Brit chip biz ARM legs it to Softbank for $32bn British semiconductor designer ARM is about to be acquired by Japan's SoftBank. The New York Times reports the deal is imminent. The FT has put a value on the deal: US$32.4bn. That's a premium of about $10bn over ARM's market capitalisation. SoftBank is a sprawling company that operates a telco in its Japanese home but is …
Indeed.
Unless ARM need investment I don't see what Softbank is bringing to the table. The risk is that when one of the many other Softbank business units is haemorrhaging cash and dragging down the bottom line, ARM could be sacrificed.
ARM losing its independence sounds like a very bad trade. I wouldn't be surprised if Apple end up as owners of ARM (again) longer term, and that will be very bad news for everyone. Seriously, seriously bad.
If worse came to worse, Apple might be interested in negotiating to acquire a perpetual architectural license. That would allow them to essentially 'fork' their own version of the ARM architecture to meet their needs, if they felt like there was a risk ARM wouldn't continue on the way it is now. They are not going to buy something that's apparently worth $32 billion just to get that. That's at least $31 billion more than such a perpetual license could possibly be worth.
It's already fairly normal in big deals to have terms to cover situations like that. IPR matters to both sides in a licensing deal, and it's not unusual for IPR to be held in escrow in case the IPR owner vanishes off the face of the earth, etc.
The fact that ARM's board are recommending this deal to their shareholders suggests that whatever the licensing terms are, IPR transfer to licensees on takeover isn't in there. That would have made ARM very unattractive for a takeover, and it would have emerged during the due-diligence that's no doubt been going on in the recent past.
Oh, and I think Apple (and Qualcomm, Microsoft, Marvell and Texas Instruments) all have a "Foundry" license, which does allow them to make changes to the ARM architecture. It's the reason why Apple can put a separate security processor deep into the heart of their chips.
Of all of these, Microsoft are the ones that I think have wasted that investment. With the ability to tinker with the architecture at a very deep level one can define a complete ecosystem (boot, peripherals, buses, etc). Now whilst Microsoft have done that for mobiles they haven't done it for servers which is where the real money is likely to come from for them, and if they don't hurry up then Qualcomm and a bunch of others will leave them standing.
p.s. @DougS, not sure why anyone's downvoting you on that, it all certainly matters a great deal to Apple.
I thought the licence those big players have ("foundry") lets them re-implement the hardware as they like, but absolutely not fork the instruction set and software interface generally (ISA, eg. cache behaviour) to make sure code portability can never be sacrificed. Intel IIRC learnt this when they acquired DEC and its Xscale designs, and tried to fiddle with the ISA, and found they couldn't because the ISA is copyrighted in some way separate from the IP licence.
WTF because this is a complete surprise. I could believe it's as simple as SB having a huge mound of cash and looking for the best possible investment, as ARM is starting to work in servers too, and the IoT bubble.
The fact that ARM's board are recommending this deal to their shareholders suggests that
it suggests that their legal obligation to do the best by their shareholders over-rides just about any other concern. This is such a large premium over the current market value that I don't think the board had any option but to recommend it, whatever their ideals may be.
Today was reporting this morning that they are promising to double the UK workforce in the next three years or something. I find that impossible to believe, and if it did happen I can foresee all sorts of organisational issues and cashflow issues that could ruin this almost lone-surviving jewel of British technology. I can see how ARM could do with a bit of a cash injection to accelerate their plans, but I can't see any reporting on how this might happen.
Bad news all around. Possibly even worse than the Cadbury's thing.
M.
P.S. slightly biased, I speak as someone who still uses an Acorn RiscPC (StrongARM) on a daily basis
"it suggests that their legal obligation to do the best by their shareholders over-rides just about any other concern."
It does not, and there is in fact no such legal obligation. The duties of the board of directors are ultimately to ensure the long-term success of the company, and while shareholder considerations are part of that, the shareholders' interests certainly do not override the prime responsibilities.
http://www.lawdonut.co.uk/law/ownership-and-management/the-board-of-directors/your-responsibilities-as-a-director-faqs#YRAAD2
Funny that most companies don't seem to work that way any more, short-term shareholder profit clearly takes priority over long-term success -- otherwise more companies would be spending more on R&D to develop new products for a few years hence, instead of cutting R&D budgets to "save money"...
