Given that NVidia can already license designs from ARM and given that companies will trust ARM less in rival NVidia's hands then in a neutral party it looks like the value of ARM is going to go way down after the sale, effectively lighting a giant pile of millions of dollars of economic value on fire. But the literature says that that's actually pretty typical of how mergers and acquisitions end up playing out. In general they're tributes to the CEO's ego that impoverish shareholders, though there are exceptions. I can't help but think that NVidia is paying $40B for a company that'll be worth $15B to them but would be worth $30B on the open market.
The actual value produced by ARM may go down, but the merger can be very effective in helping prop up NVidia's stock price.
The company is valued at $300 billion currently. The ARM deal will apparently be paid mostly in stock, making Softbank NVidia's largest shareholder. If they succeed in creating the perception among investors that NVidia controls the ARM market, the share price increase would dwarf a $15B loss in the value of the ARM unit itself.
for the arm unit itself, i agree. but what about the threat that arm poses to the rest of their business?
"acquisitions diminish value of thing being acquired" may generally be true but "acquisitions can be a way to stave off a threat" i believe may also be true, though to your larger point this may lead to less value in the market overall.
I’m curious if EU antitrust watchdogs will try to stop this.
The US will gladly accept it, since this is a massive plus for national security. This deal would mean a US entity owns pretty much every important chip architecture.
It’s already a huge strategic loss for the EU & UK that ARM doesn’t have European owners.
The power to prevent mergers is based on access to markets. EU is the worlds biggest single market, if they don't accept the merger it won't go trough. Nvidia and Amd and their customers want to sell to EU.
Big single markets like EU or US can take action against any merger between two international companies even if neither of them is based on the EU.
I don't think there is any reason for EU to take action in this case.
The US is the world's biggest single market, by a considerable margin over the EU now. It was larger than the EU before Brexit, now it's that much bigger.
US GDP for 2019 was around $21.5 trillion.
EU GDP for 2019 minus the UK is $16 trillion ($18.7 trillion with the UK).
That giant $5.5 trillion gap is big enough to fit the world's #3 economy (Japan) into. It's larger than France + Italy ($4.7t), or Germany + Spain ($5.2t).
Given the major economies of the EU haven't seen any net economic growth since 2006-2007, it's unlikely the EU will match up to the US economically in the coming decades. Over the coming 10-20 years it will persistently fall away, an increasing distant third behind the two superpowers of the US and China. And that's the good scenario and assuming there are no more exits from the EU. The US will become more aggressive about its trade imbalance with the EU, with the EU heavily dependent on exports and in a position of weakness with the US accordingly; and China will seek to compete with and replace the high value manufacturing they import from the EU (with most of the focus in both cases on Germany, their future is stagnation if they don't figure out how to move to a modern tech economy and fast). Germany will go through a very painful deindustrialization that they've largely been able to avoid thus far, as China goes after their high-end manufacturing and then exports those competing goods to the rest of the world far cheaper than Germany can.
France GDP: 2008, $2.9 trillion | 2019, $2.7 trillion
And that's of course pre-pandemic. Those are the five largest economies of the EU, accounting for 73% of the EU economy. They're going to see a likely net zero (or negative) economic growth for 20 years or more (inflation adjusted it's going to be brutal).
You're making lots of questionable assumptions here. Primarily one being that the US (and indeed the EU) continue to exist in their current form on a 20 year timeline. Given the political climate in both, that seems unlikely.
Given that the US is reindustrialising [https://doc-research.org/2018/06/reindustrialisation-us-degl...] right now, it's extremely unlikely we'll see a decline in German manufacturing - even if that means an increase in european protectionism. There's no way Germany (who hold vastly outsized power within the European Commission, especially in a post Brexit world) would willingly allow that strategic and economic advantage to wilt. Especially given the economic decline and social disintegration that the move to a service economy wrought on France and the UK. Chinese economic might is almost entirely based on exports, in an era of increasing protectionism and at a time when their manufacturing base is moving to places like Vietnam and to a lesser extent India. The US's failure to invest in infrastructure (everything from broadband to roads to dams) has left it in need of massive capital investment. And of course massive internal investment (arguably unsustainable and fraudulent) in construction. I'm not suggesting these are paper tigers, but neither are assured of 'superpower' status.
Moreover - the enormous material costs we're starting to see from climate change will accelerate. We can estimate some impacts, such as enormously larger migrant crises and water wars. But the second order effects of this are extremely unpredictable.
Extrapolating based on the last 17 years doesn't make any sense. 17 years ago you would have made a very different prediction based on the 17 years before that.
Who are ARM's actual customers? Most people don't buy anything from ARM directly, they buy from a company that builds a product with an ARM product in it. I'm sure there are companies like that in the EU, but how many compared to the US and Asia? And would that matter in terms of market access to the EU?
The point you are missing is how EU antitrust works, in the EU there is essentially a split between the Commission specifically the Directorate General for Competition and the competent authorities (aka the regulators) in each EU member state.
The DG COMP investigations are usually triggered via complaints raised by member states however they can also decide to investigate it own their own, once they done if any action needs to be taken they would authorize the regulators in each member state to take that action they do not enforce actions themselves.
However there still needs to be a violation of either Article 101 or 102 for them to investigate, this is quite different from the Chinese case where NXP and Qualcomm had to seek permission from the Antitrust regulator, in this case if NVIDIA requires one it would only be from the UK regulator and there is currently no freaking way that Brussels would interject in the midst of Brexit.
