2019 will see DRAM oversupply say industry sources

[CAT-THE-FIFTH]

Question - back in the day cheapo RAM used to be of poor quality and you'd end up with crashes and so on purely due to poor RAM. These days you don't get RAM that actually isn't fit for purpose.

So the question is are we going to see the market flooded with cheap Chinese RAM that supposedly matches the spec for DDR4 (for example) but which has lax tolerances and is prone to single bit errors which cause crashes? Is that a possibility with modern fabs / OS's?

The well known companies repackaging the RAM to sell here will probably buy the non-reject stuff and use it,but I expect a lot of production will be geared towards finished products like phones,etc. The only reason China starting prodding the RAM companies was because their own phone companies,etc were having issues.

Also remember things like export restrictions - if Chinese consumer products use parts from other countries,they are also covered by any export restrictions. If you end up using entirely Chinese designed and made components,you have far less issues there.

However,if you were to buy RAM from a Chinese webstore just be careful,just like some of the SSDs you can import! The warranty,and well reliability,might be rather lacking!

[pp05]

Isn't part of the reason that the companies focused on other more profitable areas such as smartphones. This meant a shortage for desktop memory which saw price rises.

[CAT-THE-FIFTH]

Isn't part of the reason that the companies focused on other more profitable areas such as smartphones. This meant a shortage for desktop memory which saw price rises.

It was the smartphone companies which complained to regulators since they increased prices at a rate not seen since 1978!

All the excuses in the media,was PR bumpf from the DRAM companies - they were screwing everyone over. What is the likelihood they told OEMs and regulators,the DRAM price increases for smartphones was since they didn't have enough capacity due to PCs?? They then tell the computer industry price increases are down to smartphones! With both bases covered - PROFIT!!

[kompukare]

Surely with only three players left, there is absolutely no reason for them to conspire to raise prices.
Especially since Samsung have nearly 50% of that:

Now, it is possible that in some gold course in Korea Samsung and Hynix have informal discussion, but since only Samsung has 'spare' capacity there is no need for that. And Samsung's 'spare' is AFAIK those fabs where they can make flash or DRAM so there is no point them oversupplying the DRAM market if that means loosing flash marketshare.

The far more interesting graph would be one which showed the DRAM profits over the last 20 years. I suspect it would show looses for probably 70% or more of those years. Which is why there is now so few players. I suspect it wouldn't look dissimilar to something like this:

although even that doesn't really show all the other vendors went busts. I'm sure Samsung and Hynix play this like a war game with any eventual profits being only after years/decades of losses which is why they're now raking it in.

As for consumers, well I guess I'll just not buy at this price while my current stuff is fine.

[CAT-THE-FIFTH]

Meh,as usual the internet always find some way to justify price increases in tech - I don't see them defending dairy farmers in this country who are not in a good state or asking we pay more council tax or general taxes so we can have better services?? Maybe we should ask for VAT increases and tax increases to cover the fact the country is severely indebted?? Our debt is at 88.3% of the GDP. If you want to pay more then pay more taxes. Yeah,that would go down like a lead balloon!

It was the same as the forum people saying nodes were getting more expensive so Nvidia was justified in raising prices and having more segmentation. Poor Nvidia.

Same arguments for Intel and it's segmentation since fabs cost more.

What about AMD - they made losses too,so should they suddenly jack up prices of their CPUs?

Every consumer Vega card sold is at a loss - yet people were moaning when the subsidy was cut and gamers refused to buy them. When the HD7970 was launched people on this very forum said it was too expensive - the same with Fury X and it's expensive HBM.

Maybe people should look at AMD financials for the last 20 years.



In the end nobody outside investors, employees or subset of computer geeks are that worried about the DRAM companies.

DRAM companies need to know how to a reasonable price to OEMs instead of having cycles were prices go up 3 times or more.

People on forums forget we are not paying only increases due to DRAM companies, but everybody upto the retailer who will take advantage of the increases if they can.

In the end unfortunately for them it was their own customers like phone companies which complained as they were absorbing costs.

Yet look at the actual amounts THEY are paying - less than us as PC builders who are paying EVEN more.

[Saracen]

They force a price drop, until a company or two are forced out of the market, and then the price goes up again. It has been going in cycles like that for as long as there has been dram, which is why Intel who were originally a dram company got out of the game. If the profits were that good, they would re-enter the market.

I find it hard to judge the dram companies for following the usual supply and demand rules when demand is outstripping supply.

Hmmm.

