Accounting Department Economic Faculty University of Indonesia Kampus UI Depok address : Tel: +62 (21) ,

Transcription

1 ON INTRODUCTION OF ACCOUNTING STANDARDS ON DERIVATIVES AND OTHER FINANCIAL INSTRUMENTS FRS : CHARACTERISTICS OF THE CORPORATE LOBBYIST HILDA ROSSIETA Accounting Department Economic Faculty University of Indonesia Kampus UI Depok address : hrossieta@fe.ui.ac.id Tel: +62 (21) , , Fax: +62 (0) ,

2 Abstract growing popularity of financial derivatives use in financial risk management has been started parallel with globalisation of economy started in early 1980s. However, despite its growing popularity, financial derivatives also associated with some financial scandals occurred later in the early 1990s. Responding to the circumstances, accounting regulators such as FASB and the International Accounting Standard Committee (the IASC) began introduce accounting standard on derivatives, intended to increase transparency and accountability of financial derivatives used by management. Later on, in the late 1990s, the Accounting Standard Board (the ASB) as the UK s accounting regulators introduced Financial Reporting Standard (FRS ) in September Prior to introduction of FRS, the ASB conducted several hearings to accommodate the public interests. se are: Discussion Paper (the ) in June 1996, and, Financial Report Exposure Draft ( ) in April Obviously, companies which are potentially affected by the new standard will make an effort to lobby the accounting regulators to influence the final standard. aim is either to lessen the harmful effect of the proposed standard on their interest, or, to make the final standard more favourable for their interests (Sutton, 1984). Introduction of accounting standard on derivatives and other financial instruments FRS is expected to have an effect not only on the agency costs (i.e., accounting based covenants in the executive compensation contract and debt contract), but also on corporate use of derivatives in financial risk management. In this respect, hypothetically, corporate lobbyists have characteristics that significantly different to the corporate non-lobbyists. empirical evidence shows that corporate lobbyist have bigger size, higher degree of overseas activities, higher gearing, higher earnings, have more certainty of future positive NPV investment, and, have lower degree of financial distress. Some of the empirical results that turn out to be opposite to the hypothesised suggest that the proposed standard is more consistent with the theory of financial derivatives used. This provides an early indication of the potential effect of the proposed standards on corporate use of financial derivatives in financial risk management practice. 1

3 1. Introduction Globalisation of economy started in early 1980s has force companies to operate in multi currencies. refore, not only that the companies have to face the classic financial risk associated with interest rate movements, but also the risk related to the volatility of foreign currencies. growing share of foreign market, as well as the increasing role of external source of financing, demand companies to practice financial risk management. Presumably, this leads to the growing popularity of derivatives use as a tool in the risk management, and attract researchers attentions to investigate corporate use of financial derivatives across countries (i.e., Bodnar et al, 1995, 1996, 1998, in USA; Berkman et al., 1996, in New Zealand; Mallin et al., 2001, in UK; Marshall, 2000, in UK, USA and Asia Pacific Multinational Companies, Dune et al., 2003). However, besides the growing popularity of financial derivative used, some financial scandals related to financial derivatives beginning to emerge in the early 1990s. At the same time, responding to the circumstances, international accounting regulators such as FASB and the IASC started to introduce accounting standard on derivatives. In the mid 1990s, the ASB as the UK s accounting regulators catch up to the international development by introducing disclosures standard for derivatives FRS. On this, the ASB stated specifically that (direct quotation in italic character) the risk and changes in risk that arise from the use of derivatives are not reported adequately (ASB, Discussion Paper on Derivatives and Other Financial Instruments, July 1996, page 9) Prior to the FRS, the ASB published the Discussion Paper (the ) in June 1996, followed by the publication of in April 1997, to accommodate public interests on the proposed accounting standards on financial derivatives. Having been recognised that 2

4 accounting for derivatives and other financial instruments as a complex subject, hence, needs to be developed gradually over time, yet, at the same time, ASB needs to catch up to the international standards developments, the ASB divide the target of the standard development into long term and short term target. questionnaire includes all the issues needs to be resolved in the long term, while the questionnaire consists of disclosures proposals intended to be resolved in the short term. On the, the ASB stated that (direct quotation in italic character) at a number of places the Board s views are stated. Except for disclosures, these reflect no more than the Board s tentative initial views (ASB, Discussion Paper on Derivatives and Other Financial Instruments, July 1996, page 7) In the longer term, the ASB acknowledge that, if accepted, the proposed accounting standards cause considerable changes to the present accounting practices. Among the dramatic changes are (i) to apply fair valuation to all derivatives and other financial instruments; (ii) to introduce unrealised gains and losses associated with fair valuation, and split the reporting of such gains and losses between STRGL, and, P and L account; (iii) to introduce stringent criteria for hedge accounting, in which deferral are acceptable; (iv) to increase transparency of risk profile by introducing discursive as well as numerical disclosures. By linking the project initiation with the well-published derivatives disasters, the ASB implying that the new accounting standard is expected to increase transparency and accountability of risk management decisions, so that financial derivatives will be used more prudently. This is consistent with the results of content analysis of the as well as the lobbying presented earlier, in which lobbyists suggest that the proposed accounting standards have a potential to influence financial risk managements decisions. 3

