FIFTY BILLION of anything is quite a lot, really, and it’s no different with microchips. And shortly before Financial Director met ARM Holdings CFO Tim Score in February, the company announced it had shipped its 50-billionth chip since the company launched in 1990.
What’s even more extraordinary is that more than ten billion of those microprocessors were shipped in 2013 alone. So in case you thought that the chip was already ubiquitous and demand for it exhausted, there is clearly a lot more of them being churned out every day. In an electronic world dominated by companies from the US, South Korea and Japan – and all of them seemingly using China as their manufacturing centre – this is a remarkable British success story.
Correction: we said ARM Holdings had “shipped” 50 billion chips. It hasn’t. ARM doesn’t actually make the things; it designs the technology that goes into them. It then licenses that technology to the world’s leading microprocessor manufacturers which churn out the chips, paying ARM royalties for the right to do so. Mobile phone makers have been the core market, with 95% of mobiles (including smartphones) using ARM-based chips, but this sector now represents just under half of the company’s revenues as its technology begins to be used in other applications, ranging from automotive applications to smart TVs to the so-called internet of things.
If anything, that makes ARM an even more remarkable success story – a designer of technology that the world comes to Britain to get, because it can’t get it anywhere else. Revenues of £715m (a rise of 24% year on year) rather prove the point, with a research and development spend of more than £200m and cash in the balance sheet amounting to £700m-plus – as well as a market capitalisation of £14bn. In his book Made in Britain, BBC presenter Evan Davis described Cambridge-based ARM as a “hugely important” British company, mentioned in the same breath as GlaxoSmithKline.
So what has the finance function’s contribution been to this success story, particularly over the 12 years that have passed since we first interviewed CFO Tim Score (back then, 2001 sales were £146m and the market cap was under £3bn)?
“When I joined in 2002, it had been in a very fast growth phase and lived from quarter to quarter and had an unbroken record of success,” Score recalls. “But planning was more about making sure we were going to hit the next quarter [earnings target], and maybe the one after that if we were really being far-sighted.”
A searing experience for the company – and for Score – came when he was just a few months into the job as this fast-growing company couldn’t defy what was happening in the dotcom bubble collapse, and it announced its first and – so far – only profit warning.
“The big thing we learned from that was the planning of this business needed to be much longer-term than quarter to quarter,” Score says. “Now we have got pretty rigorous 12-month and five-year planning processes. This is actually a business which is capable of being forecast in a long-term view.”
Given the volatility and rapid rate of innovation in technology markets, the idea of a component designer such as ARM being able to plan and forecast that far ahead seems unlikely. But here’s how the business model works:
• Every two or three years, ARM introduces a new generation of chip technology.
• As ARM designs the technology that goes into microchips, chip makers then pay licence fees to be able to access to that technology early on. Often, this happens even before the design is finalised, giving chip makers some input into the design as well as a ‘time to market’ advantage. ARM, by the way, would typically expect to recoup its R&D costs through licence fees alone. Note, then, that ARM is generating cash long before anyone has made a single chip or sold even one smartphone.
• Once the manufacturers start sending chips to the big-brand manufacturers, they start paying ARM royalties.
• Here’s the extraordinary part: the technology that was used in mobile phone microchips in the mid-1990s is still generating revenue for the company – because that technology may not be smart enough for a phone, but it’s perfectly fine for something like an anti-lock braking system in a car.
“Different end-markets migrate to different levels of sophistication in technology at different periods, which means that the life cycle of an ARM processor in terms of its ability to generate royalties is very long,” Score explains.
So ARM is making money from chips that haven’t been designed yet and it’s making money from chips that were designed 15-20 years ago.
In fact, deals signed more than ten years ago are generating about 40% of ARM-based chip shipments. Because chip makers have paid big licence fees, it’s in their interests to find more and more uses for ARM’s chips.
And that’s why so much of ARM’s revenue is forecastable, as Score says. For the rest, “finance owns the planning and forecasting process and therefore owns the interaction with the other elements of the organisation that one needs to build a five-year revenue forecast”, Score says.