Bear in mind that the "usual" supply and demand rules do not provide oligopolies with a free hand to collude. That's why we (most countries) have antitrust laws.

As for whether this is price-fixing, or simply companies responding individually to similar input and production costs, that remains to be seen. It's not price-fixing if s3veral individual companies decline to initiate a price war that all know would result in a race to the bottom in which all lose out. It is, however, if they collude to agree not to do so.

Ironically, neo-classical marjet pricing theory tells us that whenever there is price-fixing, resulting in unfairly high margjns, the incentive to 'cheat' for short-term gain increases.

DRAM companies have certainly been caught price-fixing in the past. Whether they are currently, or are just declining a price war, remains to be seen. I've certainly seen no hard evidence of a cartel - merely self-serving price maintenance ehich is legal provided there is no edplicit collusion involved.

And either way, a significant new entrant, like China, is going to upset the apple cart, at least in the short-term.

[CAT-THE-FIFTH]

Also to people who justify the increases - if large corporations have sent regulators after the DRAM makers multiple times for smaller increases,why should I as a consumer end up having to bare even WORSE increases??

And either way, a significant new entrant, like China, is going to upset the apple cart, at least in the short-term.

Only because recently they whacked the prices up massively and as one research firm said,last year was the BIGGEST increase for 40 years. Hence it has made it viable for China to make its own RAM especially when even doubling the price of RAM(or even a 50% increase) is significant when you import $89 billion of the stuff and are their biggest customer. That is literally meaning $60 billion of RAM one year before costs $90 billlion now,ie,Chinese OEMs have had to mostly absorb billions,since most consumers moaned even when Dell bumped up prices by 10% in 2016.

So then think if the prices further increase,ie,50% another year,then $60 billion of RAM becomes $135 billion within two years or $75 billion extra over two years. It makes China think,wait a second? Maybe we should just start making our own RAM,instead of giving away so much money to foreign countries and for our local firms to absorb those costs.

Then like a bunch of foolish people,when the Chinese governement asked them to try and stop increasing it so much,they ignored them which not only means China has a good chance of fining them (this is China after all),but will have accelerated their own plans too.

The DRAM companies should know better - if they had targeted a smaller price increase(as I said for years),they would have increased margins,and also made Chinese plans less viable.

Heck even have a reasonable minimum price level which could be agreed with their customers.

People can tolerate price increases,but not ridiculously massive ones in short periods.

This happens so many times in the tech world - companies don't engage some commonsense and then make it viable for others to usurp them.

Outside a few people,most DIY PC builders are annoyed at the price increases,but only because RAM which cost £60 one year has cost £180 or even £200 this year. If the price had jumped to £90 or even £120,there might have been still moaning but I think people would have quietly accepted it.

I have known plenty of people not bother doing builds,simply because they might as well wait until prices drop,and you can also see it a number of OEM desktops,where they have dropped the speed of the bundled RAM used,quantity of RAM and even gone to a single stick to keep costs in check.

It might even explain why graphics card prices after the mining boom are still not as low as two years ago even with a stronger pound,ie,the RAM must now cost more,so even more strain on the pockets of people wanting to build PCs!

[Saracen]

All of that justifies why people are upset at high prices, but none of it, at least in isolation, is sufficient evidence either of cartel or price-fixing.

For instance IF prices went up because a limited number of suppliers are abusing oligopoly power in an overt agreement, we have price fixing. But IF, on the other hand, prices went up because :-

a) demand shot up suddenly due to, say, bitcoin mining denanding boards, AND

b) production cannot be increased sufficiently in the short-term to satisfy demand, OR
c) production could be increased but each oligopoly supplier decides independently that, for now, their interests are best served by not increasing production ....

then .... it isn't price-fixing, On the other hand, if those companies explicitly colluded either to maintain high prices, or to restrict production, then it is. Well, it is unless either the oligopoly players are not subject to price-fixing laws (like nation-states acting under international treaty) or companies acting within a statutory exemption (which the DRAM market is not).

That is to say, high prices are an objective of price-fixing but high prices, by themselves, are not sufficient to establish that price-fixing has occurred. It's not sbout what prices are, but why the prices are ehat they are.

[aidanjt]

All of that justifies why people are upset at high prices, but none of it, at least in isolation, is sufficient evidence either of cartel or price-fixing.

...

While that's true, it shouldn't surprise the DRAM manufacturers one iota, given their past guilty pleas for price fixing, that when the price of DRAM literally doubles within a year, major accusations of, and investigations into price fixing can't be far behind.

