Hacker News Reader

Comment by Jupe | Hacker News Reader

> This will take time. Decoupling is not an instantaneous process. China spent two decades becoming the workshop of the world; the rest of the developing world isn’t going to mirror that accomplishment in one or two years.

It is all about dependency trees. The US decides to divert imports from China to other countries, which in turn may import from China.

This reminds me of my day-to-day work so much; especially w.r.t. transitive dependency management. So:

If service A is 100% dependent upon service B which is 100% dependent upon service C, then service A is 100% dependent on service C.

At some point you need redundancy to protect against disruptions, outages, supply chain issues, policy changes, wars (either physical or economic), etc.

My (probably silly) question is: Does anyone know of "building block" or "core" services (such as authentication, authorization, BLOB storage, data storage, queuing) which are multi-cloud by design? Meaning service X offers SLAs for service endpoints hosted in AWS, GCP and Azure? If/when any of these "go down", consumers can (automatically?) flip a switch to route to another cloud service implementation?

View on HN

> It is all about dependency trees. The US decides to divert imports from China to other countries, which in turn may import from China.

This is already happening with more than ~80% of products that are "Made in Vietnam" and at least ~40% of products that are "Made in Korea." Not only do the raw materials come from China, but most of the work is done in China. The product is shipped overland from a Chinese factory to Vietnam or Korea, then some screws are tightened, some paint is slapped on, or a label is applied, and the product is re-exported.

Among other things, this is effective for tariff evasion.

And it's usually at least somewhat legitimate. It doesn't take much to make a product that's assembled in China "Made in Vietnam," even in a legal sense. Consider: https://casetext.com/case/ferrostaal-metals-corp-v-us

Basically, there was a tariff on Japanese steel. US-based plaintiffs -- steel distributors -- sent that steel to New Zealand, had it galvanized, and re-exported it to the USA as a "Made in New Zealand" product. A Judge holds that the galvanization process (which is laughably cheap) is enough to render the steel "Made in New Zealand."

When all's said and done, Chinese factories haven't slowed down and Chinese GDP is still rising (by more than American GDP) which tells you that the "decoupling" has to be in some respect illusory.

> The product is shipped overland from a Chinese factory to Vietnam or Korea, then some screws are tightened, some paint is slapped on, or a label is applied, and the product is re-exported.

The whole point of this article is to rebute the point you've made. From the article:

> The Economist’s basic argument is that because many of the goods exported to the U.S. from countries like Vietnam contain a lot of parts and components made in China, “decoupling is phoney”.

Here's one of the arguments the article makes against this:

> But there are reasons to think that at least part of the drop in U.S. dependence on Chinese imports is real — at least, the recent drop in the last couple of years. First, Chinese exports overall are down substantially from 2021.

[...]

> To the extent that falling investment in China is due to corporate de-risking, it’s probably eventually going to spread all the way up the supply chain. If you’re a U.S. company that moves production from China to Vietnam because of the risk of war, and your Vietnamese factory is importing most of its parts from China, you’ll know it. And you’ll know that you’re still exposed to China risk, and you’ll look around for alternative places for your Vietnamese factory to get its parts and components. That might take time — years even, given China’s dominance of manufacturing. But as long as the risk of conflict stays high for years, companies will keep looking for ways to build entire supply chains that never run through China. And they will succeed.

But Chinese exports to the US increased.

> In 2022, both U.S. exports to China and imports from China continued to grow for a third year in a row. U.S. exports totaled $153.8 billion, an increase of 1.6% ($2.4 billion) from 2021; U.S. imports from China totaled $536.8 billion, an increase of 6.3% ($31.8 billion); and the trade deficit with China was $382.9 billion, an increase of 8.3% of ($29.4 billion).

- dept. of commerce.

Those percentage increases seem to be lower than the rate of inflation at that time, which would mean the real value of imports and exports was flat or lower.

Total 2023 Chinese exports are projected to be fairly close to what they were in 2021, which is not that bad given macroeconomic trends across the world.

The risk for companies to build supply chains that completely avoid China is that they will be outcompeted by competitors who won't and who will make use of these loopholes.

Wouldn't that be extremely bad, given that 2021 was a year with many lockdowns and COVID related disruption especially in China? Surely 2023 should have experienced a rebound?

