Dividend Imputation and BEPS. (Estelle) Xuerui Li & Alfred Tran

Transcription

1 Dividend Imputation and BEPS (Estelle) Xuerui Li & Alfred Tran

2 2 About This Study Whether and how dividend imputation mitigates tax avoidance by large Australian companies via tax base erosion and profit shifting (BEPS)?

3 3 Motivation Why important? Prevalence of corporate tax avoidance via BEPS OECD, G20: call for actions to tackle BEPS Preliminary empirical evidence on the corporate tax avoidance-reducing effect of dividend imputation E.g. Ikin and Tran (2013), Amiram et al. (2013) Perhaps a dividend imputation system can be part of the solution

4 4 Dividend Imputation System Companies: Profit: Tax =After-tax profit: P 0.3P =0.7P Total Taxes: 0.3P Franked Dividends (carry franking credits 0.3P) + Shareholders Tax on Dividends: Dividends: 0.7P +Franking credits: +0.3P Marginal Tax Rate: P MTR Franking credits: 0.3P P x MTR 0.3P = P x MTR

5 5 Dividend Imputation System Notes: Foreign income tax paid by companies cannot be attached to dividends as franking credits Foreign shareholders cannot utilise Australian franking credits as tax offsets in their home countries

6 6 Dividend Imputation System Corporate income tax does NOT reduce the aftertax returns of shareholders, provided that: Australian operations pay Australian income tax Australian shareholders franking credits can be claimed as tax offsets Distribution of franked dividends

7 7 Dividend Imputation System Evidence from prior studies Amiram et al. (2013): greater dividend imputation benefits are associated with lower corporate tax avoidance Ikin and Tran (2013): large and profitable Australian listed companies have little incentive to avoid Australian corporate income tax due to dividend imputation Both studies do not take into account the impact of foreign ownership The impact of foreign operations is not consistent across models and studies

8 8 Dividend Imputation System This study: focus on the impact of foreign ownership on corporate tax avoidance in the Australian dividend imputation system. Australian Subsidiaries of Foreign Multinationals (ASFMs) Australian Listed Companies with Mainly Domestic Operations (ALCDs) Foreign Ownership Australian Operations Australian Ownership

9 9 BEPS Base erosion and profit shifting (BEPS) is used by OECD to refer to the prevalent corporate tax avoidance practice by multinational enterprises Two widely-recognised BEPS practices Intra-group transfer pricing Thin capitalisation

10 10 BEPS: Intra-group Transfer Pricing Prices charged for the flow of goods and services between members of a multinational enterprise (MNE) that operate in different countries with different tax rates profit shifting Sell goods/services at inflated prices Low-tax country Subsidiary A OR Sell goods/services at depressed prices High-tax country Subsidiary B Profit Profit

11 11 BEPS: Intra-group Transfer Pricing Evidence from prior studies Indirect evidence: Negative relation between foreign subsidiaries tax rates and their profitability levels Direct evidence: For MNEs with either subsidiaries or parent companies located in US, when tax rate of trade country decreases, US intra-group export (import) price becomes lower (higher) than non-intra-group trade prices

12 12 BEPS: Intra-group Transfer Pricing Australia is a relatively high-corporate-tax country ASFMs are hypothesised to engage in tax avoidance via intra-group transfer pricing to a greater extent than ALCDs: Supply goods/services to other members in their MNE groups at discounted prices, and/or Purchase good/services (including patented technologies, management services, etc.) from other members in their MNE groups at inflated prices

13 13 Income Statement Sales revenue Minus: Cost of goods sold = Gross profit Add: Other income Minus: Expenses, e.g. royalties, management fees, etc. = Earnings before interest and tax (EBIT) Minus: Interest expense = Profit before tax Minus: Income tax expense = Profit after Tax H1A: ASFMs have lower gross profit to sales revenue ratios than comparable ALCDs H1B: ASFMs have lower earnings before interest and tax (EBIT) to sales revenue ratios than comparable ALCDs

14 14 BEPS: Thin Capitalisation Heavy use of debt, rather than equity, as source of finance (and possibly inflated interest rate) High-tax country Subsidiary B Heavily debt-financed Claim substantial interest expense Interest expense Profit

15 15 BEPS: Thin Capitalisation Evidence from prior studies Positive relation between foreign subsidiaries tax rates and their leverage ratios and interest expense to sales ratios

16 16 BEPS: Thin Capitalisation ASFMs are hypothesised to engage in tax avoidance via thin capitalisation to a greater extent than ALCDs: More heavily debt-financed (as opposed to equityfinanced), Incur/Claim relatively more interest expense

17 Income Statement Sales revenue Minus: Cost of goods sold = Gross Profit Add: Other income Minus: Expenses (other than interest expense) = Earnings before interest and tax (EBIT) Minus: Interest expense = Profit before Tax Minus: Income tax expense = Profit after Tax H2A: ASFMs have higher interest expense to sales revenue ratios than comparable ALCDs Balance Sheet Assets Liabilities: Long-term borrowings, etc. Equity H2B: ASFMs have higher leverage ratios (longterm borrowings to total assets) than comparable ALCDs 17

18 18 BEPS: Effectiveness of Tax Avoidance Intra-group Transfer Pricing Thin Capitalisation Income Statement Sales revenue Minus: Cost of goods sold = Gross Profit Add: Other Income Minus: Expenses, e.g. royalties, management fees, etc = Earnings before interest and tax (EBIT) Minus: Interest expense = Profit before tax Minus: Income tax expense = Profit after Tax

19 19 BEPS: Effectiveness of Tax Avoidance H3A: ASFMs have lower pre-tax accounting profit to sales revenue ratios than comparable ALCDs H3B: ASFMs have lower income tax expense to sales revenue ratios than comparable ALCDs

20 20 Other Measures of Tax Avoidance? Why not use effective tax rates as tax avoidance measure here? ETR cannot capture BEPS because BEPS reduces tax liability (numerator) and pre-tax profit (denominator) simultaneously

21 21 Table 1 Apple versus JB Hi-Fi (Year 2012) Apple Pty Ltd (ASFM) JB Hi-Fi (ALCD) $Million % $Million % Sales revenue 5, % 3, % Cost of sales 5, % 2, % Gross profit % % Other income (expense) % 1 0.0% Expenses % % Profit before tax % % Tax expense % % Profit after tax % % Effective tax rate (ETR) 40.7% 29.5% Tax expense / Sales revenue 0.67% 1.40%

22 22 Sample Sample year: 2012 Data source: IBISWorld Australia ASIC DatAnalysis Premium * Osiris (foreign ownership)

23 23 Sample Australia s top 2,000 companies (IBISWorld Australia) ASX listed companies (DatAnalysis Premium) Subsidiary of foreign MNE Wholly foreign-owned Other information Exclude financial companies, companies operating in countries other than Australia and New Zealand, companies without annual reports available from ASIC Exclude financial companies, trusts, stapled securities, foreign companies Take top 500 (sales revenue) Exclude companies with significant foreign operations ASFMs (319) ALCDs (360)

24 24 Sample Further exclude companies with pre-tax profit to sales revenue ratios being greater than 1, or smaller than -1 ASFMs 312 ALCDs 352 ALCDs (352) Perform industry classification coding (8-digit Global Industry Classification Standard)

25 25 Research Design Propensity score matching paired sample t-test Propensity score matching technique to construct paired sample The probability of a company in the sample being an ASFM based on two explanatory variables: industry affiliation and firm size Estimated based on logistic model: ASFM (0, 1) = Size (Ln Sales) + Industry dummies

26 26 Research Design Matching Statistical software package estimates propensity score and matches each ASFM with an ALCD with the closest estimated propensity score Comparison With the paired sample, compare: Gross profit / Sales revenue Earnings before interest and tax / Sales revenue Interest expense / Sales revenue Long-term borrowings / Total assets Pre-tax profit / Sales revenue Income tax expense / Sales revenue

27 27 Research Design Other test: Ordinary Least Squares regressions Gross profit / Sales revenue Earnings before interest and tax / Sales revenue Interest expense / Sales revenue Long-term borrowings / Total assets Pre-tax profit / Sales revenue Income tax expense / Sales revenue = ASFM Indicator + Size + Industry dummies Unmatched Results

28 28 Limitation Limitations: Industry coding may not be precise (multiple industries) Sample size is relatively small Foreign ownership of ALCDs is not directly controlled for Sales revenue may be depressed by companies engaging in intra-group transfer pricing

29 29 Descriptive Statistics Table 3 Descriptive Statistics Variables Mean ASFMs (312 companies) Std. dev. Min Max Mean ALCDs (352 companies) Std. dev. Min Max Mea Observable variables used in the selection model Firm size: Ln(sales in $000) Industry dummy variables Outcome variables Gross Profit / Sales EBIT / Sales Interest expense / Sales Leverage Pre-tax profit / Sales Income tax expense / Sales

30 30 ASFMs versus ALCDs - Unmatched Difference Standard Error p-value Gross profit / Sales EBIT / Sales Interest expense / Sales Leverage Pre-tax profit / Sales Income tax expense / Sales

31 31 ASFMs versus ALCDs - Matched Difference Standard Error p-value Gross profit / Sales EBIT / Sales Interest expense / Sales Leverage Pre-tax profit / Sales Income tax expense / Sales

32 32 ASFMs versus ALCDs - Matched Why not thin capitalisation? MNEs have incentives to keep after-tax profits overseas ASFMs have substantial cash No need to finance Profit repatriation by paying dividends: 0.7P ASFM Profit: P Tax: 0.3P After-tax profit: 0.7P US Parent Company Tax: P US tax rate 0.3P

33 33 Conclusion What we do: Examine whether the dividend imputation system mitigates ALCDs tax avoidance Examine whether ASFMs engage in greater tax avoidance via BEPS (intra-group transfer pricing & thin capitalisation) than ALCDs because of their foreign ownership

34 34 Conclusion What we find: ASFMs engage in intra-group transfer pricing (but NOT thin capitalisation) to a greater extent than comparable ALCDs ASFMs effectively reduce their Australian income and tax burden by transfer pricing

35 35 Conclusion Theoretical contribution Contributes to existing knowledge of the impact of dividend imputation on corporate tax avoidance Prior studies of dividend imputation systems have mainly been conducted in the field of finance Little attention has been paid to the impact of dividend imputation on corporate tax avoidance

36 36 Conclusion Policy Implications National: The Australian transfer pricing rules before 2013 may not be as effective as expected Impose arm s length standards on MNEs internal dealings, and trading of property or services between separate legal entities under international arrangements Difficult to identify comparable transactions and hence determining arm s length prices MNEs may circumvent the laws Australia should keep the dividend imputation system At least it has the effect of shifting the incentives to avoid tax away from Australian listed companies

37 37 Conclusion Implications International: OECD, G20: action plan to tackle BEPS Dividend imputation system reduces corporate tax avoidance But not for companies with foreign ownership Any improvement? Is a global dividend imputation system feasible?