5 public hearings provide companies, as one of public members who are directly influenced by the proposed standards, to lobby the standard setters. Previous lobbying studies concentrate on the potential effect of the proposed standards on agency cost such as accounting earnings as well as gearing (i.e., debt to equity ratio) and also, on political cost represented by firm s size (Watts and Zimmerman, 1978; Dhaliwal, 1982; Sutton, 1988; Deakin, 1989; MacArthur and Groves, 1993; Schalow, 1995; Ang et al., 2000; Hill et al., 2002; Georgiou and Roberts, 2004). Never before that introduction of the new standards are expected to influence risk management practice as suggested by the proposed accounting standard on derivatives and other financial instruments. 2. Empirical Construct and Hypotheses Development re has been an expanding literature of corporate lobbying behaviour with a general aim to examine the lobbying motive. Mostly, Sutton s (1984) framework of self-interests economic is assumed as basic motives for lobbying behaviour, including lobbying participation, and also, lobbying arguments (Georgiou, 2004). Accordingly, companies with a potential to suffer negative effect of the proposed standard are more likely to lobby. In this case, as the proposed standards concern with accounting for derivatives, obviously, derivatives user companies are those who are most likely to participate in the lobbying process. Hence, generally, this research project hypothesised that the corporate lobbyists are having specific characteristics which differentiate them from the corporate non-lobbyists. First, companies use derivatives to utilize economies of scale to reduce contracting costs (Dune et al., 2003). Hence, it is hypothesised that H1N: Lobbyist companies have similar size with non-lobbyists companies (SIZEL = SIZENL) H1A: Lobbyist companies have bigger size compared to non-lobbyists companies SIZEL SIZENL) 4

6 This research defines log of total sales (LOGSALES) as the size measure (Watts and Zimmerman, 1986; Dune et al., 2003). Second, derivatives are used to manage foreign currency exposures, most particularly among the UK companies (Joseph and Hewin, 1997; Solomon, 1999; Solomon and Joseph, 2000; Mallin et al., 2001). refore, derivatives users companies are expected to have more significant overseas activities compared to non-derivatives users companies. This leads to the following hypothesis, H2N: Lobbyist companies have similar degree of overseas operation compared to nonlobbyists companies (ROVSLL = ROVSLNL) H2A: Lobbyist companies have higher degree of overseas operation compared to nonlobbyists companies (ROVSLL ROVSLNL) Degree of overseas operation is defined as proportion of overseas sales to total sales (ROVSL) (Berkman and Bradbury, 1996; Dune et al, 2003) Third, companies use derivatives to enable tax and regulatory arbitrage aimed to lower tax payment or lessen financial cost (Berkman and Bradbury, 1996; Dune et al., 2003). For these reasons, companies with higher degree of tax payments and higher financial cost are more likely to use derivatives compared to others. Hence, it is hypothesised that: H3N: Lobbyist companies have similar degree of tax payment and financial costs compared to non-lobbyists companies (TAXARBITL = TAXARBITNL) H3A: Lobbyist companies have higher degree of tax payment and financial costs compared to non-lobbyists companies (TAXARBITL TAXARBITNL) Degree of tax payment and financial cost is defined as ratio of total tax charge to profit before interest and tax (TAXARBIT) in this research project (Dune et al., 2003) Fourth, derivatives are used to enable company to increase certainty of future positive NPV investment, despite its stable cash flow aimed to avoid the possibility of financial distress (Berkman and Bradbury, 1996; Dune et al., 2003). Assuming liquidity as indication of certainty of future positive NPV investment, arguably, derivatives users companies are 5

7 having lower liquidity than others. following hypotheses are drawn from these arguments H4N: Lobbyist companies have similar degree of liquidity compared to non-lobbyists companies (LIQUIDITL = LIQUIDITNL) H4A: Lobbyist companies have lower degree of liquidity compared to non-lobbyists companies (LIQUIDITL LIQUIDITNL) In this research project, ratio of current assets to current liabilities is defined as company s degree of liquidity (LIQUIDIT) (Dune et al., 2003; Berkman and Bradbury, 1996) Fifth, companies use financial derivatives to established stable cash flow and thus avoid the possibility of financial distress (Dune et al, 2003). Ability to cover the interest payments (i.e., interest covered, which is defined as ratio of profit before interest and taxes to interest paid or INTCOV) is often used as indication of financial distress. For these reasons, logically, derivatives users companies have lower degree of interest coverage compared to others. This leads to the following hypotheses H5N: Lobbyist companies have similar degree of financial distress compared to nonlobbyists companies (INTCOVL = INTCOVNL) H5A: Lobbyist companies have lower degree of financial distress compared to nonlobbyists companies (INTCOVL INTCOVNL) Besides the effect on financial risk managements, some of comments provided by noncorporate lobbyists suggest that the proposed standard also have a potential to influence accounting earnings as well as balance sheet accounts. On this, the comments letters stated that (direct quotation in italic character) companies following the proposal will experience volatile earnings (or changes in capital) ( British Bankers Associations, the responses, 4 November 1996). Previous lobbying studies often link introduction of the new accounting standards with the potential effect on accounting earnings due to the role of earnings in the executive compensation plan. If the proposed standards have an effect on earnings volatility, on which managers are rewarded accordingly, hence, firms with lower earnings will be more 6