It’s a multi-disciplinary process, involving sales, R&D engineers and legal. The revenue recognition rules are very complicated, and the market has quarterly-earnings expectations. Finding out when ARM can book the licence fees it’s generated is an important question: “It’s the achievement of the contract milestone and the delivery of the technology that frees up the revenue,” says Score.
Luck and judgment
Finance has to know what’s going on across the business to be able to know if its targets are going to be hit.
“Are they on track to achieve effectively the percentage of completion that they need to achieve? Where is each deal? Have the commercials been agreed? Where is it in the legal contract stage? What are the risks in the signatory loop at the customer? All of these things need to be factored in all of the time,” Score explains.
When you get to the end of the quarter and it’s suddenly Friday night, a deal signed by the end of the day is good – but if it defers to the following Monday morning, then that’s not so good. Since that one profit warning all those years ago, ARM hasn’t put a foot wrong – either meeting or beating expectations every single quarter.
“It’s part luck, part judgment, part good forecasting,” Score explains. “If you’re hitting one quarter and missing the next, the market has little confidence in your medium- to long-term forecast.”
It’s the predictability of ARM’s earnings over the very short to the very long term that has earned it – truly earned it – a market valuation that would have been regarded as stratospheric 12 years ago when Score first landed at ARM.
“ARM has grown earnings at 22% over the past ten years CAGR and has the look and feel of an entity that can continue to do that,” Score says.
“ARM is now valued by the finance community on a fundamental basis – ie, discounted cash flows based on a very reasonable range of assumptions about the likelihood of forward licensing and forward royalty, whereas it was all about how tolerable it was to have a multiple of 30 times revenue and 100 times earnings when I arrived.”
Apart from supporting an enormous – though justified – stock market valuation, it’s hard to imagine that the finance function’s relentless ‘finger on the pulse’ process does much to help the engineers and R&D boffins. In fact, it’s easy to imagine that the quarterly-earnings treadmill must drive ARM’s cartoon Dilberts absolutely nuts.
“When I first arrived, I’m sure some of my colleagues probably thought, ‘Who is this short-term financial guy who seems to be totally motivated by the next quarter’s earnings per share?’” Score concedes. “There’s an element of quarterly reporting being an artificially short period of reporting for a business like this.”
But he’s won them over. “There’s a distinct benefit of discipline, structure, execution, delivery on time,” he says.
“Over the past ten to 12 years, colleagues have seen that it actually works to balance the desire to invest in all these exciting projects with a need to maintain a trajectory of improving profitability and to maintain a track record of doing what you say in a consistent fashion. The organisation has actually realised you can invest in everything you need to invest in and have an improvement in profitability.
“Ten or 12 years ago people might have said, ‘Why do we need to worry about margin expansion or eps growth? Let’s just keep investing and it will be glorious in the end.’ It might or might not be glorious in the end but you only find out at the end.”
So even though, technically, ARM Holdings doesn’t have to report quarterly, it does, partly because every other IT business does, partly because it helps keep that focus and discipline in the company, and partly because “if you don’t report [to the market], they will just speculate for you. So it’s better to be in control of it,” Score says.
Before joining ARM, Score’s CV was an impressive succession of roles at the likes of Arthur Andersen, engineering group BTR, Lucas Industries and others, all good stepping stones about three years apart. Back in 2002, we teased Score that he perhaps wasn’t going to be in ARM for the long haul. “I do intend to be at ARM for a while,” he told us. So he was right, and we were wrong.
“ARM’s got lots more opportunity ahead of it,” he says today – and he has given the company the forward visibility of earnings to know that is true. “I don’t see any more attractive FD jobs around than this one – let’s put it that way.” ?
IN BLACK AND WHITE
2002 – present CFO, ARM Holdings
1999 – 2000 Group FD, Rebus Group Limited
1997 – 1999 Group FD, William Baird
1996 – 1997 Director / group controller, Lucas Industries / Lucasvarsity
1992 – 1995 Group financial controller, BTR
1991 – 1992 Group chief accountant, BTR
1991 – 1992 Arthur Andersen