> SoftBank discussed with underwriters on Wednesday whether to price Arm shares above the initial range, at $52, but settled on the lower figure in the belief shares would trade up and boost confidence in overall markets after nearly two years of pressure on valuations.
> “So many people, including SoftBank and all the underwriters, have so much riding on the overall health of the IPO market,” said one person close to the Arm listing.
Makes sense given that SoftBank and its underwriters would like to have many more IPOs, and the fear that a poor showing here could draw out the current "IPO winter" seems reasonable.
That's super interesting! 4D chess move to be sure, but also kind of crazy that companies plan strategies around affecting the entire world market based off of their behavior.
Matt Levine has a great piece on ESG actually being a market reaction to the fact most people own index funds now. Since most people are concerned about the OVERALL market now, ESG serves to optimize an externalized concern for the entire market at specific firms.
I saw a twitter commentator point out that this is higher than the $40B Nvidia wanted to buy it for. Argument was that stricter antitrust could actually lead to greater shareholder value.
Am I crazy in thinking that $60(ish) billion feels like very little when considering the value brought about by the paradigm shifts and hugely successful products (iPhone/Apple M chips/most of the SoC's on just about every phone) made possible by its chip architectures?
Makes Twitter/X seem hugely overvalued, to provide one comparison which makes me scratch my head a bit.
I guess I truly know nothing about anything.
The Apple M successful because Apple paid a lot of money to higher great chip designers and has a great relationship with the best Fab. They could likely have achieved very similar results with MIPS. Apple pays ARM a license fee but otherwise don't have much to with chips like the Apple M.
But a lot of where they make money is very attachable from the low end by RISC-V. And from there RISC-V will just move up and up. I see their future as trending down not up.
To add, PA semi was a Power ISA shop. After apple aquired them they simply ported it — modern computers are branch prediction machines and the ISA doesn't matter as long is it has the basic load/store add ect.
But why did Apple choose to do that if they could have saved money,
never paid Arm Inc a cent, and gone with RISC-V? Arm brings a lot of things to the table, and while Apple is alone in having done a CPU architecture change successfully not once, but twice, it's not something they do just for funsies, so I doubt a switch to RISC-V is imminent.
It feels like something of a missed opportunity for them (Arm), I guess. I wonder what the stock value for Arm would be, if back in the day, they had come up with a business model similar to that which Unreal engine employs today.
E.g. we'll license you the architecture for free but customization and support will cost. What we really want though, is royalties generated by any and all ecosystems created or made possible by our design, at the rate of X percent of annual sales revenue generated by said ecosystem if and when the revenue reaches more than 0.65 bajillion dollars.
They let their customers make most of the money, which is how they grew the ecosystem with all the network effects. Just as an example, Qualcomm is a 10% customer of ARM, and they made $38b in chip revenue in fiscal 2022. ARM had $2.7b. Now they want more, and are going to push people to RISC-V if they’re not careful
The value a company brings to society and the value that it captures are two different things. Arm operates in an industry where it is hard to capture value, the semiconductor industry, and its moat is strong but only upheld by them not charging a lot. The semiconductor industry is mature at this point, still having a lot of development, but also a lot of competition.
If they charge too much, the bleeding to RISC-V will accelerate: many of the customers of Arm simply don't have the profit margin to afford paying more for them. Arm has competitors in the higher priced segment (Intel/AMD) and they have RISC-V for the cheaper priced segment.
I understand why SoftBank wants to diversify their investment.
I tend to agree, but also realize most phones run on razor thin if not negative margins. Stocks are all about looking long term, and as an investor I'd have concerns about the rise of RISCV or other 'free' ISAs on that horizon.
OEMs rarely make money: they get the design in but don't get significant payments until volume gets up (if ever). And once the original version is designed, the buyer wants to increase margin and/or lower costs, which they do by pressuring their suppliers.
Furthermore, ARM isn't even an OEM; it sells licenses (patent licenses plus rights to use copyrighted designs) so Oppo is paying someone to design a chip using ARM IP and paying someone else to fab it, and then a little goes to ARM.
All in all, a shitty business unless you get lucky, which ARM did. They took advantage of their luck, don't get me wrong, but luck was what gave them the opportunity that so many others either didn't get or didn't know how to take advantage of. Microsoft was another lucky winner who knew what to do; Intel went the other way and made their own luck until they got distracted.
Unlike, say, NVidia, Intel and Apple, ARM is not in control of its fate, or rather has limited control. And despite those first three having control, that doesn't guarantee anything: one of them lost the thread, one of them had a near death experience, and one of them has so far managed to stay on top.
You should price it against the replacement value - what would it cost to replace ARM? (Or Twitter/X for that matter - in the later case you need to include the marketing needed to match their current userbase.)
The paradigm shift is utterly irrelevant to their market value.
If they aren't making money and aren't likely to be making a lot of money at any time in the future, it doesn't seem low - regardless of whatever impact you perceive. One of the problems that ARM faces is that people have alternatives or can create them. For example, you talk about Apple, but Apple is only licensing the instruction set. Instruction sets have some value, but the real value that ARM provides is in their core design. However, more players are looking to design their own cores. Apple already does and Qualcomm is looking to do the same. That would really harm ARM's revenue stream.
Did ARM bring about these paradigm shifts or were they merely a convenience? Apple started using ARM in their iPhone because it was a convenient and affordable low powered processor, but if ARM had wanted more for their IP, people wouldn't have used ARM in the first place.
It's one of the reasons that companies like to bait-and-switch if they can. Get people using your product for cheap and then hike the price once everyone is using your thing. You then argue that you've created so much value via your product - but everyone would have just used something else if you'd charged a lot from the beginning.
