西安交通大学城市学院本科毕业设计（论文）
III. METHODOLOGY
A. Sample Selection
The data used in this study are taken from the Thomson datastream. The sample consists of companies from nine sectors listed on Bursa Malaysia, which includes industrial products, consumer products, trading and services, properties, plantation, construction, technology, infrastructure and hotel. The sample data are collected for a 14 years period from 1993 to 2006. The period for official assessment system tax regime was eight years, i.e 1993-2000. However, the sample data in 1999 were removed since the year 1999 was declared as a waiver year (tax holiday year) for all taxpayers. While, the period for the self assessment system tax regime was six years, i.e 2001 to 2006.
B. Measurement of Effective Tax Rates
This study examines corporate ETRs as a proxy of corporate tax planning. Corporate ETRs basically assesses the tax performance of firms. Thus, it is the best measure to evaluate the actual corporate tax burdens. Previous studies have used various methods for measuring corporate ETRs, where the numerator was the measure of the company?s tax liability and and the denominator was the measure of its income. As for this study, current-based ETR is used. It is defined as a ratio of current income tax expense (total income tax expense minus deferred tax expense) divided by pretax income.
C. Data Filtering and Recoding
In the process of data filtering, companies with a negative pretax income are removed from the sample because a negative income create tax saving. This tax saving will bring down the company?s ETR in different years due to the carry-forward provision provided in the tax laws. Then, data recoding is performed on the calculated ETR. In the study conducted by Rohaya et al. (2008), the ETR is recoded as follows. First, companies with negative tax expenses which produced a negative ETR were recoded as ?0?. Second, companies with an ETR above 100% i.e the companies tax expense exceeded the pretax income were recorded as ?100?. The process of data recoding is necessary since the ETR does not have any economic meaning and can distort the findings in the case when its denominator is zero or negative.
D. Empirical Model and Variable Definitions
The empirical analysis in this study uses the following general multivariate model. The ETR model estimated for current-based ETR is as follows:
ETRt = β0 + β1SIZEt + β2ROAt + β3LEVt + β4CAPINTt + β5INVINTt + β6SECTORDummyt + β7YEARDummyt + β8REGDummyt + εt
ETR refers to current-based ETR. Current-based ETR is measured as current income tax expense (income tax expense minus deferred tax expense) divided by pretax income, β0 is the intercept or constant; β1SIZE is the company size, measured as log of total sales;
30
外文翻译及译文
β2ROA is return on assets, measured as pretax income divided by total assets; β3LEV is the firm leverage, measured as total debts divided by total assets; β4CAPINT is capital intensity, measured as fixed assets (property, plant and equipment) divided by total assets; β5INVINT is inventory intensity, measured as inventory divided by total assets; β6SECTORDummy is the sector dummy for industrial products, consumer products, trading and services, properties, plantation, construction, technology and infrastructure; β7YEARDummy is the year dummy for 1993 to 2006; β8REGDummy is the tax regime dummy for official assessment system regime and self assessment system regime; ε is an error term; t is the firm-years between 1993 to 2006.
31
西安交通大学城市学院本科毕业设计（论文）
IV. EMPIRICAL RESULTS
A. Descriptive Statistics and Univariate Analysis
Table 1 presents descriptive statistics for statutory tax rates (STR) and current-based ETR. During official assessment system (OAS) regime, the mean for STR is 29.5%. While, the mean for current-based ETR is 24%. Further, during self assessment system (SAS) regime, the mean for STR is 28%. While, the mean for current-based ETR is 21.1%. The standard deviations of the current-based ETR during OAS and SAS tax regimes (15.5% and 16.9% respectively) revealed significant variations in the ETR between companies in the sample. As the mean for the current-based ETR was below the mean for STR during both tax regimes, thus it is proven that the Malaysian public listed companies are involved in tax planning activities during the OAS and SAS tax regime. By looking at the divergence between current-based ETR and STR, the results revealed that the divergence was greater during SAS regime, thus companies are involved in tax planning more aggressively during that time as compared to OAS regime.
