Writing yesterday, I addressed the propensity of certain journalists to ready excitement. I examined what the Facebook deal didn’t mean. Today I want to write a little more on what it does mean. What is the likely thought process that led Facebook to this point? And how much cash is it likely to involve?
First, the thought process.
As David Quentin discusses – in an influential piece of work here – for the sophisticated player tax planning is essentially an exercise in managing tax risk. It involves a choice to seek to reduce your tax liability, which choice carries the the cost of embracing more tax risk. If my assessment of the risk as low is right, I save 100. If my assessment of the risk is wrong, I give back that 100 and I suffer the public embarrassment of fighting and losing a tax case.
I will revisit this equation in future weeks: it doesn’t operate in the public interest. But what I want to cover now is how it affected Facebook’s decision as to where to book UK sales.
Facebook understood – as did Google before it – the diverted profits tax to have backfired. It did not have the consequence of bringing Facebook within the (punitive) 25% diverted profits tax net. What it did do was increase the risk profile attached to Facebook’s tax planning. On the left side of the balance sat the prospective tax savings; on the right sat a now enhanced tax risk. But another factor sat, too, on the right hand side of the balance: the cost attached, especially for a consumer facing business like Facebook’s, to the negative publicity attendant on paying too little tax.
You weigh that tax saving against the increased tax risk and impact of that publicity on your global brand and, ultimately, you make a call as to which side of the balance sits heavier. That would have been the assessment that led Facebook’s to rework part of its UK tax calculation.
But what sat on the left side. What was the actual sum on the left hand side of the balance in my calculation. Can we know?
We would need to know (1) what sales were made by Facebook to UK customers. Of these sales, what subset (2) were effected by UK relationship managers? (You will recall from yesterday’s blog post that only these sales, and not advertising bought by smaller customers online, are affected by the announcement).
Then what profits were generated by selling – as opposed to other parts of Facebook’s value chain (3)? Under existing international tax rules the UK can’t tax what you might think of as ‘manufacturing’ profits here, only ‘retailing’ profits.
Finally, one would apply to those profits (4) our prevailing rate of corporation tax.
As to (1), the latest year of accounts available (for Facebook Ireland Limited’s accounts, where European sales have been booked up to now) gives European sales of €4.8bn in the year to 31.12.14. It doesn’t say what UK sales are, but the relationship of the UK’s GDP to that of the EU (18%) might give you a reasonable proxy for the proportion of sales made in the UK. 18% of €4.8bn would represent €860m. Divide by 1.3 to convert to £ would give you UK sales of £660m.
As to (2), we know that Google generates 60-70% of its sales from 1% of its customers. Let’s assume that the same is true of Facebook: that 1% of customers are serviced by the UK relationship managers, and they generate a (mid-range) 65% of sales in the UK. That would give you 65% of £660m of UK sales, or £430m of sales booked in the UK.
As to (3), worldwide Facebook appears to have a 40% margin (of pre-tax profits/turnover). Attribute, for the sake of argument, 10% of that margin to selling and the remainder to generating and maintaining the technology and intellectual property rights. That would imply a profit on those UK sales of 4%, or £17m.
If you apply to that £17m our 20% rate of corporation tax, you’d arrive at a UK corporation tax bill of £3.4m per annum on those sales. That’s a little less than, but not of a completely different order to, the £4-6m (independently) estimated by Tim Davies, Head of Tax at Mazars, yesterday.
As I explained yesterday, the deal is forward looking. It will apply from April 2016 going. So we won’t see any signs of it in the accounts for the period to 31 December 2015 (which should appear in October 2016). The first sign of it will appear in the accounts for the period to 31 December 2016 (which should appear in October 2017). These will contain 9 months rather than a full year of this new treatment giving rise to an additional corporation tax bill disclosed to us in October 2017 of 75% of £3.4m or £2.55m.
But that’s not all.
We also know – see my post of yesterday – that Facebook UK has approximately £10m of ‘deferred tax assets’ (effectively credits against future tax liabilities, and here comprised of carried forward losses and capital allowances). These would need to be depleted before any cash moves from Facebook UK to HMRC. £2.55m in the year ending 31 December 2016 would leave just about enough to clear the tax bills for the years ending 31 December 2017 and 31 December 2018. So the first cash changing hands in consequence of this deal would be in the year ending 31 December 2019 and we’d find out about it in October 2020.
Of course, this calculation makes a number of assumptions which may or may not be right. If Facebook’s UK revenues grow faster the deferred tax assets would be depleted earlier. Earlier depletion might also result from HMRC’s existing enquiries into Facebook UK’s affairs bearing tax fruit. Or the assumptions made in my calculations might just be off.
But, even bearing these in mind, the BBC’s Businesss Editor Kamal Ahmed’s blog post of yesterday which led with:
Facebook is set to pay millions of pounds more in tax in the UK after a major overhaul of its tax structure.
might look, in the cold light of day, a little excitable. “Set,” perhaps, but not for a good few years to come.Follow @jolyonmaugham