Yes, Google can afford to lose $9bn in Motorola sale. But did it really?
Choc Fac should max tax axe crack in Lenovo sale whack
Comment So just how much has Google lost on buying and selling Motorola Mobility? $9bn - as The Telegraph seems to think? $7bn as simple arithmetic would seem to indicate? Or how about a very decent indeed profit as the vagaries of tax law might indicate - with that tasty patent portfolio thrown in for free?
Let's start with a number that we know is wrong, that $9bn loss. From The Daily Telegraph:
Did it ever make sense for Google to own a handset manufacturer? Was its purchase of Motorola simply a $9bn play for valuable patents, with which it can now defend the entire Android ecosystem
Quite how we get to $9bn I'm not sure. The original purchase price was $12.5bn and the sale to Lenovo is for $2.91bn but perhaps that's close enough for journalism these days. But as the Telegraph separately notes, Google also netted $2.6bn from the sale of the set-top division to Arris:
Google had already sold Motorola Home, a TV set-top box business acquired as part of Motorola Mobility, for $2.6bn.
So now we're down to a $7bn loss on the whole deal. Except, as I've pointed out here before, there were also tax losses inside Motorola Mobility. There are limits on how you can use these under US law, but Google would certainly be able to use them because it did indeed take over the whole company and did indeed run it – as the existence of the Moto X and Moto G proves. And this is where the real reason Google bought Motorola comes into play:
"The tax benefits of the deal make what was a good deal into a great deal," said Robert Willens, a New York accounting and tax expert. He estimated that through the acquisition, Google can expect to reap $700m a year in tax deductions from future profits each year through 2019. Google also will be able to immediately reduce its taxes by $1bn due to Motorola Mobility's US net operating loss, and by a further $700m due to its foreign operating loss, he said.
$700m a year for eight years and $1bn immediately: call it a round $6.5bn?
The secret here is that Motorola wasn't making profits and didn't really seem to have any immediate likelihood of doing so. Therefore those tax losses aren't worth anything to carry forward: there's no likely future profits that they can be used against. But Google is, as we know, quite startlingly profitable and yes, there was a tax law change in 2009 that means that Google will indeed be able to use those losses against the greater business, not just any profits from Motorola.
Len(d)ovo us a quid?
But wait, you say: if Google has now sold Motorola then it will not be able to take advantage of those tax deductions into the future, will it? And that could be true. But do recall that we're talking about a company that employs some of the best (completely legal) tax-dodgers in the business.
It will all depend upon exactly how it has structured the sale. If it has sold the entire business to Lenovo and only kept the patent stash then yes, those accumulated tax losses will now move to Lenovo. But how likely is that?
If, instead, it has kept the holding company that owned the patents and the handset business, selling instead the handset assets out from that company, then the tax losses would stay with Google.
Yes, tax law is hugely complicated and at this stage we've no idea how the deal was done nor do we know whether the IRS will allow it. The way to find out lies at some point in the future: when Google next publishes its accounts.
In the last set there will be a number for accumulated tax losses, or provisions to carry forward. In the next set will also be that number: and if the figure has fallen substantially, then we'll know that the losses have moved over to Lenovo.
But that most certainly isn't the way that I would bet. Not from a company which, quite legally, manages to pay much less tax on its ex-United States business by routing much of it through Ireland and then to Bermuda. It would entirely boggle the mind if it didn't manage to utilise those Motorola losses inside Google. It's also worth noting that those Motorola losses aren't going to have much tax value to Lenovo either - given that it is not a US-domiciled company.
Which leads us to thinking that the actual price, nett, to Google of that patent stash was in the $1bn range. It should be said that it's not really proven even that valuable as the attempts to assert those patents seem to be failing at every turn while the $4.5bn that the Rockstar consortium paid on the Nortel stash does seem to be having some luck in getting enforced.
Toss across loss
But a $1bn loss to a company like Google is clearly rather different than a $9bn one, or even a $7bn one.
Just to complicate things further, Google has had to carry the ongoing losses of Motorola Mobility while it has owned it and I'm not even going to try and crowbar that into this calculation: and these are of course offset against those profits made elsewhere for tax purposes, so the headline loss isn't the net loss.
And finally we've one more thing to consider. Mobility had $3bn in cash in it when Google bought it. How much of that do we think is going to move over to Lenovo? "Not a lot" would be a reasonable answer there.
To this writer, it would appear that it found a bundle of assets that were worth less in their current configuration than they would be if they were split up and parcelled out to a different group of owners.
At first glance we've the set-top business, the handset one and the patents. At that first glance it would appear that Google has paid $12.5bn in total, flogged off the two physical businesses for $5.5bn and thus paid $7bn for the patents.
However, if we add in the $3bn cash (possibly retained) and the $6.5bn in tax losses (certainly some used, possibly the rest retained) we can, with a good deal of guessing I agree, make the case that Google has actually made a couple of billion profit on the deal and got those patents for free. It all depends on how good the tax lawyers are and who really wants to posit that Google employs bad ones? ®
Sponsored: Hyper-scale data management