ARM Holdings – Powering the Smartphone Revolution
ARM Holdings is a technology company with a narrow focus and worldwide impact.
ARM Holdings (hereafter referred to as ARM) is a quiet leader in the smartphone market. The company produces no physical chips nor sells any phones yet designs the underlying architecture that powers 90% of all smartphones [1]. The company started in 1990 and has a current market capitalization of $23 billion. They produce constant innovation that requires little physical assets and almost no capital expenditure.
Business Model
ARM earns money principally by creating and licensing its intellectual property (IP) to chip manufacturers, largely in embedded systems (smartphones, automotive, tablets, smart appliances) [2]. This IP is the architecture that describes how a processor will handle its most fundamental operations. There are infinite ways in which to design a microprocessor and creating even more efficient versions is an area of continual research. ARM has made its focus producing an energy efficient architecture that is ideal for applications, like smartphones, that have a limited battery. Compared this to desktop processors produced by Intel and AMD, an ARM architecture will have less raw power but last significantly longer.
ARM captures value by licensing these designs to chip and System on Chip (SoC) producers who combine the IP with additional functions to create a full processor that can handle all the processing (pictures, calls, video, graphics) that a modern smartphone requires. One such partner is Qualcomm who produces the Snapdragon line of processors, an example of which is shown below:
The above circled portion is designed and licensed by ARM, customers are willing to pay for this portion because of significant cost savings in avoiding R&D of their own. At the same time they reduce their own design risk by starting from a thoroughly tested baseline. These customers will pay a licensing fee to develop using the IP and a royalty payment as each chip is sold:
Operating Model
ARM works in a very interesting niche within the embedded processor market. There are few barriers to entry in regulatory and capital terms. Furthermore, because of the rapid pace of advancements in smartphone technology, existing designs must be continually improved on or even replaced entirely. Given this context, one would expect their high market share to be continually at risk. ARM however has built a defensible position through its ubiquity and cost efficiency. By the company’s own estimates a firm would need to spend $100 million annually to develop the architecture needed to keep pace with industry.
The company accomplishes this by investing heavily, and continuously in R&D, their 2014 income statements show £224.2 million R&D expense against £795.2 million in revenue, or 28% of the total (IFRS). Compare this to an average R&D spend of 7% by computer and electronics firms and 16% for software and internet companies [4]. One key indicator that ARM measures is the percentage of people employed who are engineers (72% in 2014), they do this to ensure they money they do spend goes directly into the product[3]. Too much other overhead immediately reduces the value proposition they bring to their partners [3]. In addition to hiring a large portion of R&D personnel, ARM makes Human Resource (HR) management a top priority. Employees are given extensive and continual training, their internal “ARM TV” channel added 300 educational videos in 2014 alone [3].
Finally, ARM aligns their R&D development with their customer’s needs. ARM knows that they can be disintermediated at any time a customer decides to do their design in house. To keep their customer tightly aligned with their own development, they host an annual conference (the ARM Partner Meeting) that brings in 600 clients to discuss development opportunities [3].
Conclusion
ARM has been a clear success story in using a company’s operating model to drive their business. They make a clear value proposition to their customers, and back up the pricing of their products by massive R&D investments. The operating model focuses on being asset light and investing resources into attracting and retaining people and enabling them to be productive.
Going forward, there is some concern that if the market consolidates enough, the remaining producers will be at such large scale that ARM’s solution is no longer cost effective. ARM can continue to leverage its effective operating model to head off this risk. By continuing to be thoroughly invested in R&D and keeping their large customer base happy, they can keep the licensing fees (even with profit) lower than a customer would be able to produce on their own.
[1] http://www.forbes.com/sites/darcytravlos/2013/02/28/arm-holdings-and-qualcomm-the-winners-in-mobile/
[2] http://www.arm.com/about/company-profile/index.php
[3] ARM 2014 Annual Report, http://phx.corporate-ir.net/phoenix.zhtml?c=197211&p=irol-reportsannual
[4] http://www.strategyand.pwc.com/global/home/what-we-think/innovation1000/rd-intensity-vs-spend-2015
[5] http://www.arm.com/about/company-profile/milestones.php
[6] http://www.ubergizmo.com/2015/02/snapdragon-810-fully-revealed-benchmarks-architecture/
[7] http://www.anandtech.com/show/7112/the-arm-diaries-part-1-how-arms-business-model-works
Really interesting article on a fascinating company! As the smartphone industry consolidates around a few large players (Apple, Samsung, etc), do you see ARM as being able to stay independent and just license out IP? Is possible they may be acquired by one of these smartphone producers or even a large chip manufacturer that doesn’t have a strong R&D capability (Global Foundries)?
Great question, this is a huge concern I have for them in the future. One of the major fabless semiconductor producers for smartphones is Qualcomm, their major product is the Snapdragon line of processors which have been largely ARM based. I just looked into the most recent launch, the Snapdragon 820:
https://www.qualcomm.com/products/snapdragon/processors/820
This is run on a new “Kryo” CPU that Qualcomm developed internally. As ARM this is very concerning, I might be forced to give up some of the high end if Apple, Samsung, and Qualcomm both go away from the architecture. There is still plenty of space in the lower/middle market where manufacturers are unwilling to invest millions in architecture R&D.
This is a really interesting model. To me, it seems they are in the business of being the go-to option for outsourcing highly specialized R&D. As chips and other pieces of hardware get more and more powerful more and more quickly, I wonder if the same will eventually extend beyond hardware to the design of architecture embedded. At some point, will marginal gains in energy efficiency be worth the investment on the design side and put pressure on ARM’s model?