0.0081 Tc
0.1834 Tw
[(W)48.1(itt, S.F)114.8(., and W)61.4(itt, C.A. \(1992\), )]TJ
/F1 1 Tf
14.72 0 TD
0.1832 Tw
[(Modelling and Forecasting Demand in T)101.4(ourism)]TJ
/F2 1 Tf
17.7467 0 TD
0.1861 Tw
(, Academic Press,)Tj
-31.2 -1.1067 TD
0 Tw
(London.)Tj
-1.2667 -1.12 TD
0.0626 Tw
[(W)74.7(ong, K.K.F)101.3(., Song, H., W)61.3(itt, S.F)114.7(., and W)74.7(u, D.C. \(2007\), T)88(ourism forecasting: to combine or not)]TJ
1.2667 -1.1067 TD
0.0078 Tc
0.1611 Tw
(to combine?, )Tj
/F1 1 Tf
5.8533 0 TD
0.0081 Tc
0.1667 Tw
[(T)99.1(ourism Management)]TJ
/F2 1 Tf
7.8 0 TD
0 Tc
0 Tw
(,)Tj
/F1 1 Tf
0.2533 0 TD
( )Tj
/F2 1 Tf
0.4133 0 TD
0.0083 Tc
0.1649 Tw
[(V)62(ol 28, pp 10681078)]TJ
/F1 1 Tf
9.4933 0 TD
0 Tc
0 Tw
(.)Tj
ET
endstream
endobj
60 0 obj
<<
/ProcSet [ /PDF /Text ]
/Font << /F1 6 0 R /F2 7 0 R >>
/ExtGState << /GS2 9 0 R >>
>>
endobj
61 0 obj
<<
/Type /FontDescriptor
/Ascent 656
/CapHeight 614
/Descent -220
/Flags 98
/FontBBox [ -184 -254 1029 833 ]
/FontName /HPBNLD+GaramondThree-Italic
/ItalicAngle -13
/StemV 60
/XHeight 394
/StemH 26
/CharSet (/T/o/u/r/i/s/m/space/E/c/n/I/t/e/f/a/p/S/d/comma/U/v/y/L/two/nine/J/K/pe\
riod/hyphen/l/colon/h/at/k/M/g/G/seven/X/H/P/w/R/C/idieresis/dieresis/do\
tlessi/F/A/D/Q/q/B/Z/Y/endash/x/b/O/z/N/W)
/FontFile3 62 0 R
>>
endobj
62 0 obj
<< /Filter /FlateDecode /Length 8790 /Subtype /Type1C >>
stream
HlPw9Ozr.]1g,
I% IT%.rwؽȤ@!5HUKMcV4It&Ӟ?yxy~σ" ^} ojZ+;\ih7*GD