You are confusing the two. The foundry license gives the licensee ARM's designs, like say that for the A72, and supporting tools etc. to help them implement it on a given foundry process. The architectural license lets the licensee design their own core from scratch to implement the ARM architecture. Apple and Qualcomm used their architectural licenses to create the A9 and Kyro, respectively. A foundry license would be used to create an SoC that includes "4x A72 cores" or other ARM designed cores.
However, even with an architectural license your design has to be compatible with ARM's specs. You can't change something in the architecture, and I don't believe you can even add your own stuff to it. If you added a new instruction, for example, the opcode you code could be used by ARM for a different new instruction, making your core no longer compatible with and able to run standard ARM code.
From what I understand, Apple had some input into the creation of the ARMv8 spec, so they wouldn't have any need to add their own non-standard stuff to it, since they would have got anything they wanted added to it already. It is more likely those trying to build ARM servers would be the ones interested in adding something, but they'd want to work with ARM so that the new stuff would be a part of a future architectural spec (i.e. ARMv8.1, 8.2, etc.)
Dunno what the plan is. The guy who runs Softbank is no fool - bought a $20 million stake in Alibaba, which is know worth $65 billion. Lucky?!
Softbank runs a mobile network and, because of how the Japanese mobile market works, is therefore a major vendor of phones. Those phones will be marginally cheaper for Softbank as they can refund themselves the license fee for the ARM core that'll be inside. But not £24 billion cheaper.
The IOT thing sounds more like it. The Japanese love their gadgets, and if IOT is ever going to really take off anywhere it will be their.
Then there's the ARM server market - if that explodes (like it's threatening to do) then ARM could become really big indeed.
will really make people sit up and take notice.
The Premium is the price you pay that is in excess of the quoted share price.
IMHO, this is just a silly amount. The board when seeing it must have had a collective heart attack.
Why would Softbank risk this large amount of money in an area of business that they have no experience?
I get the distinct impression that someone else is behind this.
Why would Softbank risk this large amount of money in an area of business that they have no experience?
They're not. This is a debt-funded acquisition, so all they're risking is somebody else's money. Due to financial repression debt is ridiculously cheap and ARM is 10% cheaper than it was a month ago.
If they're just looking at it due to the drop in GBP, they've got bad timing.
I bought at £8.55 earlier this year. It's not been above £10.75 until this morning and Independence served ARM well as most of their revenue is in dollars.
As pleased as I am to have struck lucky, I don't see ARM as being a bargain. Nor do I really think the takeover benefits any one other than shareholders or ARM and whoever gets a bonus.
I bought at 42p, and sold after 7 years, the last 5 of which it was hovering around a pound, of course it immediately took off like a rocket to the £10 mark. With the takeover I'd have stood to make over 60K in profit if I'd held on another 7 years.
I bought at £8.55 earlier this year. It's not been above £10.75 until this morning and Independence served ARM well as most of their revenue is in dollars
Friday's closing price was £11.89, and that wasn't any sudden jump.
I bought ARM for under £1 (indeed, just under 80p at best). It stayed around £1 for a year or two, and only really took off when it leaked to the financial markets that it was in the Iphone. Someone is looking for a repeat performance with IoT, after a few years in the £10 ballpark.
Why would Softbank risk this large amount of money in an area of business that they have no experience?
The Japanese economy is ex-growth. With an ageing population, and vast (and still growing) public sector debt, the wider economy has to shrink. The best thing a Japanese company can do at the moment is to spend money on businesses not exposed to Japan's domestic economy. ARM's model of being an IP house means it isn't a bet on any particular country winning the race in the Internet of Tat - ARM benefit so long as their technology is used by whoever makes the Tat, not whether that it is made in China, Taiwan, Korea or anywhere else.
With UK assets looking cheaper as FX markets react to the Brexit vote like a bunch of sheep, Softbank have taken the opportunity to buy ARM at a useful discount.
If Softbank treat ARM as an autonomous investment (much as Berkshire Hathaway treat their business units) then there's no reason to worry about their lack of sector experience.
This post has been deleted by its author
So if IoT and the ARM server market do take off, ARM could do very well out of that. It might simply be a case of Softbank seeing vast future profits ahead for ARM and fancy a piece of the action.
With ARM under Japanese control, it does give Japan an interesting strategic position. ARM could equally have been purchased by a Chinese manufacturer, and that wouldn't necessarily have been good news for Japanese chip designers and phone manufacturers. There's a lot of regional politics out in that part of the world... Access to ARM core designs arguably is very important to Japan, and there's nothing like owning the company to guarantee that. We'll get to see how important if the Chinese government starts complaining about the deal, or if Chinese manufacturers start fiddling the license fee payments.
