Their stock is falling because of future demand issues coming up with the whole semiconductor industry. Just this week, expanded export bans to China were announced and are an uncertainty on how much volume will be impacted in the future, because no one knows what will get an exemption or not.
The U.S. position on the entire Chinese semiconductor industry is getting increasingly ridiculous. It's not just chips, since China is the one country that's developed efficient monocrystalline silicon ingot production at scale, and solar panels made with that material are far superior to anything made in the USA currently.
The USA's financialized short-term thinking has resulted in a failure to develop basic technological infrastructure capability, and China is now in the lead - so the response is, stop trading with China on all things silicon? This is the Tonya Harding team's 'let's try to win by kneecapping our superior competition' mentality, and it's doomed to failure, if not in the short term, than certainly in the long term.
It's a move part of the bigger world picture. A large chunk of experts see China as the likely next world superpower. The USA, current world superpower, probably can't prevent that happening, but they can make moves to delay it.
This move might delay it happening by a few months. Combine with lots of other small moves, and the USA might get another few years of world superpower status, even if the moves hurt both countries in the end.
> A large chunk of experts see China as the likely next world superpower. The USA, current world superpower, probably can't prevent that happening, but they can make moves to delay it.
We'll see.
Lots of people thought Japan would overtake the US as well, and that didn't exactly pan out.
China has some real problems between the legacy of their one child policy and their current real estate investment crisis. It's certainly possible that they overcome those issues, but it's far from guaranteed.
China overtaking the US seems a lot more likely based on sheer numbers. China’s population is over 4x larger than that of the US. So China “just” needs to reach a GDP per capita of 25% to have a larger economy than the US. This metric has been growing rapidly in the past decade or two and from a rough search, seems to currently stand around 17-18% of the US’s figure.
Of course GDP isn’t the only factor in world power status, but it sure is a big one.
Compared to the US, China also has a far larger labor force, a greater ability to manufacture things, and greater government control over its population.
Access to natural resources is similar for many things, less for some (like oil and gas) but greater for others (like rare earth minerals).
> If worldwide carbon caps and tariffs become a thing, the oil and gas in the ground suddenly becomes nearly worthless if you can't use it.
It may be worth less, but it won't be worthless. I have a hard time seeing another fuel source for long distance aviation (although we may move to synthesized fuel in the future).
Companies will also need feedstock for plastics - that use won't be affected by carbon caps at all.
Those tariffs are also a joke in and of itself. As we can already from the oil price cap idea you can only enforce things when the world plays along, but looks increasingly like the global south is fed up by these dictates. So good luck enforcing carbon caps and tariffs when the majority of the world population doesn't play along.
You can still have them, but ultimately it's the people in western societies that will pay the price with their own pensions and social services gradually degrading.
We kinda enforce "don't trade slaves" globally... And "No child labour"... And "don't steal other countries ships when they sail near your country".
"Don't emit more than your fair share of carbon" is just another rule like the others. If most countries agree, but there are a few dissenters, we can deal with it like other international disagreements - warnings, then sanctions, then war.
Currently, we're far from that, because we don't have a majority of countries agreeing on carbon caps.
Nah - it'll go down like Venezuela [1]. The government of a country flagrantly ignores the internationally agreed carbon caps, and so some country or group of countries get together and sponsor a coup.
It happened to Venezuela because they refused to align with US interests, and with other economic and political issues, they were an easy target. The wikipedia page [1] has a nice summary of the list of countries on each side of the dispute, and you can see a nice 'pro USA' and 'anti USA' divide.
I think China could easily fix their demographics if they wanted to.
They have sufficient control of their population that if they said "everyone of child bearing age must try for a child this year and next", they'd probably double their population within 2 years.
The 'demographic problem' is only a problem because they don't see it as an issue worth fixing.
Lots of them don’t want to have kids. The government has many ways to influence the public. But I don’t think it’s true to say that they have so much control over their populace that they could do anything like double their population in 20 years, let alone 2.
People in China are just people, right? Many of them don’t want to have kids. What’s the government gonna do?
Of course, hypothetically, the government blocks people from leaving the country and literally forces them to procreate but that’s getting a bit imaginative (it would be turning the China that exists today into something very different).
Realistically I think the government is worried about their demographic problem and is using every reasonable tool in their arsenal to attempt to address it.
> See the Cultural Revolution and Great Famine etc for examples of what it can do. Basically anything.
Even assuming China is all powerful and can achieve "anything" here, the policies would still have a latency of an entire generation. Unless you include time-travel as part of "anything".
There's nothing that needs solved. It's ideal to shrink the global population while boosting per capita quality of life. The focus should be on improving per capita, not on expanding the overall economic size or population numbers.
Providing housing to young families will not make a meaningful dent to declining demographics in the affluent world. People do not want to have so many children (3-5 or more) and it's not due to lack of housing. That's a cultural change and no amount of housing is going to reverse it. And even in the financial scenario, the housing aspect is only one chunk of the huge cost of raising several children.
> There's nothing that needs solved. It's ideal to shrink the global population while boosting per capita quality of life. The focus should be on improving per capita, not on expanding the overall economic size or population numbers.
Then why do we simultaneously import millions of workers each year to make up for the supposed shortage of natural population growth (caused by policy)?
But many of those coming here are not well educated. Hard working maybe, adventurous certainly, but often economically deprived and undereducated as a result, or so I’ve read.
There is a real illegal vs legal split here, legal immigrants tend to be quite well educated. The majority of illegal immigrants have a high school degree but they are definitely less educated than the average American.
That said, while undocumented immigrants tend to be less educated being able to work when we didn’t pay for any of their education is still a boon. It’s something of a waste to pay for 12 years of education and then have someone doing manual labor.
If you immigrated to the west as a working proffesional, then you have to have job offer, education and usually you have to be somewhat well off to be able to afford lawyers and get through the visa process.
I don't think you can fix population issues in 2 years. Takes ~20 to make an adult (unless you have a liberal, multi-cultural society -- American adult citizens could be made in an afternoon if we shifted some laws around).
PRC demographics are extremely advantageous for great power competition / building comprehensive national power going forward. The TLDR is PRC leadership has been too preoccupied with sustaining 1.4B people, huge% of which are old and undereducated (to be blunt: a drain) that drags on her potential. Shedding quantity while improving quality will significantly improve PRC competitiveness.
Family planning / One Child Policy = concentrating resources on only kids = PRC now produces comparable STEM talent as all OECD countries combined. Demographic bomb collapists narrative fixate on absolute population decline without realizing PRC is starting to undergo the largest (by an order of magnitude) high-skill demographic divident in "recorded history" that will last decades. Meanwhile reducing net population = declining import dependency over time = more strategic (military) flexbility. Combine with massive home ownership % and high savings rate, and most of population likely settle around middle / low high income = no expectation for onerous social safety nets.
Over the next 10-30 years, we will see PRC (even with shit tier TFR) close/lead in technology due to having the largest and likely best talent pool (courtesy of PRC scale), that exceeds what US can produce domestically + immigration. Which will also erode western dominance/economy for high value niches it tried to fence off for itself. Also a PRC with increasingly less import vunerability due to lower absolute population, which mitigates much of the geographic shortcomings, i.e. first island chain "containment" nations that will be forever import dependant will lose their geographic leverage. Meanwhile increasingly burdensome social nets will drag down / destablize developed economies vs PRC who can sustain domestic serenity on relatively lean welfare estate.
Yes, generally, PRC nationals (Tier 2/3/4 cities) will (continue) to have worse QOL relative to developed societies. But in terms of overtaking US, and improving existing strategic enviroment, demographic trends are pointing up, up, up.
The US could solve the population difference fairly easily by having some family friendly policies, and walking back laws that are destructive to families such as by incentivizing divorce etc.
I was of the view that China could be the next super power and overtake the US.
But two things changed my view:-
1. COVID and their response to it. China still does not have a good vaccine and they still rely on lockdowns. That speaks a lot to how backward China is in basic sciences.
2. The move to remove term limits for the president. History tells us that one person can’t be relied on to make good long term decisions for a country specially a communist country.
China's covid response was extremely damaging to China, but thanks to the US and their allies response to Russia, China has received plenty of intermediary free money to add to their economy. Cheaper energy, and arbitration returns from Russian goods to western economies for a premium. Ultimately China will find a way to cope without the products from the west. Europe however cannot cope without the cheap goods from China and the cheap energy from Russia.
India is by the way the only major economy that is projected to grow over 6% next year. That's largely due to profiting off of resell of Russian goods.
The problem is that people conflate their own moral standing with actual factual realities and unfortunately it will cost _us_ dearly.
India isn't growing (projected, un huh) at 6% due to profiting from Russian exports, that's a joke.
Russia's economy is trivial in size, as are their non-energy exports.
A $20-$30 gap in the price of oil via artificially cheap Russian oil isn't the difference between India's large economy (3x larger than Russia, nearly the size of Germany) growing at -3% vs growing at +6%.
We need to put the hammer down far harder on Russia than we have so far.
> COVID and their response to it. China still does not have a good vaccine and they still rely on lockdowns. That speaks a lot to how backward China is in basic sciences.
I think this was a policy move more than a lack of science understanding. In fact, leadership in China is very science-heavy. Time will tell, but it's actually looking like possibly a good move, now that 5% of the UK population is not working due to long term sickness, the highest ever[1], and covid is looking like a contributory factor.
I could even imagine China extending their lockdown policy to other diseases which might have long term population-wide effects, like influenza.
The main downside of such a policy is you need to have effectively closed borders forever, otherwise travellers re-introduce diseases from other countries into yours.
>Lots of people thought Japan would overtake the US as well, and that didn't exactly pan out.
not even comparable situations, Japan has a fraction of the population, natural resources, isn't a nuclear power, and is effectively militarily occupied by the US to this day. People using Japan as a comparison to China are either naive or just coping with reality to comfort themselves
This is one of those opinions that sounds smart but doesn’t make any sense to anyone with more than a passing familiarity with it.
A large chunk of experts do not see China as the likely world next superpower, many experts are debating if China will survive as a unified nation in the next decades.
China has finally admitted to lying about their population for decades, they have some of the worst demographics in the world, along with the single worst food security and greatest debt load in recorded history.
No serious person outside of anonymous online forums would ever describe any of these things as ingredients to success, and this is k my a small sampling of their long term problems.
This is even ignoring India, a country with much better strategic advantages and an actually growing population.
The idea that China was next was popular by experts in 2005.
> The USA's financialized short-term thinking has resulted in a failure to develop basic technological infrastructure capability, and China is now in the lead - so the response is, stop trading with China on all things silicon?
"Force people to try to come up with alternatives rather than rely on that single external source that short-term thinking created" seems a lot smarter than you give it credit for.
This is one advantage of tariffs. Increase the costs. For some applications there are easy substitutes, for others it is harder which incentivizes the creation of alternatives.
Of course the downside of tariffs are also well-known. Retaliation tariffs that damage your own industry. Increased costs being passed own downstream in the economy resulting in less value creation. The overall deadweight loss of not utilizing comparative advantage.
It still might be worth it, but you don't get that "force" for free.
Hmm, and wouldn't that work both ways? What's stopping this from being an impetus for China to stop being dependent on US-allied countries for semiconductor fabrication?
I imagine the Raytheon/Lockheed/etc. crowd does want increased military tensions with China to justify bloated MIC budgets into the foreseeable future, and of course that's one of the few manufacturing centers in the US that hasn't been heavily outsourced to sweatshop labor zones.
It's not very productive in the long run, however - for example, a tank is only 'economically productive' if your economic model is based largely on raiding other country's resources. A tractor, or construction crane, however gives added economic value because it can be used to improve agricultural production, build better infrastructure to facilitate trade, etc.
War is a general consequence of monetary system problems, so just blaming it on the military wanting to be bigger (which is true for every organization) doesn't look totally at the big picture.
I love Shelling out: The origins of money blog post from Nick Szabo as it's a great explanation of how having a hard monetary system with small inflation shared by everybody the parties to be more peaceful.
US had a trade relation with China where China was buying US bonds, but as US was issuing lots of new bonds China lost trust, and started selling it thereby opting out of the US bonds based monetary system.
China has reduced some of its holdings of treasuries (<10%), but still have a trillion dollars worth. They are diversifying, but they still hold a ton of USD.
China could have very easily dumped all of their holdings by now. They're choosing to hold that pile. It's no longer a meaningful pile of treasuries vs the overall size of the US Government debt load.
Ignorant people used to joke that China owned the US because of their treasury position (despite the fact that Japan's holdings were roughly as large). Well that position has been massively diluted over the past decade and is now largely meaningless. The Fed can eat China's position with a snap of its finger.
It's to China's benefit to hold the position, at present.
I think the theory would be more like, dictator mismanages economy, uses saber rattling as a distraction, and then things spiral out of control from there. I think this is still wrong in the sense that pointing toward any one thing as the only cause of big geopolitical stuff is wrong, but at least it is a coherent theory.
This interest in bringing manufacturing back to the US doesn't really seem compatible with the Fed's current policy of making the dollar really, really strong.
Not if the fossil fuel sector has so much power over the US government that they can squelch renewable competition via control of funding for R&D and spending projects. Anyone who looks into actual DOE R&D budgets since Reagan will see this playing out in practice.
China notably has no domestic gas/oil to speak of, and coal has so many obvious drawbacks that they've decided to push money into renewable R&D as part of their long-term strategy.
You speak as if the security and strategic concerns of the state are shared by the businesses that exploit the state while simultaneously influencing policy. It seems naive to think that the state is always working in it's own interest when all of the money that flows into our political system is designed to keep the state from functioning efficiently or at all.
I mean these latest set of restrictions are all bad for US companies. This was an example of US strategic/security concerns just completely running over corporate interests.
Respectfully, there is so much that is just factually incorrect in your comment that I don’t want to bother to even respond. Please just save us the hassle and delete, and in the future refrain from speaking confidently about topics of which you have zero clue.
Also the long-term uncertainty of Taiwan's political situation will continue to weigh on TSMC's stock price.
For example, investors may be afraid that there might be some ominous signals coming at this week's National Congress of the Chinese Communist Party, held every five years.
>Also the long-term uncertainty of Taiwan's political situation
Republic of China move to Taiwan since 1945. that's over 70+ years. that's one person born and into retired age.
you can question every country on Earth for its long term uncertainty. North Korea constantly shoot off Nukes. Russia is invading Ukraine now. Saudi and Yemen.
American just recently left Afghanistan to Taliban after 20+ yrs of Nation building.
70+ yrs peaceful times in Taiwan with some up and down with China seem pretty stable to me.
edit: even USA have some long term uncertainty after Jan 6 (United States Capitol attack or protest?) Trump-backed election deniers is about to be elected into the US Congress.
This is a simple but important point. It's easy to overindex on the idea that a declining stock price means something is wrong with the specific company. But you look at stocks across the board (there is literally an ETF for pretty much every investable country now) and you can see that most stocks across the board are declining at the moment.
I wonder how capex is factored into this. If at some point they dropped a few tens of $B on the fab upgrade, and likely need to do so again soon, when does that get subtracted from profits?
They weren't implying that, see the quote they included. OP said they hadn't seen profit margins this high, and the reply indicates that it's not uncommon to see software even higher.
It's just a useful tidbit, related to the point they specifically quoted.
Though perhaps you could say they sell derivatives of software products?
I'm assuming they write the best custom chip design software and that's their biggest competitive advantage (and then operational excellence and savvy business maneuvering et cetera). Though I know nothing about them just guessing as an outsider.
TSMC doesn't really write chip design software. Maybe they do at some small level but that's not their bread and butter. TSMC provides IP engineers need to design chips to be built by TSMC. That IP is used by CAD tools from major players like Cadence, Synopsys and Mentor (now Siemens, I guess).
TSMC's advantage is their superior chip fabrication technology and ability to scale.
You will also need to look at R&D spend and CapEx. Those profits are funnelled back to that. Basically it is a whole cycle, in order for fat profit margin you need to be at the leading edge in Tech, Yield and Volume. And all three of those requires R&D and CapEx. So you need Fat Profits margin to sustain it.
Meaning what? The factory gets destroyed? A competitor appears? Labor costs rise? Demand falls?
So long as the world economy is dependent on the TSMC, a CCP invasion seems unlikely if that's what you mean. The profitability isn't the clearest way to measure that. It could fall for a number of reasons without the geopolitical context changing.
Should TSMC fail due to one of the causal factors being geotech political interference on the horizon from the clash of China's Communist Party and the U.S. Wario Blob mic you would think ASML has wargamed plans to supply demand.
1.5 billion cell phones were sold in 2018. This represents 20% of the world population, assuming a negligible number of people buy two phones in the same calendar year.
Between 1.4 and 1.5 billion cell phones were sold in 2019, 2020, 2021, and 2022 (projected), representing roughly another 6 billion cell phones. That is, there is roughly one cell phone per person in the world "created after 2017."
Obviously, a significant number of people bought more than one cell phone in that time period; but, offsetting that, the average cellphone is sold more than once due to the used market. While it's hard to project this to actual estimate saturation of cell phones "created after 2017", it's very hard to imagine a case that the 6 billion phones from 2019 on don't represent at least another 10% of the population.
There is also advanced hardware besides cell phones.
Less than 30% of the world population owning hardware created after 2017 doesn't pass the sniff test.
The comment I was replying to was about advanced devices in general, not specifically TSMC.
That said, MediaTek is not the greatest example here -- they're fabless, and most of what they design is fabbed by, you guessed it, TSMC. The better counter-example would be Samsung, who fabs at least some of their own chips.
Arguably a lot of people might interact with servers that use their chips (and cpus and nvidia gpus), plus there’s game consoles which are fairly popular.
It's not about game consoles, it's about cell phones. According to some estimates, 83% of the world's population has a smartphone. TSMC makes a significant portion of the chips powering those phones. Currently TSMC has a significant performance lead over all the other fabs.
> most expensive part of gadgets nearly everyone has
I think that's where your assumption might be flawed. It's the most expensive part for PREMIUM offerings on the market that command high price points (e.g. iPhone, AMD Server chips, etc).
The everyday person, globally, can't afford such luxury items.
The customer can sustainably afford to pay in proportion to the value it delivers to them, not TSMC.
Apple has a finely tuned machine for turning performance into profits, and so can afford to pay more than others... and still have it make sense for Apple.
If TSMC instead had a collection of automotive OEMs as their top-tier customers, this math wouldn't work.
The point being... be in a business where your profit margins are a rounding error in comparison to your customers' bill of materials, and you won't ever go hungry.
A combination of profit margin, profit, and volatility are relevant. Apple’s lower profit margin is offset by the consistency of its profit margin and sheer size of the profit.
Maybe, but again, that's not enough information to make your statement a complete line of thought
The iphone 14 does not cost 2.4x more because one component is 2.4x more expensive. What is it now? Idk.
A lot has changed since the iphone 13 besides the chip, but it looks like the launch price is $200 more for the 128GB model for the 14 vs. the 13 (~25%). What's the total sales look like? What would the estimated total sales be without it?
for iphone it doesnt matter I would still got for discounted 13 rather than 14 as you gain mainly GPU core which changes not much. But for things like Macbooks smaller process means a lost, less energy used means less heat and less issues with throttling etc. This is big for MAcs but less so for ipads and iphones.
I cant believe this is posted on HN. A link to Appleinsider, about TSMC charging 2.4x more for N4 compare to N5P.
And the whole point of this Topic on HN was about TSMC Q3, which, should Apple really did paid 2.4x more for their roughly 70M of iPhone Pro A16, would have show up on TSMC quarterly results or future guidance.
Let's not forget about global chip shortages and limited production capacities. Apple is paying to be front in the queue as much as they are for the chips themselves -- as a shortage-induced revenue decline would be vastly more impactful on their market cap than paying extra on the BOM for key components
i mean the wafers for N4 are smaller so the yields are smaller and in the end it costs more. Also Apple doesnt do small chips or chiplets so some silicon they make is really big. Big die = smaller yields.
Usain Bolt is barely twice as fast as me, but he get's paid a lot more than double what I do to run. No one said prices had to be linearly proportional to performance...
Still surprised how profitable TSMC is. Especially as this seems to be below the market expectations (their stock is still falling).