8 threatened by the proposed standards, as they have inadequate cushions to overcome the earnings volatility effect. refore, it is hypothesised that H6N: Lobbyist companies have similar level of earnings compared to non-lobbyists companies (ROAL = ROANL) H6A: Lobbyist companies have lower level of earnings compared to non-lobbyists companies (ROAL ROANL) This research project defines ratio of net income to total assets (ROA) as measures for earnings level (Cosh, 1975; Dedman, 2003). One of the comments letter imply that the proposed standards have a potential to increase balance sheet volatility, as articulated in the following statement (direct quotation in italic character) manufacturing (or other non-financial company) should show a fall in the book value of fixed rate debt,.potentially causing a breach of lending covenants where these are related to audited financial statements (London Investment Banking Associations, the responses, 4 November 1996). If the proposed standards have a potential to increase balance sheet volatility, and, accounting numbers presented in the balance sheet are normally used in the lending covenants, hence, companies with lower level of accounting numbers have a higher risk to breach the lending covenants under the proposed standards. In the UK, gearing (i.e., debt to equity ratio or DERATIO) is one of the accounting covenants that widely used in debt contract (Citron, 1992; Day and Taylor, 1994). This leads to the hypotheses below H7N: H7A: Lobbyist companies have similar level gearing compared to non-lobbyists companies (DERATIOL = DERATIONL) Lobbyist companies have higher level gearing compared to non-lobbyists companies (DERATIOL DERATIONL) 7

9 3. Data Description sampling frame is drawn from UK companies as of 1995 and 1996, prior to the ASB invitation to comments on the in July 1996 and in June time frame chosen (i.e and 1996) is consistent with the main objectives of the research, which aimed to examine the lobbying motive of corporate lobbyists on the new accounting standard. fact that some private companies are found among the corporate lobbyists is interesting to note, considering that the agency cost is not as relevant for the private companies. To some extent, this confirm the assumption that most of the corporate lobbyists are derivatives user firms, which concerns with the potential effect of the proposed standards on accounting covenant in their debt contract. re are 35 (thirty five) companies participate in the lobbying process, which consists of 8 (eight) private companies and 27 (twenty seven) non-financial public companies. public companies represents 24 (twenty four) industry types of the primary UK SIC Out of these, 9 (nine) companies participated in lobbying on both the as well as the. research used FAME database as the main data source, considering its specialisation in the UK and Ireland companies of any legal forms, including private and public companies. To maintain accurate identification of the lobbyists in the FAME database, the research used eight digit unique registration numbers assigned for each individual company to retrieve the relevant information. se numbers appear in each company s response letters submitted. corporate non-lobbyists are defined as the public companies with similar primary UK SIC Code. Some relevant information regarding corporate lobbyists and their peer corporate non-lobbyists are presented in Appendix 1. To increase research validity, around 340 companies belong to 7011 industry type are randomly divided and distributed into 4 (four) data panels. After considering the missing data, the depictions of sample size for the lobbying as well as lobbying are reported in Appendix 2. 8

10 Examination of descriptive statistics of the tested variables aims to investigate the normality conditions of the data, so that the type of empirical test suitable for the data (i.e., parametric or non-parametric test) can be decided. following table presents descriptive statistics of independent variables of Panel ALL for the lobbying data. Table 1 Descriptive Statistic of Independent Variables of the Lobbying Model Panel ALL (i.e., Whole Sample) PART. STATISTIC RO- TAX- LIQUI INT- SIZE VSL ARBIT -DIT COV ROA DERAT Non-Lobb. Median N: 587 Mean (96.23% of Std. Err. of Total N) Mean Std. Deviation Skewness Std. Err. Of Skwss % of Non- Lobb. N 68% 68% 75% 79% 65% 80% 80% Lobb. Median N: 23 Mean (3.77% of Std. Err. of Total N) Mean Std. Deviation Skewness Std. Err. Of Skwss % of Lobb. N 83% 83% 83% 87% 74% 87% 87% Total Median N: 610 Mean (100%) Std. Err. of Mean Std. Deviation Skewness Std. Err. Of Skwss % of Total N 69% 69% 75% 80% 65% 80% 80% Valid N (list-wise): 372 (61%) Of all the 611 companies participate in the lobbying (i.e., 24 lobbyists companies and 587 non lobbyists companies), only 372 companies or 60.98% have a complete data or valid to be 9

11 included in the empirical testing. Individually, completeness of the corporate lobbyists data are ranged from 74% (i.e., for INTCOV) to 87% (i.e., ROA, LIQUIDIT and DERAT), while data completeness for the corporate non-lobbyists is relatively lower, ranged from 65% (i.e., INTCOV) to 80% (i.e., ROA and DERAT). When looking at the total sample of lobbyists and non-lobbyists, some of the independent variables have a relatively skewed distribution (i.e., ROA, LIQUIDIT and DERAT). Although the skewness of non-lobbyists data are consistent with the skewness of the whole sample, yet, the skewness of the lobbyist s data are not always consistent. Due to the highly skewed distribution, this research project applies non-parametric as well as parametric empirical tests. As for lobbying, descriptive statistics of the independent variables of Panel ALL are presented in the table below. Table 2 Descriptive Statistic of Independent Variables Lobbying Model Panel ALL (i.e., the Whole Sample) PART. Non- Lobb. N: 637 (96.66% of Total N) Lobb. N: 22 (3.34% of Total N) STATISTIC RO- TAXA LIQU INT- SIZE VSL RBIT IDIT COV ROA DERAT Median Mean Std. Err. of Mean Std. Deviation Skewness Std. Err. Of Skwss % of Non- Lobb. N 82% 82% 84% 98% 76% 98% 96% Median Mean Std. Err. of Mean Std. Deviation Skewness Std. Err. Of Skwss % of Lobb. N 91% 91% 91% 95% 91% 91% 91% 10