[CAT-THE-FIFTH]

All of that justifies why people are upset at high prices, but none of it, at least in isolation, is sufficient evidence either of cartel or price-fixing.

For instance IF prices went up because a limited number of suppliers are abusing oligopoly power in an overt agreement, we have price fixing. But IF, on the other hand, prices went up because :-

a) demand shot up suddenly due to, say, bitcoin mining denanding boards, AND

b) production cannot be increased sufficiently in the short-term to satisfy demand, OR
c) production could be increased but each oligopoly supplier decides independently that, for now, their interests are best served by not increasing production ....

then .... it isn't price-fixing, On the other hand, if those companies explicitly colluded either to maintain high prices, or to restrict production, then it is. Well, it is unless either the oligopoly players are not subject to price-fixing laws (like nation-states acting under international treaty) or companies acting within a statutory exemption (which the DRAM market is not).

That is to say, high prices are an objective of price-fixing but high prices, by themselves, are not sufficient to establish that price-fixing has occurred. It's not sbout what prices are, but why the prices are ehat they are.

The fact is they can make up any reasons they want but as shown by history if you jack up prices,it invites alternatives and in the end THEY are setting the prices,and sadly for them they got caught fiddling stuff before.

The price increases are unusual - not seen for 40 years,and in the end the RAM makers only have themselves to blame if regulators or companies go after them,and the issue is that they are making the prices increases so blatant it invites it happening. If they had been more cunning about the price increases,ie,just not made it so massive,they probably would have got away with it too,and most likely could have mantained a better price for longer.

This is why they have gotten accused of price fixing,since when it actually happened last time it was obvious massive price increases,and hence why I cannot believe they could not have seen the same thing being levelled against them this time.

Now think of this in the context of all the trade wars we are having now.

IF China plonks a fine on them,then expect the US to deffo try it if Samsung is involved,even if it is to protect Micron.

Micron has taken Chinese RAM makers to court by accusing them of stealing stuff late last year,through collusion with the Taiwanese UMC,and then the Taiwanese/Chinese companies took them to court and now Micron cannot sell any RAM in China. So the chances of China fining them might be very big(unless they give a concession like dropping the prices,which China apparently asked them to do months before),even as a political statement especially since companies like Micron are American(plus look at what happened to ZTE). I said China will not play nice,which is no different from any of these large countries.

The issue is they are adding legitimacy to what China is doing(even though it is to protect their interests) by having massive price increases.....for whatever reason.

Plus they might get more support from the rest of the world too,but when the rest of the world also suffers from RAM prices,I am not sure many are that enthused to get involved,since they only care about their own bottom line too.

While that's true, it shouldn't surprise the DRAM manufacturers one iota, given their past guilty pleas for price fixing, that when the price of DRAM literally doubles within a year, major accusations of, and investigations into price fixing can't be far behind.

Yep,and if one regulator(even if its Chinese) does it you can see what might happen after that.

[Corky34]

All of that justifies why people are upset at high prices, but none of it, at least in isolation, is sufficient evidence either of cartel or price-fixing.

Not disagreeing just throwing this out there, there's a class action lawsuit going on regarding price fixing so we'll probably know how sound the evidence is eventually.

The short version is apparently there's clear increases in the price of DRAM shortly after statements to investors and the industry on their business plans, iirc those business plans were along the lines of we're not targeting increases in market share, we're not increasing supply, etc, etc. The sort of thing that could be interpreted as we're not changing the supply side despite increases in demand.

A long version in a PDF is here if you like the idea of watching paint dry.

[CAT-THE-FIFTH]

Not disagreeing just throwing this out there, there's a class action lawsuit going on regarding price fixing so we'll probably know how sound the evidence is eventually.

The short version is apparently there's clear increases in the price of DRAM shortly after statements to investors and the industry on their business plans, iirc those business plans were along the lines of we're not targeting increases in market share, we're not increasing supply, etc, etc. The sort of thing that could be interpreted as we're not changing the supply side despite increases in demand.

A long version in a PDF is here if you like the idea of watching paint dry.

If that holds up in court it would mean the US regulators would have to get involved too.

Maybe they will drop prices soon and say supply is back to normal!

[Saracen]

While that's true, it shouldn't surprise the DRAM manufacturers one iota, given their past guilty pleas for price fixing, that when the price of DRAM literally doubles within a year, major accusations of, and investigations into price fixing can't be far behind.

