GloBE: learning the lessons of the past

The OECD’s pillar two paper on the Global Anti-Base Erosion (GloBE) proposal has requested comments on the ‘use of financial accounts as a starting point for the tax base determination’.

At a very high level, GloBE would look at whether every member of an MNE is paying an effective tax rate (ETR) at least equal to a (as yet to be agreed) fixed rate and, if it were not, to top that up. Given that ETR = (tax paid/tax base), the $64,000 question is how do you consistently determine the tax base across multiple jurisdictions?

As the OECD’s paper notes, you could apply the tax rules from the parent’s jurisdiction to the subsdiaries. But that is likely to be very onerous and costly as that is not something that is required for any other purpose. Even most CFC regimes don’t require that for all entities due to the myriad of entity level/substance based exemptions. Further, that would mean that entirely acceptable differences between the way two different jurisdictions have designed their tax rules could give rise to a GloBE top up in circumstances where the relevant policy concern is not present.

So, the paper invites comments on using financial accounts instead. Superficially, that is attractive. Many groups who are required to prepare consolidated accounts will already have to recalculate the accounting income of their subsidiaries using an appropriate accounting standard for that purpose. Further, it may turn out to be the only practicable starting point. But there are two key learnings from the UK’s experience in seeking to tie rules for calculating tax base to accounting rules which should be borne in mind.

The first is that it effectively hands control over the rules for calculating the tax base to someone (namely the body responsible for setting the relevant standards) that may change those rules in a way that the relevant taxing jurisdiction does not want. A good example here is the UK’s loan relationship rules which were introduced in 1996 and, broadly speaking, aimed, and still aim, to align the tax treatment of corporate debt with its accounting treatment. However, the accountants subsequently introduced the concept of derecognition, allowing, for example, a company to remove a loan from its balance sheet, with a corresponding debit, simply by issuing mirroring preference shares. It took the UK some time to realise that that change had occurred, and was effective for tax purposes, and to bring in legislative changes to reset the tax base. So, the first key learning is that it is not just a case, as the OECD paper recognises, of identifying up front what departures from accounting treatment may be required but an ongoing process. And the effort involved in ensuring that changes in accounting standards are constantly reviewed, and required departures therefrom implemented in real time, to avoid a GloBE top up being applied where it clearly should not be (or vice versa), should not be underestimated.

The second key learning is that it is not a purpose of accounting standards to determine the tax base (and, accordingly, not a purpose the body responsible for setting those standards will bear in mind). This is neatly illustrated by the GDF Suez tax case in the UK. In that case, a UK company in the Enron Group, TPL, had a number of valuable claims in the bankruptcy of certain US subsidiaries of Enron. Those claims were not recognised on TPL’s balance sheet because they were ‘contingent assets’. Periodically payments were made out of the bankruptcy proceedings, at which point such amounts were recognised in TPL’s accounts. Essentially, the relevant accounting standard says that amounts cannot be recognised in respect of contingent assets until they are ‘virtually certain’. Why? In order to prevent a company overstating its financial position to protect creditors and others relying on the company’s accounts. It is a natural corollary of the accounting concept of ‘prudence’.

TPL transferred those claims to a wholly owned offshore SPV for shares. Nothing was recognised in its accounts as a result. Why? Because in substance nothing had changed. In essence, TPL had exactly the same contingent assets that it had before the transaction, they were just now in a wrapper, the SPV. It hadn’t in any sense realised income from the claims, or made it virtually certain that it would. So the accounting rules wouldn’t let TPL recognise anything at that stage. Indeed, if they had it would have driven a coach and horses through the prohibition on recognising income on contingent assets in the first place if a group could side step that, and accelerate income, by just dropping contingent assets into an SPV. So from an accounting perspective, and given the purpose of TPL’s accounts, this was clearly the ‘right’ answer.

But from the tax authority’s perspective it was the ‘wrong’ answer, because TPL was subject to tax in accordance with its accounts on the claims, under the UK’s loan relationship rules, but not on an accounts basis for the shares. So it argued, and the UK courts agreed, that TPL’s accounts did not ‘fairly represent’ the profit that had arisen to TPL on the transaction if you viewed it as a sale of claims for consideration equal to the value of the SPVs shares and that the UK’s rules allowed them to depart from the accounting treatment and tax the ‘profit’ in those circumstances.

Comments on the proposal are invited by 2 December, 2019. It will be interesting to see how different stakeholders respond. In particular, will any of the bodies responsible for setting accounting standards proffer a view on the important new role they may be offered as setter of the rules for a global tax base?