12 Table 2 (Continued) Descriptive Statistic of Independent Variables Lobbying Model Panel ALL (i.e., the Whole Sample) PART. STATISTIC RO- TAXA LIQU INT- SIZE VSL RBIT IDIT COV ROA DERAT Total Median N: 659 Mean (100%) Std. Err. of Mean Std. Deviation Skewness Std. Err. Of Skwss % of Total N 82% 82% 84% 98% 77% 98% 95% Valid N (list-wise): 456 (69% of Total N) lobbying data available for the empirical tests is slightly higher than the lobbying, which is 456 companies or 69 % of the total sample. Except for IINTCOV, proportion of corporate lobbyists data available for empirical test is above 90%, since only 22 companies participate in the lobbying. As for the non lobbyists, data availability ranged from 76% (i.e., for INTCOV) to 98% (i.e., for ROA). Almost all of the independent variables in lobbying have a highly skewed distribution, as suggested by the skewness statistic, except for SIZE. Accordingly, non-parametric as well as parametric test is applied. Further investigation of the incomplete corporate lobbyists data for the 1995 and 1996 reveals that: (1) overseas sales variables are not consistently available in the FAME database for several big established PLC s such as Zeneca, Grand Metropolitan and Tate & Lyle. Some companies, such as Shell International Ltd and Unilever PLC do not present overseas sales separately with the domestic sales. Instead they present the single total sales figures. This might relate to its status as a holding company with its global operations (i.e., Shell International Ltd, Unilever PLC). 11

13 (2) Some lobbyist companies facing financial difficulties (i.e., Syspas Ltd dissolved in 1996 shortly after the published) or under receivership/ administration (i.e., GKN and TXU Europe Group PLC) around the time when the proposed derivatives standard was about to be introduced. 4. Research Methodology and Model Specification In testing the hypotheses, this research project use Statistical Product and Service Solution (formerly Statistical Package for the Social Sciences, or well known as SPSS) computer program released for Windows. SPSS is a popular software used among researchers, hence, supported by abundant references. software is continually upgraded, so as the associated literatures. study hypothesised that corporate lobbyists are derivatives users companies, hence, are having particular characteristics that differ to the corporate non-lobbyists (i.e., H1, H3, H5, H7, H9, and, H11). Parametric test (i.e., one way ANOVA) as well as non-parametric test (i.e., Kruskal Wallis Test) is used to examine the hypothesised characteristics. null hypothesis assumes no significant differences between the independent variables of corporate lobbyists and of corporate non-lobbyists (μ 1 = μ 2 =μ 3 =. μ k ). Corporate lobbyist and corporate non-lobbyists are indicated by binary number of 1 (one) for corporate lobbyists and 0 (zero) otherwise. Several basic assumptions for the means comparison tests, both parametric and nonparametric, among others, are: (1) Homogeneity of variance of the tested population. (2) Random sample selection. 12

14 5. Results and Analysis of Empirical Investigations empirical investigations are aimed to examine as to whether corporate lobbyists have specific characteristics compared to non-lobbyists. hypotheses are developed based on the agency theory (i.e., debt covenant hypothesis and executive compensation hypothesis), and also, the theory of corporate use of financial derivatives. proxies for the tested variables are determined based on significant findings of some of the previous studies in lobbying on accounting standards (i.e., Watts and Zimmerman, 1978; Dhaliwal, 1982; Deakin, 1989; Hill et. all., 2002; Georgiou and Roberts, 2004; Georgiou, 2005) and corporate used of financial derivatives (Berkman and Bradbury, 1996; Dune et al, 2003). results of comparisons of the central tendency statistics (i.e., skewness, median and mean) between lobbyists and non-lobbyists are presented in the table below. Table 3 Comparisons of Central Tendency Statistics of Data Panels (i.e., Whole Sample, and, Panel A to Panel D) between the Lobbyists (L) and Non-Lobbyists (NL) Comparisons Results SIZE ROVSL TAX- ARBIT LIQUI- DIT INT- COV ROA DE- RATIO Lobbying Skewness* L>NL 100% 0% 0% 0% 80% 0% 0% L<NL 0% 100% 100% 100% 20% 100% 100% Median L>NL 100% 100% 80% 100% 100% 100% 100% L<NL 0% 0% 20% 0% 0% 0% 0% Mean L>NL 100% 100% 80% 100% 0% 100% 40% L<NL 0% 0% 20% 0% 100% 0% 60% Lobbying Skewness* L>NL 20% 0% 80% 0% 0% 0% 0% L<NL 80% 100% 20% 100% 100% 100% 100% Median L>NL 100% 100% 100% 100% 100% 100% 100% L<NL 0% 0% 0% 0% 0% 0% 0% Mean L>NL 100% 100% 80% 20% 0% 0% 100% L<NL 0% 0% 20% 80% 100% 100% 0% *Comparison of skewness statistic is based on the absolute value, ignoring the vector of (+) or (-)

15 table above works as follow. Take for example, the skewness of SIZE variable. Examinations of the skewness statistic of all data panels (i.e., Panel A, B, C, and, D) of the lobbying suggest that all lobbyists data are more skewed than non-lobbyists data. Hence, L>NL is 100%. In contrast, skewness statistic for lobbying shows that 4 (four) out of 5 (five) comparisons (i.e., 80%) suggests that lobbyists data are less skewed compared to non lobbyists, hence, L<NL is 80%. Except for skewness statistic of several variables (i.e., SIZE, TAXARBIT, and INTCOV), in general, the central tendency comparisons results of data panels in the lobbying are consistent with the comparison results of lobbying. To allow for mean comparison test using both parametric and non-parametric test, the data variances are assumed to be homogeneous. results of homogeneity of variance test are presented in the following table. Table 4 Results of Homogeneity of Variances Test VARIABLES Lobbying Lobbying Levene Statistic Sig. Levene Statistic Sig. Panel ALL SIZE ROVSL ** TAXARBIT LIQUIDIT INTCOV ROA DERATIO PANEL A SIZE ROVSL ** TAXARBIT LIQUIDIT * INTCOV ROA DERATIO