Oh, and Fujitsu's next supercomputer might be a tiny bit cheaper if Softbank cut them a deal on the license fee...
Other companies also have to pay royalties for the use of ARM chips, IP, etc, so it is more than Softbank not having to pay for the license.
It is also good timing since Intel is not developing the Atom for mobiles any longer. So the other mobile manufacturers either find an alternative SoC or Softbank might get sued for monopoly.
In a world where money is free because they've been printing it for years then a whole load of stupid investments start to look sensible.
"The plan" is to take ARM's nice revenue stream and use it to pay virtually zero interest on the loan and pocket the rest.
Amusingly... this acquisition of the world's most successful chip company is about half financed by the sale of a "freemium" phone game... https://en.wikipedia.org/wiki/Clash_of_Clans
Now that's "added value" :-)
I'm not sure that Softbank are really a direct competitor to Intel, so I don't see exactly what the difference is at the moment.
Sure, if Intel had any designs of their own on acquiring ARM, those plans have just become a lot harder. By the time Intel's bods wake up in America, the deal will likely be done here in the UK, so they'll miss out.
Not that anyone would have let Intel buy ARM. Nor Apple, Samsung, etc. That would have been too obviously anti-competitive, and not even the blindest of Competition Commissions would have let that one go past.
"By the time Intel's bods wake up in America, the deal will likely be done here in the UK, so they'll miss out."
Not really - the shareholders have to decide in the end and could easily be open to a bigger offer. It's the board's job to get the best value for the owners.
On the other hand it would be great pity if this company was damaged in any way. I see Softbank are suggesting doubling the (UK) workforce.
"Not really - the shareholders have to decide in the end and could easily be open to a bigger offer. It's the board's job to get the best value for the owners."
I know I know. But Intel would have to react extremely quickly indeed, and make an immediate and unbelievably good offer. Big enough to encourage shareholders to run the gauntlet of a Competition Commission inquiry instead of walking away right now with an extra unexpected $10billion in their pockets.
Especially with a new PM just in post who has expressed a distaste for ownership of firms going abroad. There's no obvious anti-competitive concern with Softbank's acquisition, but there sure as hell would be if Intel tried to buy ARM. Softbank's offer looks like a dead-cert $10billion profit for shareholders. An Intel offer would be a risky-as-hell profit, and no one is likely to turn down Softbank's already very generous offer and risk losing out altogether.
Intel would have to offer something substantially above $32billion in cash, and that'd be a huge and risky proposition for Intel right now.
Who knows. Japanese companies can be very conservative in their approach to innovation, etc. ARM haven't exactly been idle on the innovation front, but they've also been very careful to stick to what they know - CPU design and licensing, and could also be considered a conservative company.
It might turn out to be a reasonably good fit, especially as it's clearly not a merger between two CPU design outfits (Softbank have nothing else like ARM in their portfolio). It might be that Softbank will simply leave ARM to get on with it.
What would be a problem is if Softbank have borrowed heavily to buy ARM, and then saddle ARM with that debt. That would be a bad thing.
The big question is, does SoftBank actually has a plan, a considered strategy in which ARM is the cornerstone that would be worth paying an extra 10bn. Or do they just have a load of cash they want to use to diversify, and someone on the board read some blurb about IoT? In that case, they might not even be aware of the fact that ARM designs chips, but doesn't make them. (Stranger things have happened.) Anyway, anything other than option one does not bode well for ARM.
Well I just read a story about this buyout calling ARM a "supplier to Apple", so obviously some of the press are under the illusion that they make & sell chips, but I'm pretty sure you don't get to the point where you can throw $32 billion around without having a pretty good idea of what the company you are buying actually does.
It is obviously an IoT play, based on the assumption that in 10 years there will be tens of billions of ARM CPUs made every year to go in toasters and light bulbs, each one licensed from ARM.
Well I just read a story about this buyout calling ARM a "supplier to Apple",
And on Today this morning that was practically all they said in the news bulletins (the business bloke was slightly better). There was no mention of the absolutely vast range of stuff that has at its heart one or more ARM cores, just "iPhone and iPad".
Shoddy reporting.
M.
Well that's because Meeja are boneheads who want to bling around Apple gear (and not for good reasons such as UNIX-y programming) and think that is all that anyone ever buys, and have *no idea* what goes on in the real world - which is why they similarly also think all motorcyclists drive Harley-Davidsons, or want to when they "grow up" or "get successful" - God knows what they think people are riding the TT or places like Laguna Seca on.