A year, 5, 10 down the line they might decide to open an antitrust case if they believe that NVIDIA's practices warrant an investigation, but currently there is no way for them to block the sale.
Nvidia/Arm merge is large enough to require the commission be notified prior to its implementation. The time between notification, clearing or final decision is just few months.
EU is going to ban ARM chips ? Or impose sanctions/tariffs ? What do you think those will be transformed to ? Higher mobile prices for EU citizens for no apparent reason and no reasonable alternatives. And to what end ? EU has no players in this space.
There is nothing EU can or should do here, no matter how you feel about the move.
Oh really ARM is not selling/operating in Europe? Brexit or not it doesn't change a thing. Last big merger (Qualcom and NXP) was blocked by China. Even if EU don't do it China is likely going to mess Up with this one again.
Firstly the EU doesn’t have the same requirements as the Chinese anti trust authorities, secondly it’s not about what they could do in theory but want can do in practice it’s politically impossible for the EU to block it now with Brexit going on.
I'm sure they do, however if the CMA clears the deal the DG COMP can't block it politically in general and lawfully if there isn't a clear violation of the TFEU.
The DG COMP isn't able to give the CMA direction atm, the sale doesn't violate article 101 or 102 and no member state has raised a complaint against it.
This isn't about Brexit being a distraction, it's about their mandate and optics.
Until the 31st of December the DG COMP actually needs not only to show that there is a sufficient reason to investigate the deal, but that the CMA ruling wasn't in line with the TFEU.
The US govt can't protect the chip manufacturers from malicious actors. It takes an obscene number of people with the right culture and expertise to have a company that can secure its chip platforms. Even then, the chips and associated software has to be sold to specific trusted customers.
I am totally okay witht he U.S. owning it for strategic reasons. I just wish it wasn't Nvidia...Too much centralization. Another company (IBM?) would help ensure ongoing competition.
IBM? Seriously give the geriatric case control over yet another ISA?
NVIDIA was bullied by both AMD and Intel when it wanted to make GPUs that are capable running x86 code.
IBM would pretty much kill ARM at least SoftBank left it be.
NVIDIA on the other hand will come with a strong vision and competency this is by far the most exciting shift in the semi market for years and we might finally get some competition in the CPU space.
IBM? Seriously give the geriatric case control over yet another ISA?
IBM knows how to license an ISA and has heard of open source. NVIDIA hates open source almost as much as Oracle. I'm pretty sure its not going to go well if NVIDIA gets ARM.
> we might finally get some competition in the CPU space
You are talking about Desktop / Server CPUs I presume.
The problem is that this is detrimental to the competition in the mobile / phone CPU space as Nvidia is a direct (although not very big) competitor to Qualcomm, MediaTek, Samsung, HiSilicon etc., but would now own the architecture everyone is using.
This is the second major company in past few years where SoftBank pretty much served as an intermediary for merging of companies. They bought Sprint a few years ago and merged it with T-Mobile. Now they are selling ARM, which they bought just 4 years ago, to NVDA. Wonder if this was their plan all along or if they had to change plans due to financial issues or if they faced regulatory/geopolitical issues.
Isn't this the standard private equity playbook? Buy, hold, and flip? Though Softbank isn't a VC and I haven't heard of mass cost-cutting after they buy something.
There is some hypothesis in the RISC community that this sale will be a boon for RISC-V. With NVIDIA at the helm, ARM is no longer a neutral third party, so we may see a bunch of investment into RISC as it replaces ARM's SoC prevalence.
If Softbank needs cash now (re:WeWork), can they sell a minority stake to Nvidia, in preparation for future IPO? That would provide a path to Arm regaining their role as neutral supplier of ISA & IP. Nvidia would get a preferential driver's seat for a year or two, long enough to cut favorable deals. If Softbank could get a valuation for Arm larger than $40B, from sale of a partial stake now, that helps with their future return from an IPO. Why sell 100%, if they can get enough from Nvidia to solve their near-term cash needs? They have IoT assets, a market which has huge growth ahead.
Wonder what Apple’s reaction will be. Their relationship with Nvidia has been bumpy to say the least. Any ARM licensee now has a competitor as a supplier. The smaller ones will have to swallow but the larger licensees? Do they have options on the table - other than trying to throw a wrench into nvidia’s presumed acquisition plans?
Are not they already years independent of ARM? I thought they had one of these universal licenses allowing them to design their own processors which they seem to do very successfully?
What's the play here? To make it more expensive for competitors to license ARM chip designs? So that Nvidia just gets to make more chips, or to make non-Nvidia, ARM chips less tempting over x86?
I would certainly like to know the plan as well. If we're just guessing, there's a lot of benign reasons, like nVidia expects lean into ARM and they would rather buy the ARM than pay the license. I'm surprised there's not a lot of info on the reasoning yet.
"The announcement of the deal hinged on SoftBank ending a messy dispute between Arm and the head of its China joint venture, Allen Wu, who earlier rebuffed an attempt to remove him and claimed legal control of the unit.
Several people close to SoftBank said the matter was now “resolved,” though one person close to Mr Wu said he “remains the chairman of Arm China.” A spokesperson for Mr Wu declined to comment."
Yahoo shareholders got shafted when Alibaba spun out their multi-billion dollar payments division, AliPay, "without informing other shareholders ... because the CCP made us." lol.