16 Table 4 (Continued) Results of Homogeneity of Variances Test VARIABLES Lobbying Lobbying Levene Statistic Sig. Levene Statistic Sig. PANEL B SIZE ROVSL ** TAXARBIT LIQUIDIT INTCOV ROA DERATIO PANEL C SIZE ROVSL ** TAXARBIT ** LIQUIDIT ** INTCOV ROA DERATIO SIZE ROVSL ** TAXARBIT ** LIQUIDIT INTCOV ROA DERATIO ** Significant at less than the 1% level * Significant at less than the 5% level For the lobbying, the homogeneity test shows consistently across data panel that ROVSL is not homogeneous (i.e., significant at less than 1% level). Hence, the means comparisons test for ROVSL is inapplicable, due to the assumption violation. alternative way to interpret the homogeneity test results is that in term of ROVSL, lobbyists and non-lobbyists are two different kind of sample, so that further test of means comparison is unnecessary. In other words, when the homogeneity tests fail to reject HN (i.e., which stated that the variances of lobbyists and non-lobbyists are equal), the means comparison test will obviously deliver the same result (i.e., lobbyists and non-lobbyists are significantly different). Some indications of difference are also found in LIQUIDIT (i.e., panel A significant at less than 5% level, and, panel C significant at less than 1% level). 15

17 With regards to lobbying, the homogeneity test results suggests that TAXARBIT variance between lobbyists and non-lobbyists is significantly different at less than 5% level in panel C and D. Hence, the test results provide strong indications regarding the difference between lobbyists and non-lobbyists in term of tax and financial cost arbitrage. Empirical examinations of the central tendency differences between lobbyist and non lobbyists are conducted using both parametric test (i.e., ANOVA) and non-parametric test (i.e., Kruskal Wallis). following table presents the results of both parametric and nonparametric test, together with the hypothesised directions and the directions suggested by the empirical results. 16

18 Table 5 Results of Parametric Test (ANOVA) and Non-Parametric Test (Kruskal Wallis) (Two Tailed Test) of the Characteristics of Corporate Lobbyists and Corporate Non-Lobbyists SIZE ROVSL TAXARBIT LIQUIDIT INTCOV ROA DERAT Hypotheses H1: L>NL H2: L>NL H3: L>NL H4: L<NL H5: L<NL H6: L<NL H7: L>NL Results Mean L>NL L>NL L>NL L>NL L>NL L>NL L>NL L<NL L<NL L<NL L>NL L<NL L<NL L>NL Median L>NL L>NL L>NL L>NL L>NL L>NL L>NL L>NL L>NL L>NL L>NL L>NL L>NL L>NL Panel ALL Parametric Test -ANOVA F Sig. 0.00** 0.00** 0.00** Non-Parametric Test - Kruskal Wallis Chi-Square Asymp. Sig. 0.00** 0.00** 0.00** 0.00** ** * 0.01** 0.00** 0.00** 0.00** 0.06* 0.02** PANEL A Parametric Test - ANOVA F Sig. 0.00** 0.00** 0.00** 0.00** Non-Parametric Test - Kruskal Wallis Chi-Square Asymp. Sig. 0.00** 0.00** 0.00** 0.00** * 0.10* * 0.03* 0.04* 0.04* ** Significant at less than or equal to the 1% level, one tailed test * Significant at less than or equal to the 5% level, one tailed test 17

19 Table 5 (Continued) Results of Parametric Test (ANOVA) and Non-Parametric Test (Kruskal Wallis) (Two Tailed Test) of the Characteristics of Corporate Lobbyists and Corporate Non-Lobbyists SIZE ROVSL TAXARBIT LIQUIDIT INTCOV ROA DERAT Hypotheses H1: L>NL H2: L>NL H3: L>NL H4: L<NL H5: L<NL H6: L<NL H7: L>NL Results Mean L>NL L>NL L>NL L>NL L>NL L>NL L>NL L<NL L<NL L<NL L>NL L<NL L<NL L>NL Median L>NL L>NL L>NL L>NL L>NL L>NL L>NL L>NL L>NL L>NL L>NL L>NL L>NL L>NL PANEL B Parametric Test - ANOVA F Sig. 0.00** 0.00** 0.00** 0.00** Non-Parametric Test - Kruskal Wallis Chi-Square Asymp. Sig. 0.00** 0.00** 0.00** 0.00** * 0.09* ** 0.02** 0.02** 0.02** PANEL C Parametric Test - ANOVA F Sig. 0.00** 0.00** 0.00** 0.00** ** 0.00** Non-Parametric Test - Kruskal Wallis Chi-Square Asymp. Sig. 0.00** 0.00** 0.00** 0.00** * 0.06* ** 0.01** 0.01** 0.01** ** Significant at less than or equal to the 1% level, one tailed test * Significant at less than or equal to 5the % level, one tailed test 18

