Microsoft Word 2007 ichrie conference Proceedings Final-Final 06-06-07. doc

 
2007 Annual International CHRIE Conference & Exposition 
147
i
C
w
i
1
≥
,
i = 
1,…,
 n 
and
 
 
 
 
 

(7)
 
i
C
w
i
2
≤
, 
i = 
1,…,
 n
.

 
 
 


(8) 
When varying the expected arrivals of the portfolio
( )
p
r
E
, the algorithm given above will enable us to identify 
the composition of a range of portfolios, from the one with the highest expected arrivals to the minimum variance 
portfolio. This allows tourism marketers to select the level of risk and expected arrivals that are consistent with their 
preferences. For the constraints of each market, as given in Equations 7 and 8, we decided that the weights (
w
i
) of 
individual markets have upper limits and lower limits in a subjective way, as tourism marketers usually do. To allow 
the flexibility of changes in weights for markets, we used the highest possible market weights as upper constraints, 
set by the highest market shares during the last ten years times 1.5 (assuming 50 percent growth in market share), 
whereas we employed the lowest market weights as lower constraints, decided by the lowest market shares during 
the same ten-year period minus 50 percent of the lowest market shares (assuming 50 percent loss). (See upper limits 
and lower limits in Table 4.) Again, the 50 percent growth and loss assumptions for this study are subjective 
decisions by the authors for Taiwan, so other constraints can be applied to different destinations in practice, 
depending upon their unique situations and prospects. In addition, different level of constraints can be applied to 
individual markets, if necessary.