ARM is still my biggest single shareholding. My inclination as shareholder is to vote against any loss of independence. It'll take more than a just a five-figure boost to my pot to change my vote: it'll need a convincing story about being a Good Thing for the industry (can't see that), or at worst a real fabulous "retire rich" premium (which I doubt even Apple could offer).
So, what will the big institutional shareholders do? Who's pulling their strings?
With an offer of 50% premium on the closing share price? I think the Institutionals will agree with the deal. It's a lot of money right now, vs a slow-ish but steady income for the long run.
Plus depending on how Softbank do it, they might simply acquire enough shares to completely control the company (90%?), and any hold-outs will then find themselves being issued a compulsory purchase order, possibly at far less advantageous rates.
Softbank itself only has a market capitalisation of around £21bn ... it bought back around 14.2% of its own shares at around £3bn earlier this year, but it's capitalisation at around 21bn ...
And they plan to buy Arm for 24bn? Someone somewhere is ponying up a lot of cash for this one. Well, to be fair, they won't have to pony up the full amount, as Arm has around 8bn in cash that can be raided, so they just need to find 16bn ...
Well spotted.
If you have a big cash pile (And aren't called Apple) then you are a target for the commercial raiders.
All you need to do is get a short term loan to cover the amount of cash your target has in the bank and fund the rest and away you go.
We have seen this many times over the years. Here ARM's own cash pile is being used as a weapon to take them over.
I wonder what the MBA wizzards/idiots think about that?
ARM is pure IP, not a bunch of factories, so it's 100% about the skill and enthusiasm of its employees.
So, expect a few spin-offs (outside SoftBank's control) when employees realise they're viewed as the livestock that came with the farm (now run from the other side of the planet) and the atmosphere changes.
I've read that they intend to double the size of the op to 6000 developers. I think that will have a negative effect on things - dragging people away from the core (sorry) business to do customisations (presumably) will interfere with the heart of the business. I'm going to walk away with a fairly healthy wedge from this but it doesnt smell right...
It does make you wonder what all the new employees will be doing. If they're supposed to be engineers and developers in Cambridge I suspect that won't happen now that the UK has decided it doesn't want any migrant workers from Europe.
I guess this also means that Softbank is walking away from the US. Can we expect Sprint to be sold off, presumably to private equity, some time soon?
>> I suspect that won't happen now that the UK has decided it doesn't want any migrant workers from Europe.
I don't get why people keep lamenting that visa-free employment is a necessity to be able to hire.
They'd just have to get visas now. It works everywhere else in the world.
It used to be in order of priority - local labour market, EU labour market, external labour market before a visa is granted.. That's would now become the local labour market, and then the external labour market including the EU.
Free movement for employment across national borders is unique to the EU.
It's fine to have it, but it's equally fine not to.
Yes you just have to get a visa, apply a few years in advance, and hope that Tata didn't take the full quota for their call-center "specialists", and hope that your partner can also get a visa, and hope that the visa gets renewed next year or doesn't get cancelled because the rules change because some PM is facing UKIP and wants to look tough.
"Is it a good time to join ARM? Phone interview this week after applying 6 weeks ago - just wondering..."
Dunno; but a quick back of envelope show ~3000 employees in what is basically a design company with virtually zero physical assets; valued at £24bn rates each employee at about ~£6,000,000 each on average. That's football team prices.
I know it's a dumb calculation but the only other example I can think of is, HP and Autonomy, totally different circumstances but that didn't go too well.
Most of the "fanboys" out there seem to be the engineers and designers choosing ARM cores in their massive billion dollar fabrication runs. Maybe they choose ARM because of blubbering starry-eyed affection.
Beeb was Acorn which gave birth to ARM, but the BBC Micro had MOS 6502 in it FFS.
Well, they did try the brand the A3000 (or was it the A3010?) with BBC branding.
But I think the fanboy behaviour towards ARM might be less to do with the Beeb than the fact that it is just a good efficient processor (family) with a nice instruction set...
Well, they did try the brand the A3000 (or was it the A3010?) with BBC branding.
Yes, but by then Intel had transformed the computer industry into one that suited its mode of production: a demand for x86 compatibility: excellent process engineering; huge volume assembling in Asia.
It was a risk at the time but with hindsight, Acorn's decision to go fabless was the way to go.
"Most of the "fanboys" out there seem to be the engineers and designers choosing ARM cores in their massive billion dollar fabrication runs."