ARM's in a tough place where Apple has an architecture license and makes their own cores and others are looking to follow that lead. As ARM tries to generate profits, others are there to stop that from happening. I'm not as bullish on RISC-V as many on here (it just takes a long time to get to the level that high-end ARM chips are at), but it could certainly start eliminating some business and provides a hedge against ARM increasing prices.
CNBC is saying ARM's PE ratio is 170. Where's its growth going to come from to justify that? With Nvidia, we see major growth. Nvidia has a differentiated product with some lock-in around CUDA and such. ARM's largest customers are looking to pay ARM less money going forward. If ARM loses its suit against Qualcomm/Nuvia and Qualcomm stops using ARM's cores, what happens to ARM's revenue/profits?
Some businesses that have a large impact only have that large impact because they were offering an affordable product. That affordability limits their valuation. Without that affordability, the world would have gone in another direction. Right now, ARM is trying to figure out how to pump up its profits given the slight amount of lock-in it commands. However, I think it's more likely that ARM will just hasten companies moving to their own core designs and investing more in RISC-V.
Yeah, Softbank... They evaluated WeWork at ~50 billions and a few months later it was worth almost zero. If you work with SoftBank prepare to lose everything.
Readit News
> SoftBank discussed with underwriters on Wednesday whether to price Arm shares above the initial range, at $52, but settled on the lower figure in the belief shares would trade up and boost confidence in overall markets after nearly two years of pressure on valuations.
> “So many people, including SoftBank and all the underwriters, have so much riding on the overall health of the IPO market,” said one person close to the Arm listing.
Makes sense given that SoftBank and its underwriters would like to have many more IPOs, and the fear that a poor showing here could draw out the current "IPO winter" seems reasonable.
whereas the $40B would have been for 100% of them
The Apple M successful because Apple paid a lot of money to higher great chip designers and has a great relationship with the best Fab. They could likely have achieved very similar results with MIPS. Apple pays ARM a license fee but otherwise don't have much to with chips like the Apple M.
But a lot of where they make money is very attachable from the low end by RISC-V. And from there RISC-V will just move up and up. I see their future as trending down not up.
E.g. we'll license you the architecture for free but customization and support will cost. What we really want though, is royalties generated by any and all ecosystems created or made possible by our design, at the rate of X percent of annual sales revenue generated by said ecosystem if and when the revenue reaches more than 0.65 bajillion dollars.
If they charge too much, the bleeding to RISC-V will accelerate: many of the customers of Arm simply don't have the profit margin to afford paying more for them. Arm has competitors in the higher priced segment (Intel/AMD) and they have RISC-V for the cheaper priced segment.
I understand why SoftBank wants to diversify their investment.
[1] https://macdailynews.com/2023/08/04/apple-iphone-dominates-w...
Furthermore, ARM isn't even an OEM; it sells licenses (patent licenses plus rights to use copyrighted designs) so Oppo is paying someone to design a chip using ARM IP and paying someone else to fab it, and then a little goes to ARM.
All in all, a shitty business unless you get lucky, which ARM did. They took advantage of their luck, don't get me wrong, but luck was what gave them the opportunity that so many others either didn't get or didn't know how to take advantage of. Microsoft was another lucky winner who knew what to do; Intel went the other way and made their own luck until they got distracted.
Unlike, say, NVidia, Intel and Apple, ARM is not in control of its fate, or rather has limited control. And despite those first three having control, that doesn't guarantee anything: one of them lost the thread, one of them had a near death experience, and one of them has so far managed to stay on top.
The paradigm shift is utterly irrelevant to their market value.
If they aren't making money and aren't likely to be making a lot of money at any time in the future, it doesn't seem low - regardless of whatever impact you perceive. One of the problems that ARM faces is that people have alternatives or can create them. For example, you talk about Apple, but Apple is only licensing the instruction set. Instruction sets have some value, but the real value that ARM provides is in their core design. However, more players are looking to design their own cores. Apple already does and Qualcomm is looking to do the same. That would really harm ARM's revenue stream.
Did ARM bring about these paradigm shifts or were they merely a convenience? Apple started using ARM in their iPhone because it was a convenient and affordable low powered processor, but if ARM had wanted more for their IP, people wouldn't have used ARM in the first place.
It's one of the reasons that companies like to bait-and-switch if they can. Get people using your product for cheap and then hike the price once everyone is using your thing. You then argue that you've created so much value via your product - but everyone would have just used something else if you'd charged a lot from the beginning.
ARM's in a tough place where Apple has an architecture license and makes their own cores and others are looking to follow that lead. As ARM tries to generate profits, others are there to stop that from happening. I'm not as bullish on RISC-V as many on here (it just takes a long time to get to the level that high-end ARM chips are at), but it could certainly start eliminating some business and provides a hedge against ARM increasing prices.
CNBC is saying ARM's PE ratio is 170. Where's its growth going to come from to justify that? With Nvidia, we see major growth. Nvidia has a differentiated product with some lock-in around CUDA and such. ARM's largest customers are looking to pay ARM less money going forward. If ARM loses its suit against Qualcomm/Nuvia and Qualcomm stops using ARM's cores, what happens to ARM's revenue/profits?
Some businesses that have a large impact only have that large impact because they were offering an affordable product. That affordability limits their valuation. Without that affordability, the world would have gone in another direction. Right now, ARM is trying to figure out how to pump up its profits given the slight amount of lock-in it commands. However, I think it's more likely that ARM will just hasten companies moving to their own core designs and investing more in RISC-V.
Fortunately the valuation of this one is being corrected
In this case Softbank and LPs win it all back and dumps their shares on retail.