20 Table 5 (Continued) Results of Parametric Test (ANOVA) and Non-Parametric Test (Kruskal Wallis) (Two Tailed Test) of the Characteristics of Corporate Lobbyists and Corporate Non-Lobbyists SIZE ROVSL TAXARBIT LIQUIDIT INTCOV ROA DERAT Hypotheses H1: L>NL H2: L>NL H3: L>NL H4: L<NL H5: L<NL H6: L<NL H7: L>NL Results Mean L>NL L>NL L>NL L>NL L>NL L>NL L>NL L<NL L<NL L<NL L>NL L<NL L<NL L>NL Median L>NL L>NL L>NL L>NL L>NL L>NL L>NL L>NL L>NL L>NL L>NL L>NL L>NL L>NL PANEL D Parametric Test - ANOVA F Sig. 0.00** 0.00** 0.00** 0.00** Non-Parametric Test - Kruskal Wallis Chi-Square Asymp. Sig. 0.00** 0.00** 0.00** 0.00** * 0.10* ** 0.01** 0.00** 0.00** ** Significant at less than or equal to the 1% level, one tailed test * Significant at less than or equal to the 5% level, one tailed test 19

21 SPSS output presented in the table above is for two tailed test, while the hypotheses assumed one tailed test. Hence, when the test result suggests 10% significant level, it means that in one tailed test, it is significant at the 5% level. table works as follows. Take, for example, SIZE as one of the tested characteristic. Hypothesis 1 (H1) stated that corporate lobbyists have a bigger size compared to corporate non-lobbyists (i.e., L>NL). result confirms that L>NL. Also, the results of both parametric and non-parametric tests suggest that the SIZE of corporate lobbyist and corporate non-lobbyists is significantly different at less than 1% level, in both the lobbying as well as lobbying. Overall, empirical evidences of the lobbying as well as lobbying, using both parametric and non-parametric test, strongly suggest that corporate lobbyists have bigger SIZE than non-lobbyists (H1: L>NL, significant at less than 1% level). In addition, consistent with the results of homogeneity variance test, corporate lobbyists have higher degree of overseas operation ROVSL (H2: L>NL, significant at less than 1% level) compared to corporate nonlobbyists. refore, H1N is rejected in favour of H1A, and also, H2N is rejected in favour of H2A. In addition, although the results of parametric tests provide no evidence, the nonparametric test results indicate that corporate lobbyists have higher DERAT compared to corporate non-lobbyists (i.e., H7: L>NL, significant at less than 1% and 5% level). In this case, non-parametric result is more suitable for data testing, considering that DERAT is a highly skewed variable (i.e., across data panel in both the and lobbying are ranged from to for corporate lobbyists, and, ranged from to for corporate nonlobbyists). For this reason, HN is rejected in favour of H7A. Some inconsistencies between the hypothesized directions and the empirical results are found lobbying. inconsistencies are found in the following variables: (i) LIQUIDIT (i.e., H4: L<NL, while the mean and median statistics suggest L>NL); (ii) INTCOV (i.e., H5: L<NL, median statistic suggest L>NL); and, (iii) ROA (i.e., H6: L<NL, while median statistics suggest L>NL). Of these variables, only a few of the parametric tests indicate significant 20

22 difference between corporate lobbyists and corporate non-lobbyists, due to the highly skewed distribution. A strong empirical evidence of significant difference between lobbyists and nonlobbyists is found in ROA (i.e., non-parametric results of all data panel show significant difference at less than, or equal to, 1% and 5% level). Hence, H6N is rejected in favour of H6A, but with directions opposite to the hypothesised. Moreover, some indications of the difference is evidenced in LIQUIDIT variable (i.e., parametric test results of PANEL A significant at less 1%, and, almost all non-parametric results of all data panel show significant different at less than, or equal to, 5% level). Accordingly, H7N is rejected in favour of H7A, but with directions opposite to the hypothesised. Lastly, one of non-parametric test results (i.e., the results of Panel ALL for both as well as lobbying) suggests significant difference between corporate lobbyists and corporate non-lobbyists in term of INTCOV (i.e., significant at less than 1% level), with directions opposite to the hypothesised. Considering that the significant differences are not found in the other data panels, hence, the empirical evidence to reject H5N is inadequate. refore, the test fail to reject H5N (i.e., INTCOV L< NL). tests also fail to reject H5N (i.e., TAXARBIT L>NL), since none of the data panels show the significant difference. 6. Discussion of the Significant Findings Over all, summary of empirical test results are presented in the following table. Table 6 Summary of Empirical Test Results Mean Comparisons (one tailed test) Variable SIZE Parametric Test (ANOVA) Non-Parametric Test (Kruskall Wallis Test) Panel All 0.00** 0.00** 0.00** 0.00** A 0.00** 0.00** 0.00** 0.00** B 0.00** 0.00** 0.00** 0.00** C 0.00** 0.00** 0.00** 0.00** D 0.00** 0.00** 0.00** 0.00** 21

23 Variable ROVSL TAXARBIT Table 6 (Continued) Summary of Empirical Test Results Mean Comparisons (one tailed test) Parametric Test (ANOVA) Non-Parametric Test (Kruskall Wallis Test) Panel All 0.00** ** 0.00** A 0.00** 0.00** 0.00** 0.00** B 0.00** 0.00** 0.00** 0.00** C 0.00** 0.00** 0.00** 0.00** D 0.00** 0.00** 0.00** 0.00** All ** A B C D All * A * 0.05* LIQUIDIT B * 0.05* C 0.00** 0.00** 0.03* 0.03* D * 0.05* All ** 0.00** A INTCOV B C D All ** 0.00** A ** 0.02** ROA B ** 0.01** C ** 0.01** D ** 0.01** All ** 0.01** A ** 0.02** DERAT B ** 0.01** C ** 0.01** D ** 0.00** **Significant at less than, or equal to, 1% level * Significant at less than, or equal to, 5% level Empirical evidences show that, except for INTCOV, corporate lobbyists and corporate nonlobbyists have significant differences in almost all of the hypothesised characteristics. results indicate that most of lobbyists are derivatives user companies, which use derivative to: (i) utilize economies of scale to reduce contracting costs (SIZE); (ii) manage foreign currency 22

24 exposures (ROVSL); (iii) enable to lessen tax payment and lower financial costs by way of tax and regulatory arbitrage (TAXARBIT); (iv) to enable to increase certainty of future positive NPV investment (LIQUIDIT). In addition, the empirical results suggest that lobbyist have higher earnings level (ROA) and higher gearing (DERATIO) compared to non-lobbyists. Interestingly, some of the empirical results suggest that the directions turn out to be opposite to the hypothesised in both the lobbying as well as lobbying. First, lobbyist s earning (i.e., ROA) is hypothesised to be lower than that of non-lobbyists, since, arguably lower earnings companies will be more threatened by the increased earnings volatility caused by the proposed standard, hence, have stronger motive to lobby. Unexpectedly, corporate lobbyists turn out to have higher earnings compared to non-lobbyists, in particular is in the case of lobbying. One of the explanations relates to the economies of scale theory, where big companies tend to utilise economies of scale by derivatives used, hence, resulting reduced cost and higher earning level. As empirical evidence suggests that the lobbyists have bigger size than non-lobbyists, consequently, the lobbyists are likely to be more efficient, hence, have higher earning level. Second, lobbyist s liquidity is hypothesised to be lower than non-lobbyists, so that they use derivatives to increase the certainty of future positive NPV investment. empirical tests suggest the opposite results to the hypothesised (i.e., lobbyist liquidity is significantly higher than non-lobbyists), particularly for the case of the lobbying. One of the alternative explanations relates to the lobbyists level of earning, which is higher than non-lobbyists. Companies with higher level of earnings are most likely to have higher certainty of future positive NPV investment compared to their lower level peers. For that reason, the empirical results suggest that lobbyist s liquidity is significantly higher than non-lobbyists. Further empirical examinations need to be conducted to support the alternative explanations for the inconsistent results demonstrated by corporate characteristics of both earnings level as well as 23

25 liquidity. In this case, correlations between SIZE and ROA, and also, between ROA and LIQUIDIT, are expected to be positively significant. other way to interpret these findings is that in this study, the theory of corporate use of derivatives is more relevant to explain the lobbying behaviour compared to the agency cost theory (i.e., executive compensation hypothesis, and, debt contract hypothesis). This gives early indication of the effect of accounting information on corporate use of derivatives to manage financial risk. 7. Summary and Conclusion This study concerns with corporate lobbying behaviour demonstrated during introduction of accounting standard on derivatives and other financial instruments FRS. observations cover two public hearings: (i) the, which include all the critical issues need to be addressed in the long run; (ii), which limit the proposed standards on the disclosures issues only. general hypotheses developed based on Sutton s (1984) framework of self-interest economic motive of lobbying is that lobbyists are having distinctive characteristics compared to non-lobbyists (i.e., H1 to H7). Mean comparisons procedures, both parametric as well as non-parametric methods, are used to examine the characteristic differences between lobbyists and non-lobbyists. So far, most of lobbying studies use contracting cost theory (i.e., executive compensation plan hypothesis and debt contract hypothesis) as well as political cost theory (i.e., tax and regulatory cost hypothesis) to explain and predict lobbying behaviour. Besides these theories, this study use the finance theory of corporate use of financial derivatives in predicting and explaining lobbying behaviour. Consistent with Modern Portfolio ory (Joseph, 1999), this study assumed that firm s financial policy, including derivatives use for hedging, has an effect on 24

26 firm s value. Hence, this study offer more balanced views of the contracting role (i.e., related to the agency costs) as well as informational role (i.e., related to firm s value) of accounting. empirical evidence shows that some of the corporate lobbyists characteristics (i.e., have higher earning or higher ROA, have more certainty of future positive NPV investment or higher LIQUIDIT, and, have lower degree of financial distress or higher INTCOV) turn out to be opposite to the hypothesised (i.e., have lower ROA, lower LIQUIDIT, and, lower INTCOV). contradictory results are explained better when looking from the perspectives of derivatives users companies. Accordingly, the opposite results are the logical consequences of the big size characteristic as one of the strongest and most consistent characteristics of corporate lobbyists found across data panel, in both the as well as lobbying. In this case, bigger size companies have the advantage of economies of scale to reduce the contracting cost associated with derivatives use. This means that bigger size companies tend to be more efficient, hence, having higher earning, and, consequently, having higher certainty of future positive NPV investment, and, therefore, have lower degree of financial distress. As far as this study is concerned, the main contribution of this lobbying studies is providing alternative perspective (i.e., corporate use of derivatives theory) to explain and predict corporate lobbying behaviour beside the well known agency cost and political cost theory. Corporate use of derivatives theory is rooted in the Modern Portfolio ory (Joseph, 1999), which assumes that firm s financial policy, including corporate use of derivatives hedging, is having an effect on firm value. Hence, differ to the previous lobbying studies which focus mostly on the contracting role of accounting (i.e., assumed accounting information as having an effect on the agency cost), this study adopts more balanced view, by considering also the informational role of accounting (i.e., assumed that accounting information is capable to influence investment decisions and firm s valuation). results of this lobbying study provide an early indication of the potential effect of the proposed accounting standards on corporate use of derivatives in risk management practice. 25