No it would be stupid good.

PRC export value went from 2.5T in 2019 to 3.4T in 2021, to 3.6T in 2022. It increased more during 3-4 years of covid than previous 10 years combined (1.4T in 2009). Primarily because 0covid left most of PRC undisrupted, most of the time, which setup PRC to export to most of world, especially global south. Currently everyone's exports down due to global slow down, but PRC exports down from stupid high peak. It's more surprising how much of exports PRC is holding onto due to artificial conditions caused by covid. Reverting back to 2021 (3.4T) would be a 5 steps forward 1 step back situation vs, reverting back to 2020 (2.6T) which would be status quo.

You are right, thank you for the detailed answer.

The average Chinese was locked down for less than the average American, and while the lockdowns were far more severe, they weren't much worse for the economy.

2021 was a great year for Chinese exports especially due to COVID. It's one of the countries that suffered the least economic damage, and it's exports actually grew immensely compared to previous years.

Xi Jinping has already enacted likely unpopular economic policies that were failing without the US starting its economic cold war. He's already enacted purges and other violent measures to centralize power as I understand it.

What degree of economic damage would fundamentally destabilize China politically such that Xi is deposed?

If stupid me is thinking about such vulnerabilities/risks, would foreign businessmen bet on Chinese factories, or move them to the newer/cheaper labor markets I've been hearing about for the last 10 years like Vietnam and Malaysia?

And Jinping is 70, do you think he's lasting to 80? When he croaks, due to his elimination of what semblance of political power handover / elections that used to exist in China, it will be a bloodbath.

"Chinese factories haven't slowed down and Chinese GDP is still rising"

This line is so wrong. You can't have 22% (at least) youth unemployment rate, and 19th straight month collapsing land sales revenue https://www.reuters.com/article/china-economy-landsales/chin... (which is how local government get most of its money), and your biggest and second biggest real estate developer going bankrupt, and still support a 5% GDP growth as mentioned by the Chinese government. just doesn't happen in a real world.

And there are so many reports of factories closing down, especially in the Shenzhen and Dongguan districts, either online videos or news reports https://thediplomat.com/2023/04/dongguans-industrial-woes-ch..., that I don't know why you're attempting to lie outright.

Land sales down 20% YoY is why growth is 4-5% instead of 8% - the effects of 3RL / crushing RE is built into cost of CCP targetting modest growth. Western reporting getting useful idiots to think PRC RE collapse is going to crush PRC growth to 0% is retarded, RE maybe large it's but not even close to plurarlity of economy - many other sectors, increasingly intermediate / high value like auto exports outperforming and making up lost RE in economic rebalancing.

Increasing youth unemployment last few years also has little NET growth impact because PRC is graduating MORE every year, 12m this year vs 8m from ~5 years AGO. What's happening is is MORE youth in absolute numbers are entering work force, even if relative youth unemployment is higher. And net unemployment is 5%, meanning youth get work eventually. They just have flexbility of chilling at home to find job they want, or fuck around and realize they have to do something - side affect of PRC having high youth live at home + high house hold savings rate + low student debt.

Industry has been closing in coastal due to land price for years. Low value/light has been moving abroad due to labour price. New (high value) industries are being shifted to interior industrial clusters. PRC export value grew by 1T in the last 4 years, more than the previous 10 combined. West might be trying to decouple, but PRC has been coupling more than ever, especially with global south.

Truth of the matter is PRC has just undergone the most intensive phase in globalization in human history, and exports down only due to weak global economy, but it's still a 5 steps forward, 1 step back situation. The lie is motivated useful idiots who pick a few indicators to rationalize dampening growth to 4-5%, aka nearly HALF, is still not enough, and real growth must be near 0 because PRC is collapsing due to exogenous crisis when in reality PRC proactively triggered 3RL and other policies in a MANAGED unwinding because they could do so and still have room to hit 4-5% growth.

> When all's said and done, Chinese factories haven't slowed down and Chinese GDP is still rising (by more than American GDP) which tells you that the "decoupling" has to be in some respect illusory.

Can we really know that anymore? China has stopped publishing a majority of economic indicators.

Satellite photos of night-time lighting and shipping movements are probably reliable.

