This week is the start of term at LSE, so I’m back blogging. While I was away, Chris Lenon started his own blog. Lenon was Head of Tax at Rio Tinto, and then chaired the tax committee of BIAC, the industry group that lobbies at the OECD. We used to bump into each other quite a bit, and every time I log into LinkedIn it now asks me to endorse him with the question, “does Chris Lenon know about tax?”
Chris’ first post was about Starbucks, and I found it very interesting. He makes a point that I’ve also heard from other former tax officials (Lenon used to work at the Inland Revenue):
Starbucks has allegedly not made taxable profits in most of its 15 years of operations in the UK. In the real world companies can sustain start up losses for a period of time (say 3 to 5 years) but after that one has to question whether they would perservere trading at a loss, or cut their losses? It is questionsable whether a subsidiary which continues loss making beyond start up is operating at arms length as non connected parties do and it would certainly pose questions about its transfer pricing.
Over at Accountancy Age, Ben Saunders and Richard Murphy are debating whether the individual Starbucks payments were priced at arm’s length. Ben says that “Starbucks have unconnected third party licensees, an almost ideal situation for determining an arms-length price.” But perhaps there is a simpler jumping-off point, which is the business as a whole, not the individual parts and prices. And if you start there, it seems obvious that a company operating at arm’s length would not sustain losses for 15 years (even with a few profitable ones).
Martin Hearson 
To be fair, I used the Starbucks situation as an example of something intended by law but producing unpopular results. I didn’t want to get into Starbucks in depth on that site.
Personally, I don’t agree with tax avoidance, but it depends how you define it. So, I was really trying to pin him down on how he evaluates tax avoidance for the purpose of this debate. He hasn’t responded, even with a hypothetical example.
I don’t think Richard is consistent in how he applies his definitions, and I consider that he willing to argue anything to make the point he wants to make. Hence he argued yesterday that morality is “just a statement of opinion”, which is something I cannot imagine he actually believes given his previous use of moral statements.
Back to tax though, the kicking off point you suggest might be fair in certain situations (I’m not sure that is in this case, but I’m too Starbucked out at present). The law, whether rightly or wrongly, states that you apply an arms-length price to transactions. Starbucks seem to have done that judging by all the evidence at the PAC and from their accounts, so it is not tax avoidance.
Remember, these rules are as much to protect revenues in the Netherlands and Switzerland as it is here.
I suppose I’m mainly looking to categorise the debate more effectively than it is at present. There is tax avoidance and there is absurdity within the tax system.
These cannot be addressed in the same way. Anti-avoidance legislation wouldn’t touch Starbucks.
The real “culprit” in this situation is the transfer pricing rules themselves. Whether they are actually fair or not, they are not seen to produce a fair outcome. Blaming people who implement the rules in good faith isn’t productive, or fair, to my mind.
I don’t really have a view on any of the proposed solutions, but identifying the problem correctly has to be the starting point.
Thanks Ben. As I understand the argument made by Chris and others, the Starbucks (say) result is arguably not intended by law, because the law intends that the UK company should be making the profits it would make if trading at arm’s length. Sometimes the sum of different transactions – each of which meets the arms length test – isn’t itself arm’s length, because something is missing from the whole picture.
I would respond that Starbucks UK did make taxable profits over a significant period. And didn’t make sustained tax losses over 15 years, or for more than a 3 year period over the main periods in question.
They started up a few pilot stores and then invested heavily. At the end of 2002 (I think) they reorganised their group structure and moved the roasting plant to the Netherlands, so something wasn’t working with having the roasting plant in the UK.
The royalty payments in question didn’t actually commence until 2003* (in connection with the roasting plant move, I believe). By 2006 they were making taxable profits.
They only started creating tax losses in 2009 again, which is understandable given that coffee isn’t exactly a necessity in a recession. They hadn’t recovered by the end of 2010, but at no point from the commencement of the royalty payments did they make more than 3 years worth of losses. I haven’t got 2011 accounts handy, but that’s still only a third year and wouldn’t break the 3-5 year loss making streak referenced above.
As such, I would suggest the argument, whilst valid, doesn’t apply because it doesn’t fit the fact pattern for the taxable profits.
The theory fits for the UK GAAP position (and remember Starbucks draws up under US GAAP which saw the UK as profitable), but that’s irrelevant in the final examination of things from a tax perspective. The profits as adjusted by tax legislation is actually what we are concerned with here – because that’s what determines what tax is paid, which is what we’re concerned with.
I’ll check 2011 later, but I think that’s about right.
*At the end of 2002, the tax losses created by Starbucks was worth a measly 2.7m, so they didn’t store up huge losses in those first 6 years. The change in structure does represent a decision to change the business model.
I thought it might be worth adding that I don’t know when the transfer pricing adjustments (down to 4.7%) agreed with HMRC were brought into account.
I think it might be in 2008, because the expenses not deductible for tax purposes jumps by an appropriate amount. So I think that in 2006 and 2007 they were making profits despite a royalty on 6% of turnover.
I might return to the subject with some sort of graph…
By “profits” in the second para I mean taxable profits….
“As such, I would suggest the argument, whilst valid, doesn’t apply because it doesn’t fit the fact pattern for the taxable profits.”
Very interesting! Thanks for this input.
Apologies, I need to correct my statement regarding 2010. Starbucks UK actually made a taxable profit in 2010 because they show that they were utilising losses.
If Starbucks UK is making losses over 15 years, then I think you’re right that a standalone company would have fallen over by now. So the UK is obviously getting some support, but I think there are two aspects of the arm’s length principle at issue here, as Starbucks US is acting in the capacity of both trading partner (or at least having a strong influence over trading terms) and as shareholder.
It may be that the US is supporting the UK by granting it favourable trading terms, but it’s equally possible that the support comes in the form of funding the losses of past years in the hope of making future profits. In my experience people are quite happy to wipe slates clean if they think the future will be rosy – and people are always willing to believe that the future will be rosy.
That is: faced with the choice of closing the business, and taking a financial hit as well as being embarrassed next time they meet their counterparts from Costas, or persevering becuase they’ve finally worked out how to stem the losses and next year will be fine, I find it much more plausible that Starbucks executives would want to keep the UK going. Especially as Ben’s figures (showing very few losses left) suggest that the UK trade is only just underwater, and so won’t need much of a change in the wind to push it into profit.
If I remember the Starbucks accounts properly, they’ve had some fairly hefty injections of equity in recent years, which seems to be consistent with the idea that the US is supporting them as investor rather than trading partner.
There’s one other thing I realised. Viewing the transaction over a fifteen year period raises the question, how do you know what adjustments to make in year one?
It’s all well and good saying that there should have been a greater adjustment with the benefit of fifteen years hindsight, but companies have to make decisions at the time they need to be implemented (especially with transfer pricing, you always hear of the importance of contemporaneous evidence supporting calculations).
Companies self assess on an annual basis. There is a year-long enquiry window after submission and once that’s closed unless HMRC have grounds for raising a discovery enquiry those years are gone. You can only request the adjustment going forward.