Number of ARM fanboys on the reg.. minimum 10s .. possibly a few hundred .. lets shoot for the low-to-middle ground and say 50. So there are 50 billion dollars of fabrication runs being done by people on el reg? Someone is living in dreamland(TM).
/me responsible for about 100k ARM cores being shipped in products. I think the number of people that can say they have been responsible for more than 100 of anything being shipped here on el reg is single or low double digits.
I think the number of people that can say they have been responsible for more than 100 of anything being shipped here on el reg is single or low double digits.
::waves::
I used to do a bit at the end of my interviews to see if the candidates had any idea about how the numbers stacked up in volume production - I gave them a hypothetical product[1] selling 100K units per month, and told them they needed to save 50p[2] off each - what would they address? The answer, of course, is "everything". But most of the candidates just stared at me without an answer :-(
Vic.
[1] Which wasn't at all hypothetical - itr's what we were working on at the time
[2] We actually needed to save rather more than that - but I thought that would be unfair for an interview question.
"The answer, of course, is "everything"."
Ah, but is that a viable answer?
Or, to put it another way: if you were project manager and you asked for ideas of how to save 50p on a product, would you accept "everything" as a place to start? Isn't that sort of implied by the question, and the answer you would be looking for being something more specific? I don't know the product in question (hypothetical or not), so I can't make any half-assed suggestions of my own - however looking at the sort of costs that the likes of the Pi family amount to, 50p could represent a hell of a lot of changes - that's an eighth of the price of a Pi Zero, for example. Or it could be something as simple as removing FlashROM and whacking a generic MicroSD card in its place (don't laugh - my PocketBook e-reader does exactly this!).
Ah, but is that a viable answer?
It is - but all I really wanted was something vaguely intelligent. Far too many of the candidates didn't come up with anything...
Or, to put it another way: if you were project manager and you asked for ideas of how to save 50p on a product, would you accept "everything" as a place to start?
I would insist on it.
For the product in question, we ended up making the last bit of our BOM reduction by reducing the number of colours on hte cardboard box it shipped in.
But all that's by the by - the point I was labouring is that in volume production, costs mount quickly. So saving 50p per unit over 100K units/month means £50K/month more dosh - so spending £250K on a total ground-up redesign is entirely warranted so long as the product is going to last 6 months or more...
Vic.
"so spending £250K on a total ground-up redesign is entirely warranted "
At a company I worked for, another R&D department in a different country was asked to cost reduce a volume product. It made good margins but it was felt the margins might come under pressure.
A totally new gee-whiz team was created, including an MBA, to redesign the product from scratch. They came up with a brilliant concept involving a clever bit of metallurgy using an exotic material, that seemed to offer a 50% cost reduction. They spent rather more than £250000 trying to make it work. Eventually it was decided to brainstorm with all the engineers.
Whereupon the guy who had designed the product in the first place pulled out a file and said "Oh, do you want it cost-reduced? I've been putting ideas together for years, but nobody was interested."
Yes; the original designer had never been consulted by the gee whiz team, and in fact hadn't been made aware of the project. Interestingly, the MBA wanted him sacked for "insubordination", but the manufacturing VP, as the saying was in those days, awarded the MBA a DCM (Don't come Monday.)
Development versus redesign. The trick is predicting when it's time to do one rather than the other, before so much money has been spent that there isn't enough left to do either.
Japanese? Masayoshi Son (SoftBank-san) may be a Japanese citizen (even born there), but everyone there knows he's an ethnic Korean. Who are one of the biggest ARM chip makers? Samsung - Korean.
Sounds more like a Korean takeover to me.
Either way, it's a pity to see it ceasing to be British.
"Hermann Hauser, the venture capitalist who helped spin ARM out of UK computer maker Acorn in 1990"
Erm, is this the same Hermann Hauser that co-founded Acorn in 1978 with Chris Curry? That was vice-president for research at Olivetti? That acts as the head of the East of England Stem Cell Network (EESCN)?
Simply referring to him as a venture capitalist seems a tad uncharitable.
Brexit was mostly about sovereignty. Arm is a crown jewel of the empire.
Is there not anyone in the UK that can simply tell these folks to piss off?
I am a yank...
Budweiser spent its life talking about "we are THE American beer"
When the Belgian's came, they were ready to crawl in bed with them.
If I were an American politician I would tell them (see above)
You have lost Rolls Royce, Jaguar, Land Rover, Mini, ad nausea-ism...
Take a stand. This is an English company, and so should it stay.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
