According to chip makers, the plunge isn’t over yet. Now hoping it won’t turn into the mess as in 2001 when the last tech bubble became the dotcom bust.
Global semiconductor sales dropped 15.5% in the first quarter, from the fourth quarter last year, to $96.8 billion, the World Semiconductor Trade Statistics (WSTS) organization reported Monday afternoon. The three-month moving average in March has now plunged 25% from the three-month moving average at the peak last October, the deepest plunge since the Financial Crisis:
The explosive growth in semiconductor sales that started in 2017, with a final spike last fall, has created an inventory glut that is now being brutally unwound in the face of soft demand. Sales in March, on a three-month moving average basis, were back where they’d first had been in April 2017.
The 25% plunge from October, in percentage terms, is still less than the plunge from peak to trough during the Financial Crisis (-39%) and the infernal 11-month long plunge during the dotcom bust that bottomed out in September 2001 (-45%). But given the recent guidance from chip makers, the trough of the sales slump in this cycle may still be in the future.
Demand was down across all regional markets: The global 15.5% drop in sales from Q4 2018 to Q1 2019 split up regionally this way – and the problem isn’t just China:
- Americas: -29.2%
- China: -14.5%
- Japan: -13.8%
- Asia Pacific/All Other: -10.4%
- Europe: -3.1%
There are numerous reasons for this mess, not just one.
It is likely that efforts to front-run potential tariffs have contributed to the blistering surge in sales since early 2017 and to the subsequent inventory glut.
The global stagnation in smartphone sales last year, including the drop in iPhone sales, when sales were supposed to rise further, didn’t help.
Data centers that power the “cloud” where supposed to be an endless-growth business for chip makers where not even the sky was the limit. But Intel has been warning since January that sales of chips for data centers started slowing in the fourth quarter. Last week, Intel reported that quarterly revenues from its data-centric business fell 5% year-over-year.
The collapse in demand in 2018 from the crypto-mining sector triggered sales swoons in semiconductor makers, such as Nvidia, that specialize in those chips.
There are other factors. It’s not just one thing that is causing this plunge in chip sales, but a series of events – on top of an inventory glut.
And it doesn’t appear to have bottomed out just yet. Semiconductor makers have lowered their guidance for the next couple of quarters, including Intel, whose CEO Robert Swan said last week in the earnings report, “Looking ahead, we’re taking a more cautious view of the year, although we expect market conditions to improve in the second half.”
For now, everyone in the sector is hoping that the chip glut will be worked through by the second half, and that demand will pick up again in this infamously cyclical sector, and that it doesn’t turn into the long-drawn-out 45% collapse of chip sales experienced after the implosion of the last tech bubble that led to the dotcom bust.
The broad inventory pile-up in the US, a result of a six-quarter surge, shows first signs that some of it has started to unwind. Read… How the Inventory Pileup Boosted Q1 Blowout GDP and What Carmageddon Has to Do with It
Enjoy reading WOLF STREET and want to support it? You can donate. I appreciate it immensely. Click on the beer and iced-tea mug to find out how:
Would you like to be notified via email when WOLF STREET publishes a new article? Sign up here.
The retail apocalypse is also contributing to the dot com bust. Most of us don’t notice all the computer equipment used in retail stores, cash registers are really computer terminals, scanners, scales, printers, and cameras, all electronic equipment. That’s a big part of the market that is disappearing.
not sure that’s true… there is a lot of automation that goes into distribution even if there are no brick and mortar stores…it could be a wash or even a boom.
Apparently it is not a boom. At least not according to this article.
It may have been once, but like with all booms, the demand by the growth of the online sweatshop-based industries has apparently peaked as that growth has been fulfilled. By that I mean that the online companies may have been buying chip-based equipment while building their businesses, but like with all things, there is a limit to that growth and now they have met it. Amazon doesn’t need to build a second Amazon after building the first.
Wait, Amazon built itself?
Excellent addition to Wolf’s points! Seems to me that smartphone sales in general have peaked and the only viable, under-delivered market share left is kids under the age of 10. Not many chips are needed in them satellites that deliver internet to places where wifi hasn’t developed, I suppose?
