CHAPTER 6: DATA ANALYSIS AND DISCUSSION
6.5. Empirical results – testing of the direct correlations (H 3 ) and indirect correlations
Table 6.5 shows the results of direct and indirect regression coefficients for all independent variables IC components, using each performance measure (Asset turnover, Investment efficiency, ROE and Tobin q) as the dependent variable.
In terms of asset turnover as dependent variable, only the direct impact of human capital on asset turnover of sample firms is considered. The β-path coefficient of human capital is positive but weak and statistically significant at the low level (β = 0.146, p value
= 0.083). Moreover, the direct relationships of structural capital, relational capital and asset turnover are not significant at 10 percent levels. The results only support for H3a, but none of H3b and H3c when considering the direct correlations with asset turnover. The outcome of human capital is agreed with the recommendations of previous studies such as S.-L.
Chang (2007), Hong Pew et al. (2007). That is, firms in Vietnam seeking to increase productivity through the employment of tangible assets put more effort into utilizing its human resource base (Firer & Mitchell-Williams, 2003). Alternatively, a firm focusing attention on human resource assets appeared to give more emphasis to the effective use of tangible assets due to the tacit knowledge embedded in the mind of employees (Bontis &
Fitz-enz, 2002), who have higher knowledge are contributing into the increase in productivity. However, not all studies support these results. Firer and Mitchell-Williams (2003), Shiu (2006) and Hang-Chan (2009), all find that human capital has a significant negative relationship with asset turnover, showing that the efficiency with which a firm can use its human resources impacts negatively on productivity. In addition, for the relationship between human capital, relational capital and asset turnover, this study discovers that SMA practices variable serves as a full mediator. Such a mediator is indirect- only mediation, alternatively SMA practices fully mediate the human/ relational capital to productivity. The indirect effect of SMA practices in the relationships of human capital, relational capital and asset turnover are statistically significant (t value = 3.882, t value =
4.331, respectively). It means that human capital and relational capital leads to SMA practices, and SMA practices in turn leads to higher productivity measured by asset turnover. However, this study cannot find the direct effect of structural capital and asset turnover or the indirect relation via SMA practices because both direct and indirect β-path do not validate statistically. Finally, to assess the predictive power of asset turnover- related model, adjusted R2 of 0.501 for non-mediated model, 0.574 for mediated model are appropriate (see Appendix 24); i.e. the model for IC components is able to moderately explain over 50 percent of the variance in the dependent variable. This result is a strong indicator that there is a relationship between human capital, relational capital and productivity, thus supporting H5a, H5c but no support for H5b in terms of asset turnover as endogenous construct.
With respect to investment efficiency as dependent variable, only structural capital is positively directly associated with investment efficiency (β = 0.162, p value = 0.000). In relation to human capital and relational capital, the β-path coefficients are not statistically significant (p value = 0.295, p value = 0.140, respectively). The results only support for H3b, and neither are H3a and H3c when considering the direct relationships with investment efficiency. This result is consistent with the author’s prediction to confirm the author’s arguments because this relationship has not been discovered yet. Accordingly, this recommends that it is vital that firms use structural capital representing well-organized structure to facilitate the efficient allocation of capital. Nevertheless, the most intriguing issue is the role of SMA practices to mediate the correlation between IC components and investment efficiency. SMA practices variable plays the role of complementary mediator between structural capital and investment efficiency while it is full mediator between human capital, relational capital and investment efficiency. It means that IC components lead to SMA practices, and SMA practices in turn leads to higher investment efficiency.
