It's quite peculiar to encounter such simplistic perspectives about the Taiwan situation on this platform. Makes me doubt this platform's main ideology as a whole. As a European, I confess my understanding of the issue is quite rudimentary, yet I'm taken aback by short-sighted projections such as:
Suggestions that seem to disregard Taiwan's autonomy over its own affairs, as if U.S. interests can dictate decisions like purchasing all of TSMC and relocating them and their personnel.
The assumption that China's primary interest in Taiwan is solely due to TSMC, with claims like 'China will lose interest once TSMC ceases its operations'.
> The assumption that China's primary interest in Taiwan is solely due to TSMC, with claims like 'China will lose interest once TSMC ceases its operations'.
Whether or not that's its primary interest Taiwan, TSMC is partnering with the US and allies to geographically diversify its most advanced manufacturing. They are taking all the eggs out of one basket so China can't use its proximity to the basket as an advantage through threats or any designs they may have on acquiring the tech they have been unable to produce domestically.
Edit: To add a little more flavor. General population are getting more aware of this but they have a surface level understanding of it; mostly memes and soundbites they parrot like "silicon shield".
The silicon shield certainly has some merit to an extent but long term this causes more problems than it perhaps solves.
The all the eggs in one basket problem:
* China knows the world depends on those chips now, and all the most advanced chip manufacturing is sitting right off their coast. This really strengthens their hand in all sorts of ways.
* The world does rely on those chips, but the worlds major liberal democracies are unified in their commitment to ensuring this tech remains in-the-family so to speak. They can't destroy the basket because all the eggs are in it and everybody needs the eggs.
Another consideration; we were committed to Taiwan's defense before they came to dominate global chip manufacturing. Independent Taiwan is important to the Indo-Pacific strategy.
Edit2: I said "all the most advanced chip manufacturing" and that's not true. Intel has all the capability or is right behind perhaps but it does not have the capacity to meet worldwide demand.
> It's quite peculiar to encounter such simplistic perspectives about the Taiwan situation on this platform. Makes me doubt this platform's main ideology as a whole.
In theory, this platform should be about curious intellectual discussion primarily about tech.
In practice, it suffers from upvote-itis which amplifies popular opinions and worldviews and silences others. That's a horrible environment for sharing knowledge about geopolitical issues that span far beyond SV.
>Suggestions that seem to disregard Taiwan's autonomy over its own affairs
Fair, but this autonomy seems to be tenuous at the moment and grows more so as time passes. You must be aware what the US is willing to do when it comes to national interest. Chips are on par with energy/oil, possibly regarded as even more important economically and militarily. I'm not sure if the US regards Taiwan's fall as inevitable or how they would act if they deemed it so. Regardless, a war over Taiwan would significantly slow chip production.
>as if U.S. interests can dictate decisions like purchasing all of TSMC and relocating them and their personnel.
I'm sure there will be a carrot and a stick, but if this happens, an offer will be made that TSMC can't refuse. TSMC seems to be more than interested in moving their fabs off Taiwan land, so I don't think this would be necessary.
>The assumption that China's primary interest in Taiwan is solely due to TSMC, with claims like 'China will lose interest once TSMC ceases its operations'.
Yes you are correct about this. China's current interest in Taiwan goes well beyond TSMC.
> Fair, but this autonomy seems to be tenuous at the moment and grows more so as time passes. You must be aware what the US is willing to do when it comes to national interest. Chips are on par with energy/oil, possibly regarded as even more important economically and militarily.
> I'm sure there will be a carrot and a stick, but if this happens, an offer will be made that TSMC can't refuse.
I comprehend this U.S.-centric viewpoint. However, from the perspective of an independent nation state, this narrative sends a cautionary message: 'Beware of creating a product that becomes too integrated into the U.S. supply chain, because you will get rich but risk forfeiting your right to self-determination.'
It is somewhat disappointing to see that, on this forum where more nuanced discussions are expected, this is the default (and appreciated) perspective.
Couldn't the opposite also be true though? Not having a product that is heavily integrated into the U.S. supply chain means the U.S. may not come to your aid when your neighbor invades.