27 Considering that development of accounting standard on derivatives and other financial instruments are still continuing, future research should examine further how and in what way risk management practice among derivatives users companies are associated with the standard developments. In particular is when the ASB started to introduce accounting standard on derivatives which directly affect the accounting number, such as measurements and hedge accounting issues. Introduction of accounting standards on derivatives and other financial instruments are intended to increase the transparency as well as accountability of risk management decision, so that an efficient firms can be differentiated from an inefficient firms. By this way, corporate use of derivatives is expected to use prudently, hence, managers who use financial derivatives in risk management will be rewarded more fairly. However, accounting standards on derivatives and other financial instruments could have an adverse effect on the risk management practice if managers avoid using derivatives due to increased cost of compliance of the new standards. Accordingly, future research should examine the benefit as well as the costs of information of the new standards, so that the adverse effect of the new standard on derivatives and other financial instruments on the risk management practice could be avoided. 26

28 List of References 1. Accounting Standard Board (1998). Financial Reporting Standard Derivatives and Other Financial Instruments: Disclosures. ASB Publication, Central Milton Keynes 2. Ang, N., B.K. Sidhu and N. Gallery (2000). Incentives of Australian Public Companies Lobbying Against Proposed Superannuation Accounting Standards. Abacus, February: Bodnar, G.M., G.S. Hayt, R.C. Marston and C.W. Smithson. (1995). Wharton Survey of Derivative Usage by US Non-Financial Firms. Financial Management, 24 (2- summer): Bodnar, G.M., G.S. Hayt and R.C. Marston. (1996) Wharton Survey of Financial Risk Management by US Non-Financial Firms. Financial Management, 27 (4- winter): Bodnar, G.M. and G. Gebhardt (1998). Derivative Usage in Risk Management by U.S. and German Non-Financial Firms: A Comparative Survey. National Bureau of Economic Research, Working Paper, No Berkman, H. and M.E. Bradbury. (1996). Empirical Evidence on the Corporate Use of Derivatives. Financial Management, 25 (2- Summer): Cosh, Andrew (1975). Remuneration of the Chief Executives in the United Kingdom. Economic Journal, March: Citron, David. B.(1995). Incidence of Accounting-based Covenants in UK Public Debt Contracts: An Empirical Analysis. Accounting and Business Research 25 (99): Day,`J..and P.Taylor (1994). Accounting Aspects : Room for Improvement? Accountancy, London, 1 (1210): Deakin, E.B. (1989). Rational Economic Behaviour and Lobbying on Accounting Issues: Evidence from the Oil and Gas Industry. Accounting Review January: Dedman, Elisabeth (2003). Executive Turn Over in UK Firms: Impact of Cadbury. Accounting and Business Research, 33 (1): Dhaliwal, D.S (1982). Some Economic Determinants of Management Lobbying for Alternative Methods of Accounting: Evidence from the Accounting for Interest Costs Issue. Journal of Business Finance and Accounting, Summer: Dunne, T., C. Helliar, C. Mallin, L. Moir, K. Ow-Yong, D. Power (2003). Financial Reporting of Derivatives and Other Financial Instruments: a Study of the Implementation and Disclosures of FRS. Centre for Business Performance, ICAEW, Chartered Accountants Hall, Moorgate Place, London EC2P 2BJ 14. Georgiou, G and Clare B. Roberts (2004). Corporate Lobbying in the UK: An Analysis of Attitudes Towards the ASB s 1995 Deferred Taxation Proposals. British Accounting Review, XX (XXXX):1-15. Hill, N.T., S.W. Shelton and K.T. Stevens (2002). Corporate Lobbying Behaviour on Accounting for Stock-based Compensation: Venue and Format Choices. Abacus, February: Joseph, N.L and Hewins, R.D.(1997). Motives for Corporate Hedging among UK Multinationals. International Journal of Finance and Economics, (2): MacArthur, J.B. and R.E.V. Groves (1993). An Empirical Investigation into the Impact of Profit Sharing Schemes of Executives on the Content of Corporate Submission on Proposed Accounting Standards. Journal of Business Finance and Accounting, September:

29 18. Mallin, C.A., Ow-Yong, K. and Reynolds. M (2001). Derivatives Usage in UK Non- Financial Listed Companies. European Journal of Finance, 7(1- March): Marshall, Andrew P. (2000). Foreign Exchange Risk Management in UK, USA and Asia Pacific Multinational Companies. Journal of Multinational Financial Management, 10: Schalow, C.M. (1995). Participation Choice: the Exposure Draft for Postretirement Benefits other than Pensions. Accounting Horizons, March: Solomon, Jill Frances (1999). Do UK Institutional Investors Adopt a Dual Strategy for Managing Foreign Exchange Risk?. British Accounting Review, 3(2-June): Solomon J.F. and N.L. Joseph (2000).Which Corporate Hedging Motives are Appropriate? An Institutional Shareholder s Perspective. International Journal of Finance & Economics, Chichester, 5(4): Sutton, T.G (1984) Lobbying of Accounting Standard-setting Bodies in the UK and the USA : a Downsian Analysis. Accounting, Organizations and Society, 9(1): Watts, Ross L and Jerold L.Zimmerman (1978). Towards a Positive ory of the Determination of Accounting Standard. Accounting Review, LIII (1): Watts, Ross L and Jerold L.Zimmerman (1986).Positive Accounting ory. Prentice- Hall, Englewood Cliffs, NJ. 28