2022-2023 YOY decrease in china shipping volume

https://www.freightwaves.com/news/reality-of-ocean-container...

If you don't believe Chinese statistics, you can check that US import from China is back to pre-COVID volumes and rising.

Honestly, I'm tired of this "oh they're totalitarian, so we cannot believe their statistics when it's good" stance. It looks so much like a comfortable self-deceit.

> Honestly, I'm tired of this "oh they're totalitarian, so we cannot believe their statistics when it's good" stance. It looks so much like a comfortable self-deceit.

I’m tired of it too, but I think it’s better than “not use any data points” or “trust the super subjective data points put out by a totalitarian regime so they aren’t comparable to other countries.”

I don’t think the US is perfect, but economic indicators that they put out are probably as good as they get, all things considered.

> I’m tired of it too, but I think it’s better than “not use any data points” or “trust the super subjective data points put out by a totalitarian regime so they aren’t comparable to other countries.”

I understand this sentiment when it's applied to shithole countries where statistics cannot be verified in meaningful ways. But China is too big and too integrated into the world economics to fake or dismiss Chinese data. You can independently prove a lot of Chinese data by comparing it with their western trade counterparts.

I don’t see how you can trust numbers from a government which fairly openly treats truth as malleable. Also, cadres are evaluated to a certain extent on their economic metrics, and of course a metric ceases to be reliable when your evaluation depends on it. But the real kicker is that the outgoing prime minister, Li Keqiang, did not trust the official numbers and developed his own calculation using other numbers. [1]

[1] https://en.m.wikipedia.org/wiki/Li_Keqiang_index

> China is too big and too integrated into the world economics to fake or dismiss Chinese data

Chinese statistics are notoriously biased. There is a cottage industry of hedge funds using satellite imagery to correct their figures and trade on it.

Can you name some of these hedge funds?

I know someone in the commodity business who uses satellite imagery for monitoring demand.

Export data has a counterparty though, so I don't see how it's anything more than a thought terminating cliché.

Right, in this case I like using the counterparty and US imports from China in place of what we want.

But I think the complaint was more to say why they weren’t looking at other data they would like, and other countries provide reliably, and instead must limit to what the US and other countries report.

Yeah the US leadership is smart enough to almost never lie outright. There are lots of more nuanced ways to technically tell the truth but still significantly warp the information to your benefit. I’m kind of surprised that China seems to be less savvy to this approach.

> There are lots of more nuanced ways to technically tell the truth but still significantly warp the information to your benefit. I’m kind of surprised that China seems to be less savvy to this approach.

That is the national government's approach, for the most part: tweak the basket of goods you use for inflation here, pick the best looking among several measures there, and so on. The overt manipulation occurs at the subnational level: Chinese statistical work is heavily decentralized and officials are promoted (or not) partially on the basis of meeting growth targets. The national government corrects for this to some degree - but estimating just how manipulated the data is is hard, and provides another lever you can pull to push it one way or another.

Federal statistical agencies in the US conduct more and better direct research and report through fewer (though not zero) layers of politically motivated actors.

It's just not that people distrust Chinese statistics, it's that they have stopped publishing statistics.

One need only compare the official Beijing air quality number against what you see on the ground to know that you simply cannot trust official Chinese reports.

US import could be clearing out stock at firesale since lots of Chinese manufacturing is liquidating. China imports are declining and IMO (I could be wrong) more telling for long-term producer health (china is not exactly resource independent):

https://www.reuters.com/world/china/chinas-july-exports-impo...

> US import from China is back to pre-COVID volumes and rising.

Source?

Can’t respond to this other user but their source confirms that there is a consistent drop in imports from China since late last year (goes down each month.). Saying “we’re back to pre-Covid levels” is a strange thing to say when Covid clearly saw a huge spike in imports, which are now decreasing. Their framing of this entire conversation seems incredibly dishonest.

Here you go, hope US government is reliable enough for you

https://www.census.gov/foreign-trade/balance/c5700.html

This data shows that imports from China are down significantly year on year, though?

Seems like it.