Despite no directly influence of human and relational capital over investment efficiency, under the impact of SMA practices, these associations representing by human capital and relational capital are values in statistics (t value = 9.505, t value = 8.571, respectively). In other words, although the employees do not directly impact on the efficient allocation of capital, they could use strategic management accounting techniques to manage investment projects better. Not even all organization’s relations with externals seem to correlate with capital allocation, deeply thinking, managers could apply SMA techniques together with
external information available from outside connections to make more efficient investments or to decrease the risky investments and in turn to gain the state of sustainable investment efficiency. The investment efficiency model has the moderate explanatory power (adjusted R2 is 0.626, see Appendix 25). This result leads to a strong undiscovered conclusion that there is a relationship between IC components and investment efficiency via the mediating role of SMA practices, thus supporting H5a, H5b, H5c in terms of investment efficiency as endogenous construct.
As regards return on equity as dependent variable, this study again confirms the positive direct association of human capital and return on equity, but this coefficient is too weak to generalize as managerial implication (β = 0.074, p value = 0.023). This is consistent with the prediction or the outcomes of previous studies such as Ming-Chin et al.
(2005), Hong Pew et al. (2007), Maditinos et al. (2011) and others. None of the other coefficients representing the remainders of IC components variables is statistically significant in this directly multiple regression; hence these results only support for H3a, but no support for H3b and H3c when considering the direct associations with return on equity.
The direct-effect model reflects that public enterprises rely heavily on human capital to increase their profitability. Accordingly, this study finds human capital as the main component of intellectual capital implementation to contribute into a firm’s profitability because lack of human capital may be caused of ineffective people involvements in business operations, leading to a decrease of profitability. Although the direct β- coefficients only imply the role of human capital and do not affirm the importance of structural and relational capital, the effects of each IC component are confirmed via the mediator of SMA practices to a firm’s profitability. The coefficients on human, structural and relational capital via SMA practices in the return on equity are 0.482, 0.110, 0.619 and significant at the 1%, 5%, 1% level, respectively, thus supporting H5a, H5b, H5c in terms of return on equity as endogenous construct. These results also imply that SMA practices variable roles a partial mediation to human capital but full mediation to structural and relational capital in relation to return on equity. Based on the results above, it has been witnessed that sample firms relied on the implementation of SMA practices to boost the effectiveness of IC capital that are expected to contribute into better future profitability. In addition, the model with return on equity has greater predictive power than in four performance models, shown by the highest value of adjusted R2 (adjusted R2 is 0.870, see
Appendix 26), and therefore IC components are the most suitable measures in explaining profitability, compared with productivity, investment efficiency or firm value. This outcome is consistent with Ming-Chin et al. (2005), where the predictive powers in the return on assets and return on equity models increase higher than the other financial performance measures when VAIC is split into IC components for an investigation.
Finally, studies using IC components have resulted in a mixture of results related to firm value across different countries, industries and years. For instance, Ming-Chin et al.
(2005) conclude that intellectual capital is a driver of both firm value. On the other hand, Firer and Mitchell-Williams (2003), Hang-Chan (2009) conclude that firms and investors place greater importance on physical capital over intellectual capital or Shiu (2006), Hang- Chan (2009) conclude that human capital have a negative relationship with market capitalization. The inconsistent evidence does not result in a compelling conclusion regarding the correlation between IC components and firm value. An investigation with Vietnamese data is therefore undertaken to provide evidence of any relationship between IC components and firm value. However, the direct relations-based results indicate that IC components are positively associated with firm value representing by Tobin q, though positive, neither proposition is statistically significant with results as follows: human capital (β = 0.130, p value = 0.175), structural capital (β = 0.009, p value = 0.881) and relational capital (β = 0.052, p value = 0.479), thus these do not support H3a, H3b and H3c. However, in the indirect model of Tobin q, all three β-path coefficients between each IC component and firm value become substantially positive and highly significant by the mediating role of SMA practices. More particularly, the coefficients on human, structural and relational capital via SMA practices in Tobin q are 0.394, 0.131, 0.498 and significant at the 1%, 5%, 1% level, respectively, thus supporting H5a, H5b, H5c in terms of Tobin q as endogenous construct. SMA practices play the role of full mediator in the relationship between each IC components and firm value. Hereby, this study provides empirical evidence that investors place higher value on firms with better intellectual capital efficiency, and these firms having the characteristics of better implementing SMA practices yield greater IC efficiency. The adjusted R2 value of 0.751 (see Appendix 27) can be described as substantial to assess the appropriateness of predictive power in the research model with Tobin q endogenous variable.
Overall, these results show the conclusions as follows:
The adjusted R2 defines IC components best in explaining the relationship with profitability presenting by the financial indicator of return on equity.
In terms of direct correlation between IC components and corporate performance, there are a few relationships amongst such these relationships, except between structural capital and investment efficiency. Structural capital positively and significantly correlates with investment efficiency because firms are likely to use structural capital representing well-organized structure to facilitate the efficient allocation of capital. Moreover, the direct-effect model reflects that public enterprises rely heavily on human capital to increase their profitability which measured by return on equity.
Under the mediating role of SMA practices, SMA practices fully or partially mediates the positive influence of IC components over corporate performance measured by productivity, investment efficiency, profitability and firm value, except for the effect of structural capital on productivity, i.e. asset turnover.
Firms transform the knowledge (HCE) generated in their dealings with external entities (RCE) to become an element of their own internal knowledge base capable of dissemination throughout the organization (SCE) and this may, in turn, lead to enhanced corporate performance (ATO, INVEFF, ROE, TOBINQ) if managed appropriately by the techniques of strategic management accounting (SMA). This finding supports earlier study by Galbreath and Galvin (2004) who recommend that no single resource can be considered critical in determining corporate performance.
Not only support the combination of different types of resources, this finding but also provides evidence to be in favour of resource-based theory which suggests resources, e.g. physical capital, intellectual capital, are very vital inputs to develop advanced managerial systems e.g. SMA to cater the increasing requirements of sophisticated managerial information (Cleary, 2015), which in turn add more values to a firm.
Table 6.5. Summary of the results of the third and fifth hypothesis testing
Structural path
relationship H3 Expect- ed sign
Direct effect
t value
Signi- ficance
(p <
0.1)?
H5 Expect- ed sign
Indirect effect
(with SMA mediator)
t value
Signi- ficance
(p <
0.1)?
Total
effects t value
Signi- ficance
(p <
0.1)?
Type of media-
tion
Test result
ATO HCE H3a + 0.146 1.735 * H5a + 0.292 3.882 *** 0.438 5.935 *** Partial Support
INVEFF HCE + -0.053 0.295 No + 0.525 9.505 *** 0.472 5.744 *** Full Support
ROE HCE + 0.074 2.284 ** + 0.482 8.850 *** 0.556 13.132 *** Partial Support
TOBINQ HCE + 0.130 1.359 No + 0.394 5.936 *** 0.525 8.980 *** Full Support
ATO SCE H3b + -0.001 0.014 No H5b + 0.073 1.523 No 0.072 0.853 No – No Support
INVEFF SCE + 0.162 3.521 *** + 0.094 2.284 ** 0.255 4.974 *** Partial Support
ROE SCE + 0.031 0.904 No + 0.110 2.385 ** 0.142 2.605 *** Full Support
TOBINQ SCE + 0.009 0.150 No + 0.131 2.448 ** 0.102 3.356 *** Full Support
ATO RCE H3c + 0.019 0.169 No H5c + 0.387 4.331 *** 0.406 5.700 *** Full Support
INVEFF RCE + 0.075 0.140 No + 0.591 8.571 *** 0.666 10.792 *** Full Support
ROE RCE + 0.037 0.480 No + 0.619 12.093 *** 0.656 16.555 *** Full Support
TOBINQ RCE + 0.052 0.708 No + 0.498 8.687 *** 0.550 8.926 *** Full Support
Note: Significant at: *10, **5 and ***1 percent levels (2-tailed)
Source: Calculated by the author in SmartPLS 3.1