Hey Wolf–no weekly podcast? Really enjoyed listening to that! Man, we got some macroeconomic changes happening RIGHT NOW…
The cloud thingy is also starting to level off in a big way. It is great if you are a cloud-based software provider. You can use AWS or Azure to quickly scale your business. But that business has quickly matured.
The cloud is still way more expensive than just keeping your stuff in-house as long as you have a well-managed IT department. And moving your servers to the cloud does not suddenly give you a well-managed IT department.
So between the costs and maturity of enterprise software systems, I expect much slower growth coming out of cloud-based systems in the future.
Well the data center buildout in Ashburn VA is still crazy. The shear size of buildings is nuts, and how many are going up. Facebook and AWS are the big ones. AWS has over 100 datacenters here.
AWS got a $10 billion DOD contract.
But they aren’t buying more chips.
“The cloud is still way more expensive than just keeping your stuff in-house as long as you have a well-managed IT department.“
Really? I would think cost is a big part of the value proposition for hosting services. That and embedded Disaster Recovery and encryption measures.
I goes through this trade study several times a year.
If your services don’t change much (up or down), and you don’t have high availability requirements, then an internal IT infrastructure can probably be cheaper.
But if you have quickly evolving and growing services and require high availability (i.e., multiple datacenters with wide geographic dispersion) then AWS, Azure, and GCP are extremely attractive.
Agree. People who, fox example, have big demand around the time of the Super Bowl can benefit from cloud services.
People who have Christmas-related demand probably also can benefit from the cloud.
Most business users, not so much.
The problem is that you are going to need good tech people no matter what. The exact location of the servers is kind of secondary.
There is a couple of things that do disturb me regarding the cloud. The main thing is what might happen in case of a major international crisis turning hot, especially outside the US or what if a major provider of the cloud would find itself in, let’s say, less fortunate economical circumstanses. A major disturbance of the services of a cloud provider would disrupt payment systems etc and in todays world the disruption of the use of non-cash based payment methods, well, the longer the duration, the more interesting things would turn out to be …
Realist,
No one on this Forum would be able to answer your questions!
But here are some points for you to chew over;
– if a crises turns “ hot” , things like payment systems will be the last thing the Nations involved would care about.
– every Cloud service provider that is worth their salt have convinced or have to convince their clients that they have a backup option in case their ( data) goes belly up!
– for any company or government department or ministry to be wholly reliant and convinced that an outside party will provide them with absolute security and longevity for data storage is ( either insane or fatally flawed misconception that doesn’t bear any realistic understanding of their role to ( their clients/ citizens in the case of gov departments).
Electronic means of storing, transmitting and exchanging data are only as good as the physical components that ( house them) .
Buildings , super servers, super computers, and to the minuscule components that these are made of and wired up together are all subject to the norms of the physical world and thereby affected by the same.
A Nuclear war, Electronic warfare or your conventional war would to varying degrees take care of those depending on how these are protected.
The only savior that might give you some peace of mind my friend is if ( Jesus-Musk) works out that Tesla Rocket and moves our ( Human data) to the other side of the galaxy, only then! :)
Cheers
The cloud is still way more expensive than just keeping your stuff in-house as long as you have a well-managed IT department
Nope. Not by a long shot.
Computing is just doing the same thing as electronics manufacturing did in the 1980 – 1990’s: Everyone has their own little factory, but, everyone has too little volume to properly use the capabilities of the new machines and yet everyone has to buy them to remain competitive. The ones who will not be wiped out are the ones that move production to outsourced assembly lines who have the volume and the experience to run it 24/7!
Today, one would be considered insane to have invested in more that a prototype assembly line, which is about the size of about four desks. The ‘production money’ saved (or some of it) should be reinvested in Q&A!
Same with software, there are a few local servers to test stuff on, maybe even a local Git-repository for redundancy, while all production goes out to the real factory, the one that knows what it is doing and can afford the very best techies to keep it running. The cloud will take over, it is how it goes.
