Subscribe
Semiconductors

Qualcomm close to decision on break-up

Results of strategic review may be announced in the coming week, but chip giant must stop short of splitting its two businesses

Since July, Qualcomm has been going through a strategic review process, sparked by rising market pressures, the firm’s worst stock performance since the global crash, and the activist investor Jana Partners. Crucial decisions should be made at the company’s board meeting this week, but insiders believe it will stop short of a complete break-up.

Some investors have pushed for the licensing business to be separated from the chip technology group, which of course would dismantle the two-headed business structure which has made Qualcomm such a powerhouse. But, even if a recent string of antitrust probes round the world seems likely to signal reduced licensing revenues and significant changes to business practices, there would be huge risks attached to a complete separation.

Two sources told Bloomberg that the management team was keen to keep the company united, even though a break-up would appease competition regulators, such as those in the European Commission, US, Taiwan and Korea, which are currently conducting investigations of Qualcomm’s licensing terms or pricing. But the downsides would be even greater. It would be extremely hard to separate two such closely entwined units and the process of doing so would only enable Qualcomm’s rivals in China and Taiwan – already gaining ground in price-sensitive markets – to step up their challenge to the mobile chip market leader.

Qualcomm has stayed ahead of such competitors largely through the continuing strength of its engineering, which enables it always to be in the forefront of new evolutions of mobile technology. That engineering expertise also feeds into the patents mountain, of course, and the cash from patent royalties – $6.8bn before tax in the last financial year – helps to fund the huge levels of R&D.

That virtuous circle should not be broken. Yes, licensing revenues are under pressure from antitrust settlements, market forces and the fact that Qualcomm has less of an IPR stranglehold in 4G than in CDMA-based 3G. But they are still significant and profitable, and without them the chip business would be far less well equipped to remain at the cutting edge at the same time as dealing with price wars. And the picture has not all been gloomy with regard to patent fees this year. When Qualcomm came to a settlement with the Chinese authorities earlier this year, it agreed to lower royalties in the country, but these are still based on the price of the handset, not the price of the chip, as China had originally demanded. The latter approach would have reduced royalties by a far greater amount, about 20 times but in fact Qualcomm agreed to a $975m fine and accepted lower royalty rates on single-technology modems, and to calculate the fees based on 65% of the device’s sale price.

“These are two businesses that would not do better separately,” Daniel Morgan, a fund manager for Synovus Securities, told Bloomberg, reflecting the views of many on Wall Street. “The latest and greatest argument is everyone has to split. It’s getting a little out of hand.” Such a move would make far less sense for Qualcomm than for other recently broken-up companies such as Hewlett-Packard and eBay.

On the opposite side of the debate, investors like Mike Green, a fund manager at American Money Management, say break-up is overdue. Green said: “Trade commissions all over the world view Qualcomm as a company that colludes with itself. Qualcomm should spin off the licensing business to unlock shareholder value and they should have done it many months ago.”

Qualcomm’s share price has fallen by 36% this year, meaning its market value is $71bn, down from a peak of $130bn in 2014, when it overtook Intel as the most highly valued US public chip company. Such developments sparked the strategic review, announced along with 4,500 job cuts in July.

So Qualcomm may have reached the end of the road in terms of its unique business model, and it will have to make some very tough decisions, both to disentangle its two main revenue streams and to accelerate its move into new markets which can offset the slowdown, and rising competition, in smartphones. But that does not mean it should break up entirely.

That proposed solution reflects purely financial, and very short term thinking. Qualcomm is an engineering company, and only by thinking of the longer term impact of inventions – not just quarter-by-quarter figures – and plowing cash into R&D not just shareholders, has the wireless industry got where it has, in terms of technology leaps, profits and long term growth. As Google CEO Larry Page once said to someone presenting an MBA-based solution to a problem: “Go sit with the engineers”.

Yes, at the point of break-up, if estimates by analysts such as Arete prove correct, Qualcomm could see its chip division valued at $74bn and its patents arm at $87bn. In the current frenzy of semiconductor M&A, the silicon arm might then be an attractive acquisition target for a rival like Intel, or it could be better placed to make its own purchases. But if both halves remained independent, for how long would they retain their new value? Qualcomm has delivered significant growth and value for shareholders over the years, and driven many trends for the broader industry, because of its combined model.

Qualcomm would be throwing the baby out with the bathwater – in order to please activist investors and antitrust authorities in the short term, it would never again be able to achieve the success that has come from the crossover between product development and patents – both based upon its engineering excellence and huge R&D investment.

Those are the key to the argument, and make Qualcomm very far from being a troll. Yes, the IPR licensing and the chips should not be intertwined in the commercial sense, but at the R&D stage, it makes perfect sense to have two ways to monetize a huge engineering investment.

Qualcomm will have to change its licensing structure radically anyway, and adjust to the impact that will have on margins. It may not be able to resist Frand and patent pools forever as the industry moves towards royalties based on component, not device costs, and as 5G looks set to include more patents from the open world than from 3GPP.

But it will survive these transitions better if it maintains its inventions engine and allows it to feed both business streams in parallel. For a putative licensing-only company, generating patents in a vacuum is hard for engineers and the business model. Nokia knows this – its Techologies unit, which houses its huge IPR arsenal, is already developing real world products as well as filing for patents. Even with the tightest of licensing arrangements between the separated Qualcomm units, they would lose creativity and commercial motivation. Or if the chip company retained most of the engineering and R&D, that would devalue the licensing arm and restrict its ability to grow in future.

Qualcomm has proven itself able to make significant change in the past, and even to attack its own sacred cows. It has open source activities now, notably the AllSeen Alliance, which shows it learning some lessons from Google about alternative ways to achieve market power. It has embraced the old enemy, WiFi, with the acquisition of Atheros. It will certainly need to be even more open to new thinking now.

VPS hosting