Blain’s Morning Porridge Aug 30 2023: Will ARM break the Tech Market?
“ Best thing for a government to do is to eliminate obstacles like filing registrations, getting licences, and so on… tedious steps slow everything down.”
September’s ARM IPO could be a “CFP” for markets if it breaks the current everything AI consensus (aka “bubble”). There are many reasons to think it could, because it’s basically Masayoshi Son rolling it with big aplomb in the Tech Casino.
September’s market is going to be a fascinating month. I will be watching ARM with great interest….
Most financiers and fund managers are cynical enough to be aware an IPO of a Private Equity held firm may be cover for cashing out on the debt-raddled husk of a company that PE partners have leveraged to the hilt to pay them dividends. Imagine Baron Harkonnen of Dune screaming: “Bring me a new firm, this one’s burst…”, as the PE partners might say.
Occasionally, IPO’s are even more cynical… then the trick becomes to make something dull, boring, predictable and solid look like the most exciting investment opportunity of all time… It’s called Hype. When engaged in hype, sponsors don’t like answering questions, which is why this morning’s quote is Masayoshi Son, boss of SoftBank.
Let’s start with a tale…
Imagine I own the patent on a widget. I licence my widget to the market, where the widget is integral to a device everyone on the planet owns. It sounds like a good business. But the thing is…. these devices sell for $500 each, but I only get paid 11 cents on each licence. And since everyone already owns one, I get annuity-like income over a 4-year cycle as folk buy a newer device. My net income, say $500m, is good but has flatlined in line with what is now a saturated market. I am selling into a mature market with little competitive threat to my widget because it’s so cheap.
How do I create greater value in myself?
How about I sell 25% of myself to myself at a price of, say, a randomly chosen number of $16 bln… therefore establishing the value of my widget licencing at $64 bln? Clear? I then launch an IPO at said $64 bln valuation, IPO-ing (selling) 10% of my widget licencing firm while retaining full control. I then have $6.4 bln cash to invest (actually to pay off my losses at the Casino where I’ve been betting unwisely on self-driving cars, office space and other things I thought were valuable but weren’t), and that valuation allows me to borrow more using my remaining shares as collateral to borrow from banks who want my fees.. giving me liquidity to go back to the Casino with.
It is brilliant. My logic is unfaultable..
I am, perhaps a little greedy on my valuation metric of 120 times earnings, but I can probably justify it in this widget obsessed market by claiming my widgets will prove equally integral to all the new, new widget based things coming to market! Even though these new sectors are much more competitive spaces dominated by other widget makers, designers and licensers, all I have to do is hype up the expectations in my Form F-1 filing, and tell the market I will dominate these coming sectors like I dominate mature widgets..
You get the drift…..
I remember Acorn computers. In 1990 Acorn went on to found Advanced RISC Machines in a JV with Apple and VLSI. Its trick was reduced instruction set computing, leading to more efficient, lower power chips, which is why ARM now dominates CPUs in smartphones.
I am reliably informed – by my tech chaps who understand the nitty gritty of tech (I just look at the numbers and provide lashings of cynicism) – that ARM is very fine company. It makes around $2.6 bln revenue and $500mm earnings by licencing its’ Intellectual Property and Chip Blueprints to the smartphone makers like Apple and Samsung where low prices mean it faces little threat, and to chip designers and makers, including Nvidia.
Over 30 bln chips per annum are based on ARM IP. It was taken private by SoftBank in 2016 at a $32 bln valuation, and since then it’s not really done very much except remain a very fine company licencing designs to smartphone makers, but barely denting the high value AI data-centre and Automotive chip sectors – but with a reputational niche in the overhyped and now generally forgotten about IoT niche. (If you had to look up IoT it tells you something about how fickle fashion is in Tech investment.)
Funnily enough… Softbank just bought 25% of ARM from Masayoshi Son’s Vision Fund at $16bln, justifying the firm’s valuation at $64bln – just like my widget maket. How coincidental… Son has determined ARM is worth $64 bn because he bought 25% from himself at $16bln. He’s as much a genius as I am.
To be worth the $64 bln Son is hyping expectations on future AI sales in September’s IPO of ARM – potentially the biggest IPO this year, meaning it’s potentially a classic Critical Failure Point for the Tech Market. If it fails…. Then how much heat comes out the market? It’s a CFP for Son as well – if he can’t prove he’s doubled the market value of ARM over the last 7 years, that severely dents his credibility as the tech super-investor.
In fact, the whole ARM IPO is balanced upon a number of Single Points of Failure – SPF’s that could each sink the deal.
The first SPF is Son’s ability to convine investors ARM is worth $64bln. He has to show how it’s going to be able to licence its chip design sets to an increasing number of firms wanting to be in the AI sector with their own chips (Amazon and Apple are expected to anchor investors) and persuading them to pay up for the IP. Much is being made of the need for Chip Makers and Users to be close to ARM and to support the IPO. That means persuading them a P/E of 120 for a mature business is fair – and it faces theats to its RISC moat from new software.
At the moment all the indications are ARMs value is in smartphones, if it can’t convince investors it’s also an AI player – where growth is expected, that’s a second SPF. ARM numbers have flatlined this decade in phones, but it provided the systems architecture that powers Nvidia’s critical Grace AI chips – Nvidia is making all the money on them already!
A third SPF is pricing its IP. In terms of microprocessor revenues the problem is: it’s not chip designs that are in scarce supply – its Chips. Who controls the supply of Chips stands to win – which is why everybody loves NVIDIA, which curiously is “Fabless”, but farms out its designs and production to TSM! ARM has no such production partnership or route.
(As an aside, when I was young Silicon Glen in Scotland was touted as the UK’s hope for dominating chip production, but years of mis-planning, mis-investment and selling out early put paid to that. One of my smarter chums from that era was recently in Taiwan and visited a leading foundry. He asked about how to set up such tech. He was brutally told the UK was “too poor” and “too inexperienced” to make the 30 year investment that would be required to establish modern foundries.)
The number of SPFs increases. The IPO documentation and filings reveal ARM China is ARMs biggest customer at 25% of revenues, but ARM only owns 48% of the company, giving it, de-facto, zero control. The Chinese seem to do what they want, appoint who they want, and have delayed royalty and licence payments to the parent firm. And since the current market is really about Tech, and China vs USA is all about not sharing Tech, ARM faces another SPF in relation to its biggest market and customer in a teck lockdown with the US where it wants to list!
The big risk is that the market spot’s Son’s game here – its bought the stock from Vision Fund to justify the $64 bln valuation, persuaded its 24 bank underwriting group (in line for big fees) and big firms like Apple, Google, Nvidia etc (all wanting ARM’s IP) to anchor the IPO, and all to support SoftBank being able to use its ARM stake to borrow more.. or something like that.
Does the ARM IPO begin to sound slightly dubious? It does to me..
Five Things To Read This Morning
Out of time and back to the day job
Strategist – Shard Capital