Electronics Weekly Electronics Design & Components Tech News
According to Bloomberg it would now cost $12.5 billion to buy ARM – a lot of dosh for a company with $600 million in revenues.
On the other hand the tech industry has a lot of dosh – apparently companies are sitting on a collective $276 billion of cash.
Even so, with a bidder having to pay 56x EBITDA, or 87x profits at the current ARM share price of £5.30, it seems unlikely that anyone would shell out.
And that includes Apple, one of ARM’s founding investors, which is sitting on $27 billion worth of cash and short-term investments.
Of course, the electronics industry would be outraged if a bid was made to take over ARM because it gives the world a compatible processor at a fair price.
Intel would love to give the world a compatible processor – but history shows it would never be at a fair price.
So ARM is a good thing for the industry. Would the industry mount a collective counter-bid if ARM was threatened? It’s possible.
If a takeover happened would ARM’s licensees go elsewhere so destroying the value of the company?
It’s possible, but ex-ARM licensees might find it tricky to use ARM code on a different processor architecture, and it might be too late to economically write enough code on a new processor to replace existing ARM code.
So, will anyone buy ARM?
The 150% rise in ARM’s share price could suggest either that a predator is going after ARM, or simply that analysts have over-estimated the tablet market.
Magnifique, Frederic, aux ARMs!
Sept 2nd : Intel is rumoured to bid on ARM
Sept 3rd : ARM top execs secretly launch a new company (possible name : ReArmed) with architectural licences for every existing ARM core
Sept 15th : Intel buys ARM, making ARM execs and head engineers rich enough to quit and join ReArmed
Sept 16th : Apple, Samsung, Panasonic, Toshiba, Philips and Sony all issue press releases explaining their commitment and confidence to ReArmed
Dec 9th : ReArmed goes public, stocks skyrocketting
Hadn’t thought of that, Stooriefit, it wuld be an interesting court case to see the directors of a PE-owned company sued for breach of their statutory duties becuse they had implemented the usual PE strategy of loading their company with debt, selling off the sellable bits, firing lots of employees, cutting all the long-term spending like R&D, selling the IP and then trying to make a quick return by IPO-ing it after two or three years. You’d have thought a good lawyer could make a case for saying that was an abuse of directors’ responsibilites as the legal guardians of a company’s assets. However I don’t know what the law is relating to the directors of privately owned companies. Maybe they don’t have an legal obligations. On the other hand, if they have publcly tradeable debts like bonds, which most do, then it’s possible they do have legal obligations vis a vis the company’s assets. If so, then a court case for misconduct would be fascinating!
On not so sober reflection, maybe the analysts do understand that “Gross Margin” doesn’t equal “Costs of a seat at the table” and don’t care. As our object lesson in greed’s capacity to shout down judgement Ed shows us, all the PE brats care about is a sharp exit. Gross Margin is a great indicator of controllable costs (you can sack all the smart folk, turn the lights out and still sell the IP for another couple of years before you become a Zilog).
This, of course, is formally illegal for plc directors to do, as they have a duty to manage the business as a going concern – again from the most accurate source of instant information on the planet: “The first director’s duty under section 171 is to follow the company’s constitution, but also only exercise powers for implied “proper purposes”. Prior proper purpose cases often involved directors plundering the company’s assets for personal enrichment,[97] …”
Of course the boys holding Ed’s strings aren’t the ones who would be up before the beak but Ed, as he is formally the man carrying the can (3 years suspended sentence and disbarred from being a director for 5 years?). It would never happen anyway – or maybe it does in a future instalment of Ed’s diaries?
Samsung, Apple, TI, etc need to take a big enough position in ARM to make sure the PE guys can’t afford to rape the business in this way, which would kill their long term technology partner and cause them an enormous headache. Maybe this is what the valuation is based on?
I hope so too, Stooriefit, but if you look at the valuations private equity analysts put on NXP and Freescale, and the valuation banking analysts are putting on Facebook, you get the impression that a financial analyst couldn’t put a reliable valuation on a five pound note. Excellent points though, I hope your message gets across.
From Wikipedia (so it’s got to be true right?) “Gross margin, gross profit margin or gross profit rate is the difference between the sales and the production costs excluding overhead, payroll, taxation, and interest payments.”
That doesn’t sound like a figure which reflects the R&D effort required to implement these cunning architectures, or to keep your R&D efforts rolling so that you’re still in the game 3 years down the line. I doubt it includes all the test kit production and development costs, or the costs of putting together the tool chain to support your product. It specifically excludes paying the bright folk who design and implement the stuff and the expensive tools they need to do their jobs.
Your comment about sustaining 90% gross margins upsetting partners is true – much more sensible to load the service costs, but this 90% is an accounting anomaly, not a true reflection of the costs ARMs faces just to be be in the processor design business. I hope the analysts realise this.
Well I rather wonder why people say ARM underprices its cores, Geoff, after all ARM’s success depends on the perceptions of it held in the minds of its partners, and ARM currently makes gross margins in the 90s of %. Any more than that and the partners might take umbrage that ARM is over-pricing its cores. According to BusinessWeek, commenting on ARM’s most recent results, “ARM retained about 94 cents per dollar of revenue after subtracting the cost of goods sold, the highest gross margin of 29 global semiconductor companies with market values greater than $5 billion”.