Figure 1 depicts the mean of the current-based ETR for both tax regimes. Based on the chart, it revealed that the current-based ETR (24.7%) during the OAS regime was higher than the current-based ETR (21.1%) during the SAS regime. Typically, companies with higher ETR paid higher taxes than the companies with lower ETR. In addition, it indicated that companies with a much higher ETR are involved less aggressive in tax planning as compared to the companies with a much lower ETR. Hence, based on the results, it showed that companies undertook more aggressive tax planning during SAS regime than during the OAS regime due to the much lower ETR during the SAS regime. This is because of supreme utilisation of tax incentives by companies during SAS regime.
B. Regression Results
The regression results are presented in Table 2. The result revealed that the current-based ETR was positively associated with size during both the OAS and SAS regimes. It indicated that larger companies faced higher income tax burdens, therefore supporting the political cost theory. The same result was observed for pool sample where in overall, there was a positive association between current-based ETR and size at the 1% level. In addition, there was a negative relationship between current-based ETR and ROA during SAS regime at the 1% level.. The negative relationship indicated that more efficient and highly profitable companies faced lower tax burdens. The possible reason was that profitable companies managed to take advantage from the availability of tax incentives and other tax provisions to lower their ETR. The result for pool sample was also similar where the current-based ETR was negatively related to ROA at the 1% level.
Then, the result from regression analysis showed that current-based ETR was
32
外文翻译及译文
negatively associated with leverage during OAS and SAS regimes. The results indicated that companies with higher leverage had lower ETR. This suggested that levered companies benefited from interest tax deductible expenses which further reduced their taxable income. The similar result was observed for pool sample where in overall, there was a negative association between current-based ETR and leverage at the 1% level. Moreover, as revealed in the Table 4, there was a negative relationship between current-based ETR and capital intensity during OAS and SAS regimes. The negative association revealed that the companies with a larger proportion of fixed assets tended to have lower ETRs. This was due to the tax preferences where the capital-intensive companies benefited from high capital allowance on the fixed assets which led to a lower taxable income. The result for pool sample was similar where the current-based ETR was negatively related to capital intensity at the 1% level.
Lastly, the current-based ETR was positively associated with inventory intensity during the SAS regime at the 5% level. The positive association indicated that the inventory-intensive companies faced higher ETR. This is because, unlike capital-intensive companies which could enjoy tax preferences, the inventory-intensive companies did not have those tax shields. Further, with respect to the result for pool sample, there was a significant and positive relationship between current-based ETR and inventory intensity at the 10% level.
C. Sector Analysis
Figure 2 depicts the level of current-based ETR for each sector during both tax regimes. As can be seen, the industrial products, trading and services, consumer products, plantation, technology and properties sectors reported a much lower current-based ETR during the SAS regime. This result showed that the companies within those sectors were involved in more aggressive tax planning during SAS regime. Hence, it proved that various tax incentives that were offered during SAS regime have given the companies the chances to engage in more aggressive tax planning. While, the other two sectors that were construction and infrastructure sectors reported a much higher current-based ETR during the SAS regime. Thus, it revealed that the companies within the two sectors were involved in less aggressive tax planning during SAS regime. This is may be because of limited tax incentives available to them during SAS tax regime
33
西安交通大学城市学院本科毕业设计（论文）
V. CONCLUSIONS
This study revealed that the Malaysian corporate tax system indeed provided a substantial amount of tax incentives to companies, thus encouraging the companies to engage in aggressive tax planning. Moreover, the difference in ETRs between sectors suggested that the tax incentives only benefited companies within the particular sectors. Therefore, there is an issue of non-neutrality of the corporate tax system. These findings could provide suggestions to tax authorities to undertake tax auditing and investigation to trace illegal tax planning activities. Since this study found that the sampled companies were involved in aggressive tax planning, thus future research should investigate the tax planning strategies undertake by companies , as well as the between ETRs and corporate governance.
34