Oh, indeed. I'm not surprised at allegations, and for the record I'm emphatically NOT saying I don't think it's price-fixing. Maybe it is.

My point was that price-fixing isn't the only possible and plausible reason why prices have been high. One very plausible reason is that with BC mining demand causing temporary shortages, each supplier decided on their own to not increase production, thereby keeping prices high provided nobody else ramped up production.

Each company, utterly independently, can make that decision and it isn't price fixing. If they collude about it, it is.

With a very limited number of suppliers, it's entirely plausible they each decided to maintain production, raise prices and watch what the others did in the marketplace. One very plausible reaction is that someone breaks cover and cuts prices to go for market share. It's basic Monte Carlo game theory in action.
Prisoner's Dilemma.

[Saracen]

Not disagreeing just throwing this out there, there's a class action lawsuit going on regarding price fixing so we'll probably know how sound the evidence is eventually.

The short version is apparently there's clear increases in the price of DRAM shortly after statements to investors and the industry on their business plans, iirc those business plans were along the lines of we're not targeting increases in market share, we're not increasing supply, etc, etc. The sort of thing that could be interpreted as we're not changing the supply side despite increases in demand.

A long version in a PDF is here if you like the idea of watching paint dry.

Agreed. But those statements are still consistent with either price-fixing or not.

Also, re the lawsuit, that's clear evidence someone thinks they can establish collusion, but it isn't evidence of collusion until the case is decided.

It's also not the first time. History gives us an example. BUT .... that it was done before isn't evidence of current guilt any more than accusations are.

As a young and somewhat naive man, I made the mistake of suggesting something similar about a recidivist burglar at a dinner party. This somewhat irascible old dude treated me to a lecture about the difference betwern human psychology and legal guilt, and precisely, and at length, why even a lengthy string of previous burglary convictions is NOT evidence of guilt in a specific new burglary and why, except in very rare and specific circumstsnces, such a record wouldn't be admissible at trial.

The old dude? Oh yeah. One Lord Denning, at the time Master of the Rolls, law lord and the country's most senior judge. I guess he knew what he was talking about.

[tweake628]

I keep waiting till prices drop to build my desktop .but for now my laptop is more then I need to play games

[Saracen]

The fact is they can make up any reasons they want but as shown by history if you jack up prices,it invites alternatives and in the end THEY are setting the prices,and sadly for them they got caught fiddling stuff before.

The price increases are unusual - not seen for 40 years,and in the end the RAM makers only have themselves to blame if regulators or companies go after them,and the issue is that they are making the prices increases so blatant it invites it happening. If they had been more cunning about the price increases,ie,just not made it so massive,they probably would have got away with it too,and most likely could have mantained a better price for longer.

This is why they have gotten accused of price fixing,since when it actually happened last time it was obvious massive price increases,and hence why I cannot believe they could not have seen the same thing being levelled against them this time.

Now think of this in the context of all the trade wars we are having now.

IF China plonks a fine on them,then expect the US to deffo try it if Samsung is involved,even if it is to protect Micron.

Micron has taken Chinese RAM makers to court by accusing them of stealing stuff late last year,through collusion with the Taiwanese UMC,and then the Taiwanese/Chinese companies took them to court and now Micron cannot sell any RAM in China. So the chances of China fining them might be very big(unless they give a concession like dropping the prices,which China apparently asked them to do months before),even as a political statement especially since companies like Micron are American(plus look at what happened to ZTE). I said China will not play nice,which is no different from any of these large countries.

The issue is they are adding legitimacy to what China is doing(even though it is to protect their interests) by having massive price increases.....for whatever reason.

Plus they might get more support from the rest of the world too,but when the rest of the world also suffers from RAM prices,I am not sure many are that enthused to get involved,since they only care about their own bottom line too.




Yep,and if one regulator(even if its Chinese) does it you can see what might happen after that.

Again, if a company reads current market forces and trends and sets it's prices to mximise returns, that isn't price-fixing. If it colludes with others that they all do the same, it is.

The price rise being unusual, or huge, isn't proof of ptice fixing, especially if some unusual external factor say, huge demand from BC miners (for which there is no historical precedent) causes an unusual demand pattern.

Taking advantage of high denand is not price-fixing. Nor is limiting or even reducing production to raise prices because doing so opens the door to competitors to jump in a seize market share by doing so.

If the competitors ALL adopt that same strategy, independently, it still isn't price-fixing.

It's a basic economic law that high demand and limited supply causes prices to rise until either demand dtops off of supply increases .... perhaps by a new entrant.

The basic definