Live data from Hacker News

The semiconductor bust – still coming?

asianometry.substack.com

61–70 of 75 posts

Re: The semiconductor bust – still coming?

#61
This article uses the wrong measurements on inventories.

-You have to look at inventories in relation to sales, either shipments or new orders (inventory-sales ratio)

-Manufacturers’ inventories are up sharply because of materials and works-in-progress. Their finished goods inventories are sill low relative to new orders. It’s the bulges in the durables finished goods-to-new orders ratio that indicate a crash in semis coming. As of March, there are zero signs of that in any of the silicon-heavy categories.

-Anyway, wholesale inventory-sales ratios are a better indicator. https://fred.stlouisfed.org/graph/fredgraph.png?g=OVAY

Re: The semiconductor bust – still coming?

#62

Earlier quoted context omitted.

Who is participating in bad faith? If you want long term security of supply pay for it. Suppliers love long term contracts and orders. Buyers usually don’t go for them because it’s cheaper on average not to. No one signed a contract. The market price moved. That’s capitalism. If you want to prevent it happening again you can pay for it, by having a long term contract, stockpiling or multiple suppliers. Many, many mar…

> Who is participating in bad faith? Hoarders. (In this instance, as per upstream comments, possibly coordinated, possibly by CCP.) There’re fairly illustrative examples in my comment that you are ignoring. You should address them if you object to that idea. > If you want long term security of supply pay for it. A single player cannot outbid a state and/or many players acting maliciously in accord.

Hoarding is not in "bad faith". If people anticipate a shortage in the future, storing some of the existing supply is socially beneficial. It raises current prices less than it lowers future ones.

Re: The semiconductor bust – still coming?

#63

Earlier quoted context omitted.

Who is participating in bad faith? If you want long term security of supply pay for it. Suppliers love long term contracts and orders. Buyers usually don’t go for them because it’s cheaper on average not to. No one signed a contract. The market price moved. That’s capitalism. If you want to prevent it happening again you can pay for it, by having a long term contract, stockpiling or multiple suppliers. Many, many mar…

> Who is participating in bad faith? Hoarders. (In this instance, as per upstream comments, possibly coordinated, possibly by CCP.) There’re fairly illustrative examples in my comment that you are ignoring. You should address them if you object to that idea. > If you want long term security of supply pay for it. A single player cannot outbid a state and/or many players acting maliciously in accord.

Maybe a more relatable example: ticket scalping. These are people who rush to buy tickets to popular concerts or sports events and then resell them at the venue priced much higher than sale value (and a lot higher than the venue/artist/sports team got). Are these people adding value in any way or form or just leeching?

Re: The semiconductor bust – still coming?

#64

Earlier quoted context omitted.

Switches and raspberries got really hard to find for a while too

Raspberry Pis are still widely out of stock. And not just Pis! The car industry has been facing chip shortages, Fujifilm's flagship digital cameras and lenses are still impossible to find and have been for years since the chip shortage started... Two years after Nvidia released their flagship GPUs they're finally able to be purchased, but they're only six months away from releasing the next generation, which will als…

Even for Apple stuff, it takes 4+ weeks to order anything outside of base models and a few other configurations. Normally It’s about 1 week.

Re: The semiconductor bust – still coming?

#65

Earlier quoted context omitted.

> Who is participating in bad faith? Hoarders. (In this instance, as per upstream comments, possibly coordinated, possibly by CCP.) There’re fairly illustrative examples in my comment that you are ignoring. You should address them if you object to that idea. > If you want long term security of supply pay for it. A single player cannot outbid a state and/or many players acting maliciously in accord.

Hoarding is not in "bad faith". If people anticipate a shortage in the future, storing some of the existing supply is socially beneficial. It raises current prices less than it lowers future ones.

The problem being discussed is when it wipes the supply, isn’t it?

Re: The semiconductor bust – still coming?

#66
post #46

Earlier quoted context omitted.

I'm not trying to advocate a pro-China stance here, but is there any business that is done in China that isn't in part sponsored by the government? One book I read even told the story of a grey market manufacturing operation that had to go legit not because of a government "bust" but because the local figurehead wanted the output of the factory included in his productivity figures.

Of course there is - I've gone to Shenzhen several times and hired people to build my kickstarter - including supplying of most of the parts, building the PCBs, assembling the boards etc - the only time I ever interacted with the government was the guy behind the immigration desk on the way in and out again (I have an APEC business card so I don't need to apply for visas, just show up at the door and show it)

I wasn't implying the Chinese government actually directly runs the business. Instead the business has a contact in the government who at times may make decisions for the business. In the case of Chinese businessmen visiting the US it literally took the form of a CCP member who shadowed them at all times, participating in all meetings and potentially censoring communications.

They never actually stepped in and did anything, presumably because the businessmen aren't dumb enough to say something like "what can you tell me about Tiananmen Square?"

Re: The semiconductor bust – still coming?

#68
post #22
post #13

Earlier quoted context omitted.

I consider this economic warfare. I suspect disruption and chaos are the end goal, and the monetary gain is just the cherry on top. Once the supply chain starts to catch up, they will dump everything at once and possibly ruin the manufacturers if they aren't careful. And careful mfgs means supply disruptions will be extended even longer. I mostly posted because it doesn't seem like anyone is even noticing the coup. T…

> “I mostly posted because it doesn't seem like anyone is even noticing…” What do you read / use to stay current on what you are referring to as a coup? Are you noticing these trends / events because you are a frequent buyer of related technologies or is it out of interest or maybe some other reason? Do you think more manufacturing will be moving to the states or out of Asia in general as a result of what’s happening…

I'm a design engineer. I've been crossing components almost every day for over a year now. There is no end in sight and it is getting worse. Companies will start to fail in 2023 if they can't ship product. And then the ripple effect will start.

I doubt wafer manufacturing is coming back to the States in less than 10 years.

I'm not sure the best way to make laws, but something will be done when the damage is clear. Chinese sanctions wouldn't surprise me.

Re: The semiconductor bust – still coming?

#69
post #11

Earlier quoted context omitted.

> “…strategically bought…” That’s a key phrase here. Did others do the same? If they didn’t then why not? And if they did, then they are surely profiting from the shortage as well. I wonder how long this can go on for… I mean if you can make less of something, and make 2x or more (in your example) profit from it then I’d guess some businesses might not want this kind of arrangement to end.

Did others do the same? Yes, and that's part of the problem. It's not just hoarding by Chinese brokers. There is a massive, unknown, and unknowable shadow inventory of parts sitting on shelves at EMS facilities right now, far in excess of near-term production demands... just as many/most of us had a garage full of toilet paper last year that we are only now working our way through. Our factory rep says it's getting h…

The majority of the buildup is because 99% of the parts are in, but production cannot start until the missing few parts appear.

That's the genius of the bottleneck hording approach. But you still need ~$100 Billion dollars to pull it off, which mean only a few players in the world could do it.

The problem comes from sitting on that 99% inventory for so long that the bills come due. Normally you get 45-60 days to pay and you sell your product by then so your cashflow is fine. But now, you need to get loans, interest rates are rising, etc. There is a finite time that companies can hold on before it all comes crashing down.

Re: The semiconductor bust – still coming?

#70
post #12

I think it is rather easy to say that the rapid buildout of more and more fabrication plants may be fruitless if the entire industry is still going to rely exclusively on ASML for EUV and even more so if there are no new ingot providers. Apparently, the fab companies know something we do not in those regards. Beyond that, while we are facing high demand now and low supply, a ton of new fabs will increase supply and……

The problem is in much more mundane ICs than cutting EUV stuff. Power regulators, converters, microcontrollers from a year back or 10 years back, everything.
Post reply on HN