Honestly did you even read the article? "Yes, there are cases of Southeast Asian nations slapping their labels on Chinese goods and then sending them to America to allow Chinese manufacturers to get around tariffs. But look at the graph of U.S. import shares above, and you’ll see that if 7% of imports from Asia ex-China are due to re-exporting, that would only shave 1.7 percentage points off of Asia ex-China’s share — only a fraction of the 4 percentage point increase in U.S. imports from that region since 2019." It refutes this point like 2 minutes in.

The article does nothing to refute the case, because Noah Smith has engaged in a little bit of sleight of hand in that section you quoted.

The 7% number is from 2018 data (!) -- yet Noah is talking about "since 2019." And, looking at the graph again, the major shift occurs in 2019.

So Noah's "7%" (again, from 2018, largely before the "decoupling") is totally irrelevant, and he handwaves away the possibility that China has since done an increasing amount of exporting to SE Asia for re-export to the USA.

I'm in China, and I think you're wrong. The situation is absolutely catastrophic and we're slowly falling down the cliff. Those GDP numbers are BS, in the US, France or China, since they count public spending. Anyone can build ghost cities for billions, add it to GDP and show a growth of GDP: meant nothing more than a big dick number.

Truth is when they ask a random dude on TV if he's ok with drinking Japanese radioactive water (the trendy distraction these days), he replies that he didn't intend to make children so he doesn't care if he loses fertility due to contaminated water...

Nobody gives a fuck about shit, we all pretend to work while they pretend to pay us, we get jailed for months in our own homes when there's a stupid sneeze in the city and get all released under the "it's just a fever, wait a few days" admonitions of the same doctors who captured us when we all get it at the same time, we're either completely ignoring the communist bullshit and doing nothing they expect us to do, or totally believing in it and doing only the BS they expect us to do, which produces nothing.

Kids are unemployed, elderlies are fascists, politicians are spineless and factories impotent. The GDP growth is as real as the president's official salary.

>our own homes

our leased (up to 70 years) homes

This is the same as in England, the leasehold vs freehold system is difficult to differentiate in practice, a bit like common law vs civil law.

You buy a house, there are expectations it's yours. Each time it's been tested the government had a choice between murdering desperate owners or accepting to just renew on the same terms. They chose to renew.

Freehold offer no more security: I come from a civil law, freehold country, well, the government can seize houses it wants at low cost any time anyway. Sometimes to build a highway, sometimes to rent to immigrants.

> This is already happening with more than ~80% of products that are "Made in Vietnam" and at least ~40% of products that are "Made in Korea."

The Philippines are famous for this too.

https://www.cbp.gov/sites/default/files/assets/documents/201...

https://www.cbp.gov/sites/default/files/documents/Vol_48_No_...

Isn't Vietnam labor cheaper? also even if these countries are just giant warehouse of Chinese crap wouldn't that mean if Chinese supply did go down there's more stuff in the pipeline which helps buffer shortages and price spikes etc.

> This is already happening with more than ~80% of products that are "Made in Vietnam" and at least ~40% of products that are "Made in Korea."

I wonder how long they are willing to do that. Those countries aren’t on the best terms with China either. Doubt they want a dependency on China if they can help it.

Wouldn’t they want to take over the manufacturing (along with it the profit) too instead of just re-exporting for a profit? They are in prime position to do so given they are the gateway for China’s goods - and can cut China’s goods out and substitute their own.

If they were able to build a toaster (or whatever) profitably than they would have already done it. Doing last stage assembly most likely requires very little capital investment and much less labor.

I feel like that position is similar to being a middleman - in which case its clear the skillset to be the producer is not the same.

> If they were able to build a toaster (or whatever) profitably than they would have already done it.

For countries like Vietnam, that might be due to lack of investment.

For the last few decades, everyone was concentrating on China. This I believe has began to change - foreign investment in China is down if Im right.

On the other hand relationships between Vietnam and the West has been thawing the last decade or so - especially with the US.

Countries like Mexico are also looking to take on some of China’s manufacturing load - their proximity to the US is an advantage.

Tariffs on China effectively gives all their competitors a competitive advantage which will only help develop said competitors.

How is this suddenly possible against the "physics" of free market forces, that were unable to move when it was about wages, workers rights or the environment? Almost as if it's all made up..

Agreed. I'm no macro-economist; I was trying to relate the dependency tree aspect back to my are of work.

