The political and organisational difficultiesin spinning off the intellectual property (IP) or assets from a business for the first time requires the stars to be aligned.
For UK-listed semiconductor designer ARM, the astronomical alignment of strategic need, strong management and funding happened with the creation of Cognovo last year.
ARM makes money by charging between 1% and 2% in royalties for the use of its designs by chipmakers in areas such as smartphones and computers. US investment bank Goldman Sachs estimates ARM’s potential market at $50bn by 2015.
Cognovo has a similar strategy for mobile communica-tions, creating technology that allows devices to use a mix of cellular, wireless and broadcast standards – under acronyms such as HSPA+, LTE, LTE-A, WiFi, WiGig, DVB and DMB – to talk to each other across a range of frequencies, called soft modem baseband. ARM had found itself with IP in baseband after an acquisition in 2002.
Bruce Beckloff, vice-president of corporate business development, said: "ARM faced a classic dilemma in corporate development. We focus on general purpose microprocessors, and dedicated baseband technology requires dedicated engineers and marketing that does not fit the core business easily. However, baseband applications are extremely important to ARM so we want to know how our core technology would intersect with areas of changing technology."
To square the circle, ARM’s research and development team worked on soft modem for three years from 2005 to 2008 before a decision had to be taken whether to move the unit to one of the group’s five product divisions or wind up the effort.T he soft modem research was still too nascent to be turned into an established product line, and given the financial downturn after the credit crunch started in mid-2007 the most probable option was to kill the project.
However, Beckloff said: "For ARM, baseband is extremely strategic, so keeping the technology going was important regardless of whether it was done inside or out of ARM."
This led ARM to consider a third option – spinning it off. Beckloff said: "Not being active in the market is the risk for a corporation. Baseband is in 1.5 billion phones. ARM partners ship 6 billion chips a year, so it represents an appreciable portion of the total.
"For a spin-out that remains connected to the mother ship you need the strategic connection to compensate for not being directly active in the market. Otherwise you are just taking operating expense off the [balance sheet] by spinning out the team and technology."
Once a spin-out was being considered, ARM’s next challenge was finding suitable management. In this, it looked at the potential of a start-up in its local area of Cambridge, UK, run by Tony Milbourn. Milbourn, former chief executive (CEO) of TTP Commu-nications from the time it started in 1988 through a flotation on the London Stock Exchange in 2000 and onwards to its acquisition by Motorola in 2006, is now executive chairman of Cognovo, while TTP’s former chief operating officer (COO), GordonAspin, is CEO of Cognovo.
Milbourn said: "We have been a management team in cellular and wireless technology for 20 to 25 years. Software-defined modem is a new era and so we started a business in 2009 after TTP’s sale, but needed a computing platform and ARM was our preferred partner as it has great technology and is just up the road.
"Out of the blue, Warren East, CEO of ARM, called and asked me to have a chat to Bruce to get connected up. He understood our direction and the merit of putting his internal unit and us together. At the same time, we needed to be bigger as we had enough cash to keep the lights on but not much else. ARM had already invested 100 to 200 man years in its programme and was looking at other ways of getting value."
Beckloff described it as a "marriage made in heaven". He added: "This was a relatively easy spin-out as we had a great engineering team and technology and there was a great management team out there. If ARM had relied solely on internal managers it would not have been as easy to do."
Mark Radcliffe, senior partner at law firm DLA Piper, said: "Spin-outs play a very important role in a company’s larger innovation strategy. The keys to success are understanding the need of third-party financing, which will mean the need to give up control so the nascent company can work with third-party investors and customers. Such release of control can be difficult for corporations, but it is an inevitable result of the decision to spin-out the technology. Most investors will wish to have exclusive rights and the technologists to go along with it."
This is the third requirement for a spin-out – arranging its funding and share structure. ARM provided the majority of funding in Cognovo’s initial round, estimated to be about £5m ($8m), in return for at least 15% equity, a further convertible loan, and a board seat for Graham Budd, ARM’s COO.
Beckloff said ARM had no right of firstrefusal to buy Cognovo or ability to dictate future investors although it would encourage against close competitors. He said: "Our investment objectives are, first, active management to participate in the ecosystem, second, a window on technology, third, to increase overall ARM group revenues, and, fourth, a financial return on the investment.
"Our investment structure allowed ARM a seat at the table but was below the accounting threshold that then requires a business to be consolidated into the investor’s accounts as a subsidiary."
Milbourn said: "On the financial structure, ARM recognised we needed good incentives, while we as the management team were mature and not stupidly aggressive. ARM also knows other investors will come in, but if the business goes belly up the technology will revert to ARM."
However, the equity, and the convertible debt that is in place to help maintain ARM’s interest in future rounds, allows the chip designer to remain close to Cognovo. Milbourn said: "ARM is helpful in three way – first, money, second, we can leverage its sales force, particular in Asia, and, third, Graham Budd is on our board and if we need something from ARM often he can help.
"There is ARM technology around our core Modem Compute Engine and so the technical relationship is very smooth. We would not have it so smooth or deep if ARM had no equity."
However, Milbourn said baseband technology took time to develop and was a challenging area, as the IP licensing business model paid royalties three to four years from a licence being agreed for a new product. As Milbourn said: "It is a big boys’ game and we are little."
This makes finding external funding difficult. Cognovo has applied for grants in Belgium but independent venture capital (VC) firmshave shied away from long-term technology investments, Milbourn said. This left corporate venturing as the option. He said: "Corporate investors can see the strategic advantages and return on investment from Cognovo having £100m in turnover by 2020 for £25m investment. The issue with corporate venturing is it usually needs others to set the round terms, which takes us back to financial investors.
Beckloff said: "Cognovo was difficult for financial VCs to invest in as the time to revenue and early investment in this industry is significant. Cognovo was only established in 2009 and first revenues are modest in 2010/11 until significantrevenues start in 2013/14. Start-ups in the semiconductor industry are not very appealing to VCs currently, which is why Smoothstone (now named Calxeda) which ARM also invested in with several other financal VCs was the first appreciable series A investment for a fabless [non-fabricating] semiconductor company for almost two years.
"But this absence of money coming in and fewer start-ups means returns will increase as demand for strategically important acquisitions by larger companies remains constant."