Our recent journal article in Accounting & Finance investigated two questions:
- the extent of cross-border profit shifting activities in Australia via two main channels: intra-group transfer pricing and intra-group debt financing and/or interest expense loading; and
- the change in the extent of cross-border profit shifting since 2013 to empirically assess the effectiveness of the Australian tax base erosion and profit shifting (BEPS) countermeasures targeting these two channels.
This is part of our research project to assess the effectiveness of the Australian BEPS countermeasures since the inception of the Organisation for Economic Co-operation and Development’s BEPS project in 2013.
In a previous article, we have provided some evidence of Australian BEPS activities using the modified Hines and Rice approach, which investigates the relation between profit before tax reported by foreign-owned Australian companies (FOACs) and the tax rate differentials between Australia and other countries where the related foreign multinational groups operate.
Profit shifting channels and Australian countermeasures
Using intra-group transfer pricing, foreign multinationals manipulate the prices charged on the export and import of goods and services between group members located in countries with different tax rates, so that profits are shifted from high-tax to low- or no-tax jurisdictions.
To tackle the problem, the Australian Parliament modernised the transfer pricing rules through two reforms. An earlier reform was the inclusion of Subdivision 815-A in the Income Tax Assessment Act (ITAA) 1997 in 2012. The second reform was the shift from ITAA 1936 Division 13 and ITAA 1997 Subdivision 815-A to a modern and comprehensive transfer pricing regime as articulated in ITAA 1997 Subdivisions 815-B, 815-C and 815-D from 1 July 2013.
The Australian Parliament also inserted Subdivision 815-E that contains country-by-country reporting requirements into the ITAA 1997 in 2015.
On the other hand, internal debts provide opportunities for foreign multinationals to minimise overall tax payments through deduction of interest expense in high-tax jurisdictions (intra-group debt financing). The interest rates applied to intra-group debts can even be inflated (intra-group interest expense loading).
Australia has a comprehensive thin capitalisation regime as articulated in Division 820 of the ITAA 1997. Division 820 was amended to tighten the safe harbour limits and the worldwide gearing debt limits effective from 1 July 2014. It was further amended to require an entity to use the value of the assets, liabilities and equity in its financial statements from 8 May 2018. The tightened rules, however, use a methodology different from the best approach recommended by the BEPS Project: the former uses debt-based tests, while the latter recommends earnings-based tests.
Research design
We first adopt an identification strategy that relies on the differences in a set of accounting ratios between FOACs that are Australian subsidiaries of foreign multinationals and mainly domestic-owned listed Australian companies (DOLACs) to measure the extent of BEPS through these two specific channels used by foreign multinationals.
DOLACs are selected as a counterfactual that mimics a situation without profit shifting under the Australian dividend imputation system, as the system reduces DOLACs’ incentives to pursue tax avoidance strategies. More importantly, a DOLAC is the ultimate parent of a group, so its consolidated financial statements only reflect transactions with independent parties outside the group.
In contrast, a FOAC is only the Australian subsidiary of a foreign multinational, so its consolidated financial statements reflect transactions with foreign affiliates within the group as well as transactions with independent parties outside the group.
To achieve a fair comparison, we apply two alternative matching methods – propensity score matching and coarsened exact matching – to match FOACs with comparable DOLACs in terms of industry and firm size.
Second, we adopt a difference-in-differences approach to estimate whether and to what extent cross-border profit shifting by FOACs has been reduced after the implementation of related BEPS countermeasures from 2013.
The first difference is the difference in the profit shifting behaviours between FOACs and DOLACs. The second difference is the change in the profit shifting behaviours of FOACs relative to DOLACs before and after the implementation of related Australian BEPS countermeasures.
Our study period is the 14 years from 2007 to 2020: the six years from 2007 to 2012 is the pre-BEPS period, and the eight years from 2013 to 2020 is the post-BEPS period.
We start with the samples of FOACs and listed Australian companies from the lists of Australia’s top 2,000 companies in 2012 and 2016 on the IBISWorld website. Only listed Australian companies with less than 20% foreign ownership are included in the sample of DOLACs to control for the impact of foreign ownership on tax avoidance behaviours under the dividend imputation system.
Empirical findings
We find that the main profit shifting channel used by FOACs is intra-group transfer pricing. FOACs mainly used suppressed selling prices and/or inflated purchase prices to shift profit, rather than incurring higher royalties, management fees and other non-finance expenses in related-party transactions. Royalties might, however, be embedded in the prices of trading stock purchases to avoid royalty withholding tax.
Based on the regression coefficient of the propensity score matching model, we roughly estimated that the average gross profit shifted out of Australia by each FOAC in our sample using transfer pricing is approximately AUD43 million per year using matched DOLACs as the benchmark or counterfactual. In addition, compared with matched DOLACs, FOACs may pay more interest expense and claim more tax deductions by bearing inflated interest rates on related-party debts, but have a similar level of total debt.
As a result of profit shifting, FOACs report lower pre-tax profit and lower tax expense for every dollar of sales revenue in Australia than comparable DOLACs.
As for the effectiveness of Australian BEPS countermeasures, despite a couple of additional tests, we cannot find significant evidence suggesting that up to 2020, the relevant BEPS countermeasures targeting transfer pricing and interest expense loading were effective in reducing profit shifting by FOACs via these two channels.
Challenges in combating BEPS activities
The lack of effectiveness of Australian BEPS countermeasures might be explained by law enforcement or administrative time lags, because the Australian Taxation Office may take years to conduct tax audits, raise amended assessments and go through the lengthy tax dispute resolution process before the effect of BEPS countermeasures on outward profit shifting can be reflected in financial reports of FOACs.
The lack of effectiveness might also indicate the defects in the current transfer pricing and thin capitalisation rules. For example, it is difficult to identify a comparable condition or transaction between independent parties under the arm’s length principle. Therefore, governments may consider introducing a digital service tax or finding an alternative to the arm’s length principle, such as formulary apportionment.
Current Australian thin capitalisation rules do not restrict the level of interest expense directly as a proportion of earnings. Consequently, the tightened debt-to-equity ratio under the rules may not be effective in limiting excessive tax deductions related to interest expense, because inflated interest rates can be charged on related-party debts unless denied by the transfer pricing rules.
In the October Federal Budget 2022–23, the Treasurer has announced that the government will amend thin capitalisation rules, replacing debt-based tests with several earning-based tests to directly limit interest expense deductions.
Where to next?
With the results in mind, we call for future research to extend the study period, especially the period after the implementation of BEPS countermeasures and perform similar analysis using improved datasets and research designs.
Future studies can also develop identification strategies and construct proxies to detect cross-border profit shifting through channels other than the two investigated by our study (for example, hybrid mismatch and the utilisation of tax treaties).
Journal article
Tran, A. and W. Xu, 2023, “A study of profit shifting channels: Evidence from Australia”, Accounting & Finance, 00: 1-33. Available at http://doi.org/10.1111/acfi.13166.
Recent Comments