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

Table 2 
Descriptive statistics and independent samples t-test
 International
Domestic
Int’l – Domestic 
Mean S.D. Mean S.D. 
t statistic 
INTL .39974 .348793 0 
0 
10.873*** 
LEV .28799 .150904 
.35114 
.268753 -2.506** 
ROA .02772 .058348 
.01729 
.145999 
0.653 
SIZE 6.69791 1.868311 
4.27302 
2.041269 9.527*** 
RISK 2.8640 1.72144 
1.02155 
2.108451 7.755*** 
Q .0279 
.47641 
-.05461 
.552964 1.217 
PPE .48538 .259835 
.64165 
.292248 -4.291*** 
INTG .10013 .137984 .02788 .058660 
4.767*** 

---

 
2007 Annual International CHRIE Conference & Exposition 
141
Note: 1. ***: significant at 1%; **: significant at 5%; *: significant at 10% 
Before proceeding to the system of equations analysis, Hausman tests were conducted to check the 
endogeneity in order to choose the appropriate analysis procedure. Although internationalization (p-value: 0.1238) 
and leverage (p-value: 0.2802) were not found to be endogenous with ROA, internationalization and leverage 
decisions are indeed endogenous (p-value: 0.0002). Consequently, the 3SLS was employed to analyze the system of 
equations and the results are presented in Table 3. Because the coefficients of leverages terms in equation 1 and the 
square term of internationalization in equation 2 are simultaneously significant, Hypothesis 3 and 4 are supported. 
The inverted U shaped relationship indicates the existence of optimal internationalization, at which the leverage 
level is maximized. Theoretically, it can be explained as that while internationalization could generate coinsurance 
effect (Lewellen, 1971) to allows a higher leverage, potential default costs of creditors also increase as well because 
higher level of non-redeployable intangible assets (i.e. brand equity, operational expertise, and distribution channels) 
accumulate in the process of internationalization (Markides, 1995). Moreover, the reciprocal relationship dictates 
that leverage exerts its influence on internationalization in the same fashion and an optimal leverage for maximum 
internationalization exists. Such results signify the influences of financial decisions and financial resources on 
strategic decisions. 
By examining the effects of internationalization and leverage on profitability in the same model, this study 
discovered a phenomenon that is not reported in extant studies. Internationalization does not directly affect 
profitability given the presence of leverage but is a pure moderator on internationalization’s influence. The above 
conclusion is supported by the results of profitability equation (equation 3), which shows that both 
internationalization terms are not significant but the square term of leverage and its interaction term with 
internationalization are significant. It suggests that business hedging, such as diversifying into foreign markets, 
might not contribute to the profitability if financial control is in place. In other words, controlling financial risk 
through leverage is a more direct and efficient way to improve profitability than controlling business risk through 
diversification. However, because internationalization moderates leverage’s influence on profitability, the financing 
decisions must be made with the consideration of consequences from internationalization. The coefficient of the 
interaction term show that for every unit of increase/decrease of internationalization, the coefficient of leverage 
increases/decreases by 0.199. Leverage would have a greater influence on ROA for hotels of higher level of 
internationalization. Such results indicate that although internationalization does not directly contribute to ROA in 
the presence of leverage, it does interact with leverage and moderate leverage’s effect on ROA.