Close
Close

Published

Intel pops champagne corks over top EU Court anti-trust reversal

Intel looks to be in the verge of winning its anti-trust appeal with the European Union, courtesy of the Court of Justice of the EU. To be fair, all the court has done so far has passed the decision back to the lower General Court, requesting that it once again looks at Intel’s arguments on whether its chip rebate system actually affected competition.

That’s like code for saying that it did not. Let’s remind you of the case and then let’s look at the implications for similar cases coming before the European Commission.

The case was original seen as proven by the Commission in 2009 and then the General Court agreed with the Commission in 2012, and Intel went on to file an appeal.

The key thing to remember in anti-trust cases is that what is fine for a small operation which does not have market control, is sometimes unfair for a dominant force and while this was not brought in our video sector, it will have far reaching implications for our sector.

The European Commission felt that all it had to prove was that a dominant force was offering discounts for exclusivity – which it was, at the expense of AMD which makes Intel compatible chips. Without those chips PC and laptops were virtually a 100% monopoly for Intel at the time. Hewlett-Packard was contracted to buy 95% of its desktop PC chips from Intel, leaving just 5% as a maximum for AMD, unless it wanted to get a big bill from Intel, repaying that discount.

The General Court reasoned that although Desktops made up around 28% of the Intel compatible chips that HP took, that limiting all rivals to 5% of this sub-market was still enough to distort the market. Similar deals were in evidence with Dell, NEC and Lenovo.

But this week the higher court has pointed out that the case law that the European Commission relied upon was based on cases where proof was offered that the market had come to harm, while the Commission thought that was not necessary this time and that tying people to exclusive arrangements and being dominant was enough on its own. At Faultline Online Reporter we had to agree with the General Court – it’s not like the pharmaceutical cases this law was based on – PC software will only work with a chip that has the X86 instruction set or emulates it in some way – whereas a similar, but different, drug will have similar effects on patients, but can have a very different construction.

But the superior court reasoned that these exclusivity deals only affected 14% of the market, a detail that has never come out before, and that this was in-sufficient to pervert the market and create what it calls an anticompetitive foreclosure effect. It makes the point that if you cannot avoid Intel, and none of the PC suppliers can, this makes the market even more tricky, but still says that the Commission has not proven that harm has come to AMD. The Commission relied on something called the AEC test, meaning an economic test based on an equally efficient competitor (As-efficient-competitor or AEC).

The key behind this test is to differentiate between businesses conduct that is part of a healthy competition and that conduct which is part of an abusive strategy of the dominant company. To cut such a competitor out of business the dominant company has to be shown to price below its own costs.

The Commission built a case around the average avoidable costs (AAC) and it showed that Intel in those 14% of deals did not cover even its unavoidable costs, which showed that it was sacrificing profits to kill off a rival. It sounds convincing, but the Court of Justice was not convinced.

Perversely the Court of Justice of the EU has now decided that it has not proved this because 14% of the market is too small to make any difference. It even says in its summing up that selling below cost is a perfectly normal strategy for many companies to limit competition.

At stake for Intel is a $1.3 billion fine, and the lower court must now re-examine the case, and no-doubt the Commission will provide yet more evidence. The likely result will be that Intel pays nothing because it has virtually instructed the court to find that Intel’s rebates to customers were not anti-competitive.

So presumably now, if Google put in its contracts that customers who used its search engine in a set top or a phone, will get lower Android prices and more integration help than those who refuse to, then this also would not be anti-competitive. Or if Broadcom said to set top makers that if they used MaxLinear or Intel chips inside set tops, they would get lower discounts, that would also be okay.

The biggest worry is, what if the European Commission gets weary from trying to police the world’s markets, something that the US Justice Department should perhaps be more active in. If that happen then we lose the ability to pursue antitrust cases such as those in play against Qualcomm, Apple and Google.

The case against Apple is based on Apple having better terms with MNOs than all the other smartphone rivals in terms of promotion and prominence – or else they must pay more for the devices. That’s why, when you are perfectly happy with your Android phone, you get constantly harassed by your MNO asking you if you’d like to switch to an Apple iPhone.

If this decision stands, then Qualcomm could discard its habit of selling all of its patents in one license, yes or no, frand or not, and instead insist that if you used someone else’s chip, you get lower volume discounts and make just as much money.  It is, in effect, asking for trouble.

There is a bit of naivety going on here – with the senior court saying that by Intel limiting these deals to 14% of the market, there is no unseen shadow effect – the idea that no-one has said something to the effect of, “If you buy too many AMD chips in this other sector, we will bring in the same policy there, and cut your discounts,” but with none of that threat being written down in a contract.

Such effects are sure to be operational here, and everyone in the market is conditioned to how Intel operates, believing it will be the market leader in the PC and server market for the foreseeable future. Sourcing some cheaper AMD chips is perhaps just to remind Intel it has competition, but you don’t want to do it to the extent that Intel gets upset with you.

However in the server markets already, and of course in other segments of the device market, ARM has a presence, and in phones, a stifling one. The threat to Intel in the PC market will come from Chrome devices, tablets and smartphones limiting PC usage and from software being able to work on ARM server farms in the cloud, and therefore being compatible for many PC operating systems such as Windows. Would the court allow Intel to sue ARM for competing with it on laptops, just because it runs the same software? And suddenly we worry that anti-trust legislation somehow turns into legislation that supports the creation and preservation of monopolies.

Let’s hope the Commission builds it case against Intel and not only sustains this fine, but asks for another, larger one for the delay and interest. That’s what the Commissioners of old would have done.

Close