Chapter 4: analysis and presentation of results
There is scholarly debate as to whether externally recruited CEOs are a stronger choice for companies that are showing weak performance (Chan, 1996; Daily, Certo, & Dalton, 2000). Studies show that successors promoted from inside the corporation tend to be a cost-efficient choice, and tend to have industry-specific and company-specific skills and knowledge (Zhang & Rajagopalan, 2003).In Vietnam this problem is particularly important, as Vietnamese stock markets are in their early stages of growth (www.hsx.vn; www.hnx.vn, 2012). As expansion of corporations takes place in Vietnam, CEO origin is vital to the growth and shareholder value of corporations.
Turnover in management is in general negatively related to firm performance and those firms with low stock returns often tend to replace their CEOs, yet evidence is not clear whether there is a clear best choice of either hiring internally or recruiting externally (Malatesta & Parrino, 2004).The present quantitative study uses a quasi-experimental correlational design to examine measures of profit margins, price movement, return on assets, return on equity, and net asset values for companies with internally appointed CEOs versus those with externally appointed CEOs.
Descriptive Statistics
Data were gathered from company archived for a total of N = 258 CEOs and measures of shareholder value, profit margins, return on assets, return on equity, and net asset value, for their corresponding company. The sample included a majority of CEOs with bachelor’s degrees (198, 77%). In addition, a large majority were internally appointed (224, 87%), which may have implications for the results, as the two groups in question were not represented to a perfect extent. Demographic features for the sample are presented in Table 1.
Table 1
Frequencies and Percentages for Sample Demographics
Variable n %
Education
No bachelor’s 3 1
Bachelor’s 198 77
Master’s 42 16
Ph.D. 12 5
Origin
Internal 224 87
External 34 13
Note. Due to rounding error, some percentages may not sum to 100%.
Participant ages ranged from 32 to 75 (M = 50.98, SD = 7.21), with equity ranging from 0 years to 31 years (M = 7.71, SD = 4.99). Participants reported an average of 16.20 years of service with the current company (SD = 9.39), and an average of 7.45 years as a CEO at the current company (SD = 4.54). Continuous demographic information is presented in Table 2 below.
Table 2
Sample Demographics for Continuous Descriptive Information
Variable min. max. M SD
Age 32 75 50.98 7.21
CEO equity 0 31 7.71 4.99
Years of service with current company 0 48 16.20 9.39
Years as a CEO at current company 0 31 7.45 4.54
Research Question One
To what extent has company performance differed between companies with internally and externally appointed CEOs as evidenced by measures of shareholder values?
SQ1- To what extent have share price movements (as evidenced by measures of shareholder value) differed between companies with internally and externally appointed CEOs?
SQ2- To what extent have net profit margins (as evidenced by measures of shareholder value) differed between companies with internally and externally appointed CEOs?
SQ3- To what extent have return on assets (as evidenced by measures of shareholder value) differed between companies with internally and externally appointed CEOs?
SQ4-To what extent has return on equity (as evidenced by measures of shareholder value) differed between companies with internally and externally appointed CEOs?
SQ5- To what extent has net asset value (as evidenced by measures of shareholder value) differed between companies with internally and externally appointed CEOs?
To examine research question one and all subsequent questions, a multivariate analysis of variance (MANOVA) was conducted to assess differences in price movement, profit margin, return on assets, return on equity and net asset value by origin of CEO (internally appointed vs. externally appointed). Prior to analysis, the assumption of normality was assessed by conducting Shapiro Wilk tests for each dependent variable. Results of the tests showed significance for price movement (p< .001), profit margin (p< .001), return on assets (p< .001), return on equity (p< .001), and significance for net asset value (p< .001), suggesting that the assumption was not met for any of these variables.
Multivariate normality was assessed by examining Mardia’s test. Mardia’s test was significant for skew, p< .001, and significant for kurtosis, p< .001, and thus multivariate normality was not met. However, Stevens (2009) suggests that non-normality has little effect on Type I error in analyses with large sample sizes (i.e., N > 50). The assumption for equality of variance was assessed with Levene’s test for each dependent variable. For origin of CEO (internally appointed vs. externally appointed), results of the Levene’s test showed no significance for price movement (p = .291), profit margin (p = .021), return on assets (p = .697), return on equity (p = .946), and no significance for net asset value (p = .319), suggesting that the assumption of equal variances was not met for all dependent variables with the exception of profit margin.