But, you appear correct, the decoupling does appear to be an illusion.

That’s step one. Then the other developing companies will need to climb the dependency tree to make more and more of the materials locally. It’s what India did with car manufacturing, eventually building the entire vehicle in country. They are now the biggest exporter of tractors in the world. Same thing with pharmaceutical manufacturing as well. The space industry was the same way with imported tech from the Soviet Union but now it’s all built indigenously along with local advancements.

It seems like in 2021 Germany was the biggest exporter of Tractors in the world, with 16.2% of the total https://oec.world/en/profile/hs/tractors But maybe you have more up-to-date or different data showing India ahead.

Mahindra is the largest producers of tractors by volume in the world. This may be on the company level and not national level

https://www.mahindrafarmequipment.com/services/tractors#:~:t....

GM traces all their dependency trees down to the mines or other sources of raw materials. This is often 6-7 levels deep. However they know what to watch for in the news and so when something happens they can jump to find an alternate, which often gives them 3 months before the production lines stop.

They also do this so that they know nobody is using slave labor.

Yeah, I'm not convinced, given GM's (and most automakers) reaction to Covid and the continuing chip shortage.

Maybe things have changed; I'd certainly be glad to hear it. But we all know of Wall Street's long history of hatred towards physical inventory.

> Yeah, I'm not convinced, given GM's (and most automakers) reaction to Covid and the continuing chip shortage.

Automakers (and everybody else) reacted very quickly and shut down/decreased staff almost immediately.

Where they got caught out was in assuming that the semiconductor industry would return to normal quickly. That was a huge mistake.

Everything in business is so hyper-optimized to hold no inventory that nobody can absorb demand increase shocks of any amount, anymore. In addition, the fact the companies actively downsized during Covid meant that things couldn't even go back to pre-Covid levels quickly.

GM miscalculated what COVID would do and made mistakes. Having information isn't always enough to make the right decisions. All we know from COVID is GM isn't perfectly able to mitigate issues, but they still try.

> But we all know of Wall Street's long history of hatred towards physical inventory.

Accounting and taxes on physical inventory are pretty stupid and cause a lot of Wall Street's hate for inventory. The more costs there are to storing inventory the less desirable or even viable it is to store inventory.

When all countries shut down, it does matter? There is no alternate or backup. What could you have reasonably expected GM to do, even if they had perfect information?

Not cancel all their orders and end up at the end of the line for chips.

Answering your technical question: Some companies let you deploy their "building blocks" onto multiple clouds (you typically choose the exact region/availability zone), and then also allow you to build proper multi-cloud deployments. There is usually no switch to flip, but these deployments (if properly setup) should tolerate if one region/cloud goes down.

Two examples from the top of my head:

Confluent Kafka: https://developer.confluent.io/courses/hybrid-cloud/intro/ MongoDB Atlas: https://www.mongodb.com/docs/atlas/cluster-config/multi-clou...

For higher level services, I can't think of an example. For the examples above, you're still thinking at VM level, whereas you typically don't do that if you purchase a higher level service.

This has always been one of the selling points of CockroachDB, the ability to seamlessly run multi-cloud redundancy.

Awesome - good to know. Thanks!

I'll read this later: https://www.cockroachlabs.com/blog/multi-cloud-deployment/

The article addresses this point directly.

In order for automatic switching to work though, you'd have to propagate everything between all 3 services, at all times, no?

Isn't that cost prohibitive?

That’s exactly it. It’s not that hard to store files on x number of cloud services.

It’s very hard to convince the money people that we should double our cloud storage bills in case all of AWS goes down.

Every new level of dependency is an opportunity to refactor and swap deeper dependency out. There is a big diff between using library A and using your own frontend B that now depends on A but can swap it out easily for X.

Let's say you have a dependency A from China. But all the software developers that work and specialize in A, live in China. If you ban dependency A, what you'll have is a dependency B from Mexico that is calling dependency A.

In order to really replace A, you'll need to hire, train and form software developers specializing in A. This is both time consuming and expensive.

This falls in the same line of some people (economists) who think that printing money will generate wealth. Sounds familiar?

There is also on shoring, which isn’t switching China from a direct to transitive dependency.

Reply on news.ycombinator.com