Self hosting is cheaper. Here’s a real world example, https://www.servethehome.com/falling-from-the-sky-part-4-leaving-the-cloud-5-years-later/
StackExchange is also self hosted for the same reason. The cloud is expensive.
Strange as Oracle is reporting that they get a return of 3-4x the cost of self-hosting when people move to their cloud.
Not saying the cloud won’t eventually take over. It just won’t do so at the existing price points. Servers and redundancy are ridiculously cheap these days. My company has had zero downtime from server and network outages in 6 years.
The same level of service on AWS and Azure would be triple the cost of our standard equipment amortization schedule + labor, while still providing poorer availability and performance.
We do use the cloud extensively, but in the opposite way you propose. Our test assembly lines are all created in the cloud. That’s an excellent use case.
If that’s the case, then I can only conclude you either have a strange use case, or you don’t take the cost of personnel, security, redundancy, etc. into account. Every project I price out including all costs, ends up with cloud cheaper and safer
How could it? We have six large database applications for different divisions and standard websites and email. Exchange is run on O365.
All done over Hyper-V VMs on a geographically dispersed pseudo-HCI (Microsoft’s) with triple deep failover. With standard, but very rigorous attention to security. I’ve got 3 sys admins who moonlight as data analysts because the system just works so they can really spend their time devoted to management reporting.
Server replacement schedule is every 6 years and network equipment is every 8. Our biggest cost is software maintenance on the various applications, which I get regardless of hosting site.
I can’t believe this is an unusual use case.
Let’s all please observe a moment of silence for the trillions of Yellen Bux that are going to perish in the implosion of Tech Bubble 2.0.
Hey Apple, people are keeping their Iphones for longer and sales are slowing down.
Apple: I am gonna release like three Iphones a year, that will make things better!
Yeah… I think not.
Apple: Shut up! I know best!
Apple: We are jumping into mature, low margins business so SC*EW YOU!
Ha! When they issue a dividend I’ll buy the mature part. Greedy buggers ;-)
Apple has a huge China problem which will make future high ROI very unlikely. It’s a safe bet that no company can sustain a high ROI indefinitely, and AAPL is no exception, to which you are allowed to say “duh!” because you know that unusually fortunate conditions are usually temporary.
Spoiler alert. It does sit on a huge pile of cash, though, so it will be just fine for the duration, even if its investors will be unhappy. Don’t be surprised when it ramps up its stock buy-back program to compensate for its tax liability on repatriated profits.
I have no financial exposure to this sector, so any deep-seated concerns about me that have been keeping you up at night are unfounded. You can rest easy. I’ll be okay, really.
Meanwhile, back in the real world, APL stock is up 40%+ in the last 5 months (PE 16.6)…
APL = AAPL
Yes, it has been a nice investment (entered in early January). It will be even nicer if/when they use dividends instead of buybacks to return cash.
Just double the price. Easy.
The Semiconductor slowdown is to be expected. Worldwide sales have gone from $315B in 2013 to $476B in 2018. That is almost 9% a year over that time frame.
Worldwide Smart phone sales were down 5% in 2018. Again, to be expected. Blistering paces cannot continue indefinitely.
I’m surprised the decline is not greater. Smart phones are expensive and a lot of them are “owned” (used by) kids. There have been no recent dramatic improvements in functionality to cause people to “upgrade”. I also wonder if the very largest new phones are selling. They appear very cumbersome to carry around.
The smartphone number i got was from a recent IDC report. The Semi number was from a recent Gartner report.
Apple has followed the same pattern as other emerging techs.
In the early days, the stuff is a bright new idea.
Then come a series of upgrades that add new and desireable features.
(Apple has their own version of this where they watch the rest of the tech world advance, and then slowly trickle the same upgrades to the members of the Apple cult)
But eventually, it is no longer an emerging tech but now a mature one. People have the smartphones that do what they want them to do. There isn’t really a killer new feature that can be added on anymore. At some point, a television does what people want a television to do, so the “upgrade to get the new feature” drive fades. There might be at times a new improvement to be rolled in, like Std-Def to Hi-Def TVs. But those are rarer. For example, when 5G is truly rolled out, people will likely want the faster speeds. But Apple and Smartphones are beyond the upgrade every six months to get the new stuff part of an emerging tech. Now they are more like Tv’s.
> They appear very cumbersome to carry around.
Yeah I don’t understand this either. Every iphone since the iphone 4 looks like a ClownPhone to me.
Do you have cool apps on your ClownPhone?
That bubble should be from increased production for crypto mining. The timeline certainly goes well with that. Also, the drop puts us well back into the previous channel, so it’s not the apocalypse, yet.
=> it’s not the apocalypse, yet.
They want it to be a surprise. If people get tipped off and pile on nobody will make any money.
High profitability resulting from correctly anticipating the apocalypse is likely to be very temporary. Still, shorting everything in sight just before the collapse of civilization is technically an excellent strategy. Technically.
I’ll finish it off for you:
The problem with betting on the end of the world is that if you’re right, you don’t get to collect.
When I look up parts on DigiKey sometimes I see a note about tariffs. Kind of interesting….this note is on a large number of parts.
Our Valued Customers,
The tariffs announced by the Office of the United States Trade Representative (USTR) under the Section 301 Action went into effect on July 6th, 2018 with additional products added to the list on August 23rd. All items identified, as classified, under the Harmonized Tariff Schedule of the United States (HTSUS) with a country of origin of China are subject to a 25% duty upon importation into the United States. See List 1, List 2, and List 3.
Digi-Key continues to work toward minimizing the impact of the tariffs to our customers. We are working with the ECIA (Electronic Component Industry Association) and our 800+ suppliers to minimize impact on the supply chain yet adhere to the law. Due to our inventory position, we have been able to delay the immediate impact of the tariff in some instances.
Digi-Key is responding to customer requests to identify products impacted by this tariff in the search results on our website. Again, the tariff only impacts products with specific HTS codes, with a listed country of origin of China, and consumed within the United States. There is no impact on pricing for products within Digi-Key’s stock shipped outside the United States.
This is a dynamic situation and we will continue monitoring the situation and exhausting all avenues to mitigate any impact or uncertainty in your designs or supply chain.
Any immediate questions not answered in the FAQ can be sent to tariffs@digikey.com
Now Mouser and Digi-Key, those are businesses I understand.
I find it interesting that they’ve been the big two sellers of components for over 30 years. No one’s replaced them. They were the big two in 1985 and they’re the big two now.
Well they’re pretty damn good at what they do. Digikey moved their whole damn warehouse closer to the airport so they could push their next-day air ship time back half an hour.
They were the Amazon of electronics components long before there was an Amazon. And they still are.
I agree, Digikey has been great on anything that I needed for many years.
=> Harmonized Tariff Schedule of the United States
“Harmonized”? Really? Is the US tariff schedule some sort of glee club? How about “coordinated” or “unified”?
Bureaucrats. Last couple of years they’ve just gotten world-weary and given up.
> Last couple of years they’ve just gotten world-weary and given up.
Nice fantasy.
In the hardware universe, I have seen belt tightening since last year. Cancelled Christmas parties, forced vacations, layoffs. This is not going to be pretty.
Executive compensation remains higher than ever, though.
That’s what I’m trained in, electronics hardware.
Don’t go into this, kids.
It’s only through knowing someone, and dealing in surplus electronics often decades, some of it many decades, old, that I have work at all.
Maybe you could look at my transilience thought unifier. Model 14.
I’m looking for some old shortwave and ham sets.
Keep an eye on TSM (Taiwan Semiconductor). It is the canary in the coal mine.
More like the 800 pound gorilla in need of euthanizing.
Dare say there is a “DATA” bubble? That huge USG building in Utah filled with what? Phone conversations tracked with airliners flying circles over us listen to Aunt Bee swap recipes? If there was a tulip moment in all this it might be that everything is data. And that data is worth money.
Worth money indeed: but neither knowledge, nor wisdom.
> That huge USG building in Utah filled with what?
Alien poop.
Don’t forget the bitcoin mining apocalypse. Bitcoin mining centers had fueled explosive sales of Nvidia and other parallel processing chips which was probably also part of that boom. That all collapsed last year.
Also besides the plateauing of cellphone sales worldwide, the reality is that computer CPUs have hit a wall in terms of maximum speed per cpu. Silicon based CPUs can’t switch any faster than 5 GHz, and so max processing speeds per cpu have been stuck for a number of years now. The main advantage for buying new CPUs are mainly in increased energy efficiency not speed.
Hard disk drives wear out before CPUs, but total unit sales have declined steadily since 2011. That’s because of the rapid increase in disk drive capacity, from 1TB to 4 TB then 10TB and now 12TB and more. The capacity increases have come faster than the demand for data storage
The tech crash will be epic. I hope it does not hit LA beach real estate too hard. The bay area will be smoked.
The bay area people always tell us this time it will not crash, but it always does.
Housing will TANK soon! LOL
Entire bay area will sink and SoCalJim is going to pick up houses before the next bubble blows and he will raise the rent to squeeze all the juicy bubble W2 pay checks into his hand.
Here is the thing, this time, the bubble is NOT in silicon, the bubble is in SoftWare. I am NOT sure chip sales drop will pop the bubble. To pop this bubble, you have to let people “VALUE” companies based on profitability. These days, the more money you lose in these software companies , the more you are valued.
A bit confusing this drop as there are chips in absolutely everything these days, from stoves and washing machines to outboard motors…even new chainsaws will self-adjust their performance and it isn’t a mechanical interface with the operator. (screwdriver). My buddy picked up an electronic analyser the other day for $20. Plug any part into it and the display tells you what the hardware is, what each pin is for, and if there is a malfunction in any part of it. $20, on ebay.
A bale of straw at the local feed store costs more than this powerful product.
Does anyone else remember going to the hardware store as a kid and watching their Dad plug a vacuum tube into the tester? This world is crazy and far too complex for meaningful resilience if anything burps desert, and that includes our ones and zeros economy run and influenced by people (persons) who don’t even read. Just sayin’. Be prepared for anything but ‘steady as she goes’. Everything is simply too complex and inter-connected, and there is much unrest in the World.
Sure, but the crappy Chinese capacitors in all of this junk will do them in.
Another big issue is the slowdown of speed improvement that is happening as we reach the limits of Moores Law. Intel only makes the great margins that allow them to pay for its Multi Billion Dollar Fabs (chip factories) when they are producing a cutting edge chip that is an order of magnitude better and faster than the last generation. Intel has been struggling to produce a cost effective 10NM chip at its DX1 fab at the Ronler acres campus. The Dutch UUV ( ultra ultra violet) machines that were supposed to make this happen are 6 years behind schedule and just being installed now, and getting useful chips out is still buggy. Some of the simpler chips from the Taiwan fabs are down to 7NM using UUV. Just recently Intel has broken ground on a new 7NM R&D fab at Ronler to try make up for lost time. But it is not clear that the technical or cost hurdles can be overcome.
What about Samsung? They are there and bailing out IBM from the global foundries fiasco.
More like IBM is bailing out Samsung for not having a research division. Every recent Glofo+Samsung (“common platform”) process has been the custom IBM process minus the cool features. POWER9 was the only GloFo14 chip with eDRAM. Still won’t let foundry customers have that feature, probably never will.
IBM is still in the driver’s seat. But when the car crashes they’ll be as dead as the passengers.
Moore’s Law Ended a while ago.
14 nanometer started shipping in 2014. Wasn’t the next factor-of-two supposed to take 3-4 years??
The problem is tunneling current. At room temperature, you simply can’t get devices smaller than about 10 nm to work. There is too much thermal noise, and lowering the gate voltage only helps up to a point (0.3 V, looks like).
Intel is a legacy company, whose major money-maker was pretty much invented 40 years ago. They are not likely to be a major part of the tech industry, once 64-bit ARM chips start to take over the Server Farms. . .
Full disclosure: I worked on the original Pentium bring-up. The one which ran at 66 MHz, at 5 Volts. Both Alpha silicon and beta silicon.
Samsung develops 5 nanometer chip.
Currently uses 7 nanometer in Samsung smartphone Galaxy S10.
https://www.zdnet.com/article/samsung-develops-euv-5-nanometre-chip-process/
Perhaps this is just marketing hype, or specsmanship.
When companies like TSMC and Samsung failed to get the 12 nm stuff to work, they simply declared that vertical-structure 18 nm is “the same thing as 12 nm”.
In the olden days, describing something as, say, “65 nm” meant “The metal interconnects are 65 nm wide, the features have a 65 nm separation, and the gate oxide is 65×65 nanometers.”
Today, it’s kind of like Alice in Wonderland. You know, “6 nanometers means whatever I say it means, nothing more or less.”
To be fair, modern technology crams a lot of stuff onto a wafer, and there are a lot more layers than there used to be. It’s just that you _really_ need to read the fine print.
Intel really, totally ate dust with its decision to try to do 10 nm with MORE multi-pattern UV rather than going to EUV, which they reserved for their 7nm process. The result- really low yields and a non viable process for mass production. I suspect by the time enough 10nm chips are produced to sell as product that the 7nm node will have matured. Their 10nm process is likely to have a very short lifespan.
In any case, silicon can’t switch faster than 5Ghz, and making smaller transistors makes it even harder to switch at those top speeds. The only gains from die shrinkage are for reducing power consumption
So, the future will require new semiconductor or other types of materials, Gallium arsenide is one known and previously used material, used in the original Cray supercomputers (which are now laughably slower than any top of the line gamer quality desktop PC). Other possibilities are- carbon nanotubes, etc. Not clear if Intel is doing much development in these areas.
Btw, Intel has long claimed that their main competitors TMSC and Samsung use a looser definition of node size such that transistor densities in the Intel 14nm process are equivalent to their competitors 10nm, etc. There does seem to be some truth in that.
We used to make x-ray intereferometers and x-ray spectrographs. The problem with them is that you also had to own a particle accelerator. (1980s)
“Trump calls for Fed rate cut, praises China stimulus…”
Wait to he learns about ECB and Switzerland buying corporate bonds and stocks, and negative interests rates.
Any guess on when Sarah Palin become the next Fed Chairman?
We need more QE, don’cha know.
You betcha!
I’m scared about that very thing……he’s like a grandpa that never went to school and never wanted to but thinks he’s a genius….a greedy, delusional genius. I bet he has already had a talk with the Fed team which is why the market has gone straight up from the crash that was already taking place.
I just can’t believe this is where we are at, these people and ideas should be laughed at like children asking for 100 candy bars, no of course not! I also can’t believe it’s allowed that other countries like the Swiss are allowed to just print money to buy our stocks….in essence we are just giving them ownership in our companies for free and if we start doing it it’s just nationalization and somehow no one understands enough about this to be outraged and just thinks “oh, those smart finance guys have it figured out” when nothing could be further from the truth. The Swiss are essentially nationalizing or just stealing our companies no?……any argument?
You nailed it.
from what I have been reading AAPL makes its growth profits from its services. AAPL has been cannibalizing its hardware products for years. Its solution? 1000.00 phones. But, now realizing it sells more used older model phones. I have a iphone 6… I will muddle along with it for however long before it bricks, then the replacement will not be an AAPL. AAPL can go pound sand. AAPL also has the whiff of desperation in the air, trying to throw anything at the wall to see what sticks, Driver- less cars, credit cards, perhaps flame throwers next? I figured when they built that 1 billion dollar spaceship HQ… that was a tell. As for a cash hoard, MSFT had a cash hoard and went sideways for 13 years… its all about the promise of growth to make a stock move up. As for the SOX index , the weekly 50 ma is severely lagging,suggesting caution to investors.
I just upgraded my 2012 Mac Book Pro — myself. No need for a faster CPU. Newer MBPs have less ports anyhow. Also, no butterfly key board.
I replaced the entire battery and upgraded my sata drive to 1 T. Next to boost CPU performance I replaced all the thermal paste on the CPU — now the intel processor will burst performance because the temps are lower.
Prior to Steve Jobs death, he realized a new temple ( corporate HQ ) needed to be constructed. He thought it should be the shape of a zero because upon his death there would be zero innovation at Apple. :)
I refuse to have any Crapple product in my vicinity. I was given an iPhone 5, and it signals that it’s fully charged by emitting noise that at least covers the FM broadcast band, how is that even legal? No NPR for you! The local “Fix Laptop” place gave me $10 for it.
Come to think of it, there’s an old Bondi blue iMac sitting here in the warehouse, guess I’d better get that listed on Ebay to get it out of here.
You can tell organizations that are in a bubble. They build cathedrals.
Much like cities building the world’s tallest skyscraper, corporations moving into super-glamorous new headquarters often marks the peak before the start of a decline.
While I agree with the sentiment of this article I would caution Wolf’s readers against shorting or otherwise buying puts in anticipation of a move down in semis or the market.
We are in a froth at the mouth, no questions asked mega bull market. I watch company after company miss earinings yet, the pull-back is a day or two then they’re right back to their old pre-earnings price.
The magic trick is that the companies merely say that the second half looks good (no evidence), and it reinforces the notion that China’s recent mega stimulus will boost the US economy as it did back in 2016.
In addition (and I hope Wolf addresses this in future articles), is that I hear the FED will cut interest on bank excess reserves held at the Fed soon (tommorow?).
This will send another flood of mindless money into the markets to boost indicies to all time highs.
You’re long this market or get out. There is no middle ground until we reach the magical second half to discover if it was all a mirage.
And yet it moves:
https://www.bloomberg.com/graphics/startup-barometer/
How much of the sequential decline of 15.5% is pricing? WDC reported today that NAND prices fell 23% in the same period. Same old story: price declines lead to volume gains in highly elastic markets. Units were actually UP in some categories.
When you strip an entire generation of its savings (thru rates lower than taxes and inflation) and transfer that wealth to business owners (who don’t need it) and the working middle class and poor (who don’t save any of it) eventually that donkey will not hunt. The funds end up overseas because of the payments deficit. Trump is screaming for lower rates with a 3.8% unemployment rate. This will not end well. Inequality……what do they think is happening to the small savers because of this party put on by low rates.
Raising taxes on the middle class/wealthy and spending it on infrastructure is a good idea. It will bring rates up and end the party. Those who benefit from the party will be paying for it.
Right on, Mr. Flintstone!
I have $20 in the bank right now. I’ve paid $2200 to the IRS on a gross of $15600, listen to the radio (no TV) and use a several years old laptop and will replace it with a few years old laptop when it bricks. In the last year I’ve replaced my old flip phone with a new flip phone. I’ve also had maybe $100 worth of work done on my bike at the bike shop. I do read a fair number of books, not having any interest in those digital readers.
I am even going to sell my ukuleles (they consume strings) and have sold off my trumpets (consume valve oil and slide grease, and need fairly extensive cleaning occasionally) in favor of the shakuhachi, a Japanese flute that’s about as simple as you can get.
The only way for me to have a tolerable life and an enjoyable retirement in 5-10 years is to, as much as possible in this modern world, adopt the lifestyle of a Buddhist monk, and they’re noted for not being bigtime consumers.
In the US you can be as smart as you like and work as hard as you can, and you won’t get ahead, you’ll only die tired.
This might be a bit short-sighted; but in just more than a year, I purchased 3 LCD TVs with soundbars, Nvidia Shields, PS4 Pros, more than 2 dozen Echos, another 2 dozen Wifi cameras, Routers and Wifi Extenders, etc. etc. And this is only for my home. I also equipped my Dad’s and my sons’ homes.
Considering that even an InstantPot has a semiconductor device, you begin to wonder if the terms life = semiconductor = cloud. There’s no getting away from it. My whole house is on Samsung Smartthings and Amazon Alexa. I don’t even turn a light switch manually anymore. Before you know it, you’ve already spent a fortune.
Two big reasons not to go all totally wired into the IOT fad:
1. Hackers
2. EMPs and other electrical strikes. It doesn’t even have to be a nuclear EMP, a good size solar flare or big lightning strike or transmission grid spike will turn your electronic IOT into fried silicon. Hope you have everything backed up with manual controls and have surge protectors out the wazoo.
I first realized this was a problem in the 1980s when a perfectly good microwave oven stopped working because the motherboard got fried. Now I have even my washer drier plugged into industrial strength surge protectors
I’m on my second GE Monogram microwave and that died, too. My old Samsung portable does the job.
If you’re worried about EMP or something similar, you’ve got bigger problems than light switches. For example, I might freeze to death because my forced air furnace has a motor that can just get fried up. No telling how our hospital can even function.
I live on a top of a hill in the woods far away from the public street. I’m not worried of hackers. I’m just an old man who likes the lights turn on automatically. My pet peeve is having to turn on the lights with a wall switch when my hands are full especially on a stairwell. Home automation is helpful to sroke victims. Does repetitive simple chores that I might forget to do. My son’s mixed pit has learned to trigger the lights.
Apart from the demand reduction mentioned in the article, when the US refused to sell certain chips to China, alarm bells rang.
China has embarked on an ambitious plan to produce most chips locally.
SMIC China’s largest chip producer is enjoying record sales but profits are down as they plow money into R&D.
I doubt American producers will ever recover, and China will dominate the market in the future.
Sinbad, SMIC is 2 to 3 generations behind Samsung, TSMC and Intel. TSMC slapped SMIC with 65 counts of stealing patients and IT and SMIC lost yet they still refuse to honor agreements.
The R&D they are spending is to figure out how to utilize the 1 ASML DUV Photolithography tool they purchased 2018. Photo is the backbone of the chip as it etches the features (to build on) on the wafer. DUV is the latest greatest. Meanwhile TSMC bought an additional 11, Samsung 8 and Intel 3.
These tools take up to 2 years to build. So obviously SMIC is not ramping up.
American fabs make up half of the top 15 fabs. See link: https://anysilicon.com/top-15-semiconductor-sales-leaders-2018f/
Also note Hynix and Samsung have fabs in the US. Two European fabs in the top list have multiple fabs in the US.
There are also a hundred other small fabs in the US who service Military and other particular customers.
If China ramped up today it would take them 15 years to catch up with both capacity and technology.
@ TheDona
\\\\
If you study the technology (referred to as Xnm) vs. performance you will see that since 21nm there was no significant move forward. Some achievements in reducing power vs. performance were made but nothing the customer would mind, given he gets a discount. In other words, they don’t need to be cutting edge to take the market.
\\\
The breakthrough technology everyone is working on right now is EUV, as it would reduce complexity and increase yield significantly. It would be a paradigm changing technology. If the “big ones” (TSMC, Intel, GF, etc.) get this ready before the small ones pick up pace, they are safe. If not, they will eat into their market share and reduce revenue desperately needed to pursue a very expensive EUV R&D phase.
\\\
@Sinbad
\\\
The semiconductor manufacturing is a very integrated process and it would require emense effort to gain foothold in such a demanding, secretive and technologically complex* field.
\\\
* On a 1 to 10 scale it’s a 55 in complexity.
That sounds very plausible, but the sales figures tell a different story.
They call it a cyclical industry, but investors are still surprised to discover…..it’s cyclical! Sales sometimes go down! Then they go up again! Who would have thought it?
I’m still looking for the industry, and the stock, that goes straight up.
I have a warehouse full of old Cobalt 60 irradiation machine guts. Maybe I can set up a chip fab that uses gamma radiation, and get rich.
No mention of AMD. AMD has been on a tear, and seems to be bucking the general trend.