This is very true! Which is why this is a fine balancing act on Taiwan, and not the simplistic 'hurr durr US interests can't allow that'. They need to carefully weight their options, and decide what is best for their nation under the circumstances.
The US absolutely would be very carefully considering options.
Their primary preference would be for no conflict, so they put up a front of bravado to discourage any aggression. He question does remain, if they'd actually respond to such an invasion.
Could we say that almost any powerful country will try to protect their critical supply chains with every available mean? This is just plain realpolitik.
The majority of USA's most critical supply chains come from China proper (PRC), not Taiwan.
If this was fundamentally about economic security for the average American, Washington would not jeopardize those with economic and military brinkmanship.
Could you list some of these critical supply components? I mean, TSMC is unique, ASML (Dutch) is more unique. What other unique components does China have? Obviously Apple depends on Foxconn but we can envision that they could be replaced within a 10-year period while TSMC or ASML not. Really curious.
Simplistic but more realist. Most TW analysis/commentary are simplistic, but few are realist. TSMC is one of those refreshing topics where even American useful idiots who shill LIO values tend to goes full mask off on realisim because the geopolitical interests involved is so stark. 99% of the time on western platforms, endless paragraphs of ideological propaganda drown out realist narratives. Until it doesn't. Then it's might is right, but frequently expressed in equally simplistic terms. It's not an exclusively HN issue and TBH like half the dumb hot takes are directly from US gov voices and dessminate their way down.
Because transporting any large company from location A to location B involves a lot more than just shipping the machinery. TSMC is the tip of the iceberg, the thing you can see jutting out of the waters. We don't see the invisible - but implicit - support structure that makes the company possible.
Freakonomics podcast had a couple of episodes on Taiwan some time ago, and one of the points brought forward was that TSMC is a spearhead. In Taiwan, there is more semiconductor industry, and an established education structure able to produce the ongoing stream of skilled workers at every level. These feed into each other, and TSMC is at the apex of the pyramid. To produce an environment where you could transport TSMC to and have it continue to thrive, you'd likely need to first maintain a strategic focus for something like 15 years to build the necessary support structure.
I suspect that China is looking at this, but looks at it from another angle. If they invade Taiwan, they will end up disturbing the support structure that makes TSMC, well, TSMC. The edge of Taiwan's semiconductor industry is not that they have TSMC. Their edge is the entire environment that makes TSMC possible in the first place.
Or to be really blunt about it: once the CCP believes they can do without Taiwan's semiconductor industry, they'll send in the troops.
> once the CCP believes they can do without Taiwan's semiconductor industry, they'll send in the troops
That's still a simplistic point. Will they not consider the practicality ot invasion?, like Taiwan beating them on its own, USA getting involved as promised, etc?
If not, then we'll likely have a war between major powers or they'll get to declare victory as with vietnam.
Of course they do! And it would be totally fine to make an offer. But to purchase an entire strategic sector from a country would warrant some sort of government intervention, no? TSMC getting fully bought out and moved out of Taiwan, as comments suggest, would not be possible without the government's approval, and why would a government approve that if it had its own interest at heart?
It's why I was giving the example of EU or China or whatever offering up to buy the entire US military industrial complex. It would be impossible, and not only due to lack of funds.
So is yours. OP used the word "dictate". You transformed it into "offer a better location". And e.g. USA is not a better location for TSMC for a multitude of reasons, so that "offer" wouldn't fly unless it were coercive.
TLDR is no, strategic companies with massive state industrial funding have little autonomy. Even TW (and US) semi workers don't have autonomy to work in PRC semi industry for greater pay. Why is it unseemly to offer a person better pay and locations - TW workers sure liked getting 3-5x salary while being an hour flight from home, until TW gov said no. Ditto with US but on pay. Because it undermines strategic position of incumbant / host.
“I don’t like its location, and I’ve reevaluated that.”
“I feel better about the capital that we’ve got deployed in Japan than in Taiwan”
“[TSMC is] one of the best-managed companies and [most] important companies in the world.”
“There’s no one in the chip industry that’s in their league, at least in my view”
“Marvelous people and marvelous competitive position and everything, [but] I’d rather find it in the United States.”
Buffett said his reassessment of the company was in “light of certain things that were going on.” He had previously pointed to geopolitical tensions as a concern.
---
TSMC is a key supplier of many companies, including Nvidia.
Not an ideal situation.
What do people expect plays out over the next few years?
> What do people expect plays out over the next few years?
China fantasy version: China takes Taiwan and TSMC and everybody minds their own business. China miraculously converts this move into de facto monopoly. America can't get chips and tries to negotiate with China's tight new ally Mexico, losing California in the process! Holy sh*!
Maybe more real: China takes TSMC like a chess piece, tries to put it back on the board like in Xia...Shogi (sorry, my Japanese is better here). Except somehow (sanctions, brain drain, sabotage, etc) the piece metaphorically crumbles to ash. Buffett is still so very glad to be out.
Another one: Taiwan and foreign partners with lots to lose like US, Japan, and SK successfully turn Taiwan into a meat grinder. TSMC becomes a godforsaken piece of charred land. Miraculously a very-TSMC-like corporation appears split between those countries with some kind of Taiwanese stake as well.
Lots of possibilities to think about. Could happen in the next 5-20Y rather than a few though. WW2 for example had a very long build up from expeditionary moves to the bigger strategic conflict stuff if you look from the Japanese perspective, and this pattern has a lot of benefits to China given factors like US administration churn, etc.
Interesting how none of these imagined scenarios involve China not attacking Taiwan. Like, for example, Xi dies slightly earlier than expected and the next, younger, head of state doesn't have the "Oh god I need to cement my 'legacy' before I turn 90 years old" problem and can indefinitely kick the Taiwan can down the road.
Yesterday, I read something in the FT where it was suggesting a lot of Chinese leaders resent the west for goading it into attacking Taiwan. The problem is the west isn't doing that, but when China acts as though it is, it escalates tensions, and the west responds in-kind. The obvious solution for China is don't attack Taiwan, but Xi thinks he's entitled to be the leader that reunifies China. He lacks Deng Xiaoping's patience.
Ahhh, I mean there has historically been an ebb and flow in the focus of Chinese leaders. Some are outward focused, some are more domestically focused, some are more interested in legacy, some are more interested in financial overhauls. Yada yada.
In context of the article, I think this scenario would still be expected to bring the market cap of TSMC to near-zero. I believe GP's second scenario was intended to capture the predicted realities of a Taiwanese military victory.
The fact that Putin and Xi are close and Putin made a move unthinkable a few years ago shows that these autocrats mean serious business. It's not some fiction novel story anymore.
Taiwan's chip fabs are not the only reason that Taiwan is strategically key to any country that is not CCP. Enormous amounts of shipping, especially to/from Japan sail through those straits, and ceding control to CCP is not a viable strategic option.
Even if TMSC and other fabs are entirely moved off Taiwan, or reduced to Bhakmut-style smoking rubble, CCP will still have a serious fight on it's hands to gain control of Taiwan.
I tell my friends and family, "If you ever see a headline saying Taiwan is getting blockaded, or receiving airstrikes, or whatever it may be, drop what you are doing and run to the nearest Apple Store and buy everything you can afford. If you still have money, go to every Best Buy and repeat. Anything Apple, along with Game Consoles and Nintendo Switch Game Cards, will be among the worst affected. If it doesn't work out, you'll sell it for only a 10%-20% loss. If it does work out, it will be years before you can buy those products new again."
I suspect the temporal order will be reversed. From what I hear, China will have to spend months amassing troops and material near the coast, in a way which cannot be hidden, giving the world at least 3 months to "prepare".
I expect speculators will deplete Apple Stores' JIT stocks well before any belligerence begins. Some of those speculators may place their bets too early and lose money.
> China’s military flew 38 fighter jets and other warplanes near Taiwan, the Taiwanese defense ministry said Friday, in the biggest such flight display since the large military exercise in which it simulated sealing off the island earlier in April.
I'll buy that Chinese in China have that view, but flying 38 fighters near Taiwanese airspace is China actually using force.
At the same time, I do see China's larger gripe. Between Japan, South Korea, Taiwan, The Philippines, and Singapore, China's access to the open seas is boxed-in by western allies. Russia has a similar gripe with a warm-water port in Crimea and NATO slowly closing in on Russia by growing the alliance east.
Yea I totally agree. That’s literally the thing I thought about as a rebuttal. This Johnny Harris video elaborates on it as well https://youtu.be/VNZ0so0LCoM
But at the same time. You would think that if China does in fact believe Taiwan is theirs, then why would they bomb their own (eg. valuable TSMC, etc) infrastructure vs play the long game and take over politically.
Also, their sorties are more like a dog barking. It’s showing force, like North Korea doing those parades and launching missiles.
> Russia has a similar gripe with a warm-water port in Crimea …
Russia has multiple ports on its own black sea coast and before 2014 was building them out, especially Novorossiysk, to host the Black Sea Fleet, for when the lease of Sevastopol runs out.
The US should offer to pay for relocation of all TSMC personnel and rebuilding of all fabs here on US soil. I'm not sure there's much of an upper limit to what this would be worth. $2B? $10B? Something like that. Of course the patriots won't want to leave - and rightly so. It's their home, and democracy is worth defending against autocracy. But the offer should be made, IMHO
EDIT: Estimate is off by 100x. I also agree that Taiwan would not like to lose the leverage TSMC gives them. I still think the offer should be made.
TSMC is owned by the state, they can have their cake and eat it too. Send low nm process to the US in exchange for military hardware and continue to own the IP which means they could withhold it from the United States during time of war unless they have the support of the United States. It’s not clear why this can’t work, it’s what they are currently doing…
A single 3nm fab is ~$20B. In the last decade TSMCs investments in Taiwan capacity (5nm-28nm) exceed $100B. All the other fabs and infrastructure many $100Bs. Not including contractors or support personnel TSMC employs ~50k in Taiwan. It would be an enormous undertaking.
Seems starting with moving the most modern lines out of harms way, and building all new lines not on Taiwan, they would soon reach a state where what is left behind is multiple generations old, so worth far less
Certainly a good idea even if moving nothing, to also build nothing new within China's reach.
Isn't that how a fab gets built? Equipment is moved to the site and assembled? Seems it would be more expensive due to needing to disassemble plus you'd be offline a fab.
Yes, of course moving would be an extra expense and involve downtime. This would be weighed against the risk of losing it all due to CCP invading, either through destruction or just being seized.
[Total Equip Loss Cost] * [Probability of Loss]
vs
[Additional Cost of Moving] + [Additional Cost of Downtime]
Perhaps there's an argument to be made about U.S. hegemony when such proposals concerning the strategic assets of another country are considered. What if we Europeans devised a plan to acquire all strategic manufacturing assets from the U.S. - including tech, military, and energy sectors - along with their personnel, and relocated them to the EU? Wouldn't the federal government likely have something to say about that?
We call that Capitalism. If you don't have a plan to stop it, assume it's going to happen. I'd like to refer you back to the documentary idiocracy. There is line in that film that seems to resonate with many humans especially Americans, "I like money, you like money too? We should hang out"
I'm not buying free market capitalism with regards to strategic resources of another country. If 'we call that capitalism' is all it took, the US wouldn't need the military it has. Again, this feels like a simplistic view about a country with its own vision of the future...
You should because that's what these companies or "resources" are actually thinking. If you asked the CEO or shareholders who the company belongs to they'll claim it's theirs and that where it operates from is irrelevant. The US has the military it has because of capitalism, warfare costs tremendous amounts of money and resources and so a lucrative market was birthed.
It's just impossible to get my point across to some people. What happens when a company gets too big and is split by the government? Do the CEO and the shareholders go request this?
> We call that Capitalism. If you don't have a plan to stop it, assume it's going to happen.
Maybe that still holds true for Europe since the US doesn't want to piss of their strategic allies too much, but unless you're living under a rock, you've seen how that works for other less friendly countries.
If America pays for this because of fear of Chinese bullying, what is next? If you let the Chinese know America will pay, and people can flee and the Chinese can take what is left behind, what else does America let the Chinese take over? Middle East, Africa, South America?
I don't think being a coward and running away from a bully is the right thing to do.
Er, you might consider looking into China's so-called Belt and Road [0] initiatives with regards to Africa. The boomer-era IMF strategy of predatory lending to third-world countries with an eye to controlling export infrastructure via impossible loans has been noted and improved upon by the Chinese. IMO "we" (read: the Pax Americana) have already handed over much of Africa.
I believe TSMC will diversify away from Taiwan and they will probably need to do it as quickly as they can. This is already happening; they are building in Arizona, but it takes a while to build a foundry and get it running. They should have started a decade ago. Hong Kong was probably the catalyst today. Companies who rely on TSMC will also need to diversify as much as possible as quickly as they can incase TSMC can't. I'm not sure if Berkshire doesn't think they can do it, or just doesn't like the risk. They seem fairly risk-averse.
Once this happens China will probably feel more emboldened to take Taiwan because US industry won't be so reliant on it and less likely to intervene. I'm not saying the US won't, but they would be less incentivized economically and in regards to self interest. I don't think the populace has the stomach for a long protracted, potentially nuclear war over Taiwan.
China will become significantly more isolated on the global scale if this happens, but they seemed determined to do it. They've fallen into a single-man dictatorial power structure under Xi. The China global economic experiment seems to have failed. It's a waiting game.
Want to point out that the global economic importance of Taiwan will decrease if TSMC is able to diversify away from Taiwan in any significant manner, meaning China stands to gain significantly less economically by invading Taiwan. Most of Taiwan's importance currently seems to be directly related to their position as the primary supplier of chips for global corporations.
No. China has wanted Taiwan since before TSMC was major factor.
This is about geopolitics, and fulfilling the destiny of the Party/Nation that has for decades been embarrassed that their (in their eyes) territory was held by others and supported by the West.
Sorry, I should've specified that I was specifically talking about potential economic incentives for China. I am aware to a lesser degree that they wanted the land for alternate reasons.
It is shocking how poorly they treat their employees. No wonder they are lagging: They can't keep people on the floor because they are poorly compensated and treated poorly.
The amount of middle manager bloat is STAGGERING. 5 bosses for every person actually doing work.
As my friend put it: "What if we retained employees instead of running a strategy where we expect 20% turn-over every quarter?"
I bring this up because I see Intel as TSMC's primary competitor, and I don't see how they can hope to compete with how inefficient and incompetent they are.
TSMC is pretty famous for working their researchers 80 hours a week and making them work night shifts. They might have more autonomy and less bureaucracy but I don't think most would consider them a company that treats employees great.
Their advantage is that they've made themselves into Taiwans entire national security strategy so they're able to push an idea of working yourself to death as a selfless patriotic act.
Do you think Apple will stop using TSMC sometime in the future? Berkshire does own 1/20th of Apple, and I wonder if sentiment from one of their largest investors that TSMC is not a safe investment might sway them working with other chip manufacturers instead. Or maybe even furthering Apple's vertical integration by having them make the chips too (we're talking far, far future here, obviously they can't set up chip manufacturing within even a decade)
I'd probably lean towards Apple making their own chips, it would definitely seem to be a logical continuation of Apple's agenda, and they are certainly among the few companies that have the capital to make it happen. They already seem to be close to running into issues in terms of securing chip supply (e.g. having to purchase 90% of TSMC 3nm process ahead of time), so I suspect they already have plans to expand into chip production in the long term future.
Apple has always had a multi supplier strategy. They also used Samsung at 1 point (and still do). It's whatever works at Apple so no difference. They'd eagerly switch as needed.
Not sure how many producers have the available long term supply to meet Apple's demand. Samsung might already have contractual obligations for their short term supply with other manufacturers so it might not quite be as straightforward as just switching overnight.
Does anyone know of a good piece of writing about what has made TSMC so successful, what makes their management so good, etc.? Seems like an operational exemplar I'd like to learn more about.
That rock must be heavy on top of you. Pushhhhhh! I mean if China invaded right away they would have lost money, so clearly the know something that the public does not. If I had that kind of intel, I would pull out also (that is what she said)
Suggestions that seem to disregard Taiwan's autonomy over its own affairs, as if U.S. interests can dictate decisions like purchasing all of TSMC and relocating them and their personnel.
The assumption that China's primary interest in Taiwan is solely due to TSMC, with claims like 'China will lose interest once TSMC ceases its operations'.