Well a little bird once told me that ARM often struggle to get licensee’s to take on new core licenses because they got such a good deal on the original core. If they need more processing power then can add more cores of the same old type rather than shell out for the newer faster cores.
However that’s exactly why the advance of the smartphone and even the tablet could be so good for ARM – because such systems are app hosters not embedded systems, and thus the consumer expects more and more processing power from the processor running their app.
Now I don’t know if ARM has exploited this potential fully to raise their per core license fee – but I suspect that while they got a better price they were probably inhibited by the markets demand to not stray too far from the original core start point.
If I was an analyst this is what I’d look most closely at to determine the potential market valuation for the company.
Wise words, as always, Stooriefit, thank you
ARM probably under sell their processors – or have traditionally, in order to get the design wins. Maybe the analysts have got a sniff of this tide turning given that ARM have, through this extraordinarily low price for what is a fantastic architecture, won the lions share of a number of potentially high value markets. Lock in is starting to look more certain and then it could be time to turn the screw.
I’m not so sure. Firstly, I doubt if licensee lawyers allowed any more than a very parsimonious price escalator into their contracts. Secondly, in my experience selling hardware + software into blue chips, it is much easier to get the buyers to swallow your margin on hardware & service than on software. They see the cost of sales of software widgets as essentially zero, whereas they appreciate that there is a significant cost of sales of physical widgets & providing service. If you are selling into bluechips load the hardware, and service contract prices, make the software licenses essentially free and you won’t get your margins destroyed in the haggling. ARM essentially have a software (the architectures) strand and service strand – and their figures show the service work to be more lucrative as I would expect.
Intel cover almost the whole semicon value chain, and a much firmer lock in, with no license unpleasantness – their hardware (in the form of their processors) is a crack-proof dongle for their software (in the form of their architecture) such as to make licensing unnecessary. Because they don’t need licensing they can ramp their prices as and when they know their OEMs have nowhere else to go, notwithstanding any volume supply agreements they might enter into.
Either there is a phoney bidding war going on or the analysts really don’t understand the semiconductor value chain. Whatever, ARM’s independence is worth $12.5B to the semi-con industry, but to a single buyer from within that industry they are worth more like $3B.
Interesting to hear you say ‘devices are too expensive’, Chris, maybe the semi industry has woken up to the fact that over-capacity = low prices = barely profitable semi industry. Today’s under-capacity may be a sign that things have changed and ‘too expensive’ is the new normal.
I don’t particularly think it’s going to get taken out for the reasons you state – price and logic.
I suppose Intel (if they were allowed) or AMD (if they have the money) might pay a premium in an attempt to get control which might get them some more influence over the way the market evolves, but its a pretty penny!
Does make much sense for an end-user to buy the company as they can just licence the technology anyway.
My point isn’t really about the fact that its cheap for the user, its that ARM don’t get proper value from their product. Though speaking as an end user, devices are already too expensive so perhaps I should be quiet;)
Yes, Chris, there’s a lot of buzz over this but at $12.5 billion! It would be a huge purchase for even Intel which has somewhat depeted its cash coffers recently with acquisitions of questionable value-add potential. As to price – well we all know the value of a sq cm of silicon – so what is a fair mark-up – Intel’s or ARM’s? I would respectfully submit that from the point of view of the profitability of the electronics industry, ARM’s value proposition is so infinitely superior to Intel’s that anyone using x86 when there’s an ARM alternative would be bonkers.
Some speculators clearly think it’s going to be taken out hence the bonkers price movements over the last year or so. The fundamentals don’t justify the price, tablets or not, windows or not.
Have to take some issue with your claim they offer a processor at a fair price… more like a bargain basement price. I maintain ARM do a very poor job of getting value from their inventiveness.
‘Industrial bids’ would no doubt get into all sorts of cartel issues. Of course an individual bid may end up being anti-competitive.
Bet a few of those ARM management who sold at £3/share are feeling a bit sore!
I’m not saying ARM’s fortunes are based on tablets, fabrice, I’m merely pondering on the reason for the 150% ARM share price increase in the last 12 months. As you rightly say. I’d love to see the end of the Intel monopoly and I hope it happens in the way you describe, but Intel won’t be sitting there waiting for it to happen, they’ll have hundreds of very clever strategists working on how to retain their market position via stratagems like MDF. MOAP, FUD, arm-twisting and downright bullying. They’re not going down until they’ve pulled out all the stops. But I hope they do – they just suck too much of the profit from the chip industry.
“have over-estimated the tablet market” …. well, why restrict ARM to this market ? Which processor power your cellphone ? And with Window on ARM, the potential are just huge. You always complains about the Intel monopoly, and you don’t realize that it will disappear in the coming years. With Window8 on ARM, there is 0 reason to have a x86 preferably to an ARM … and the ARM price will be so cheap … and the performance will be so “good enough”.
An please, stop with the Intel process advantage …. because you just forget the x86 architectural disadvantage … an instruction decoding scheme which is just insane … see how well ATOM fare in cellphone, see the Nokia/Megoo fiasco, see the Larabee fiasco … a company whose engineers seriously state that they have envisioned several instructions sets for Larabee, and that finally the x86 one was the best suited for the task, is off ground by a long shot.
Empires are mortals … as anything we know …
(beside this, Intel as not process advantage in low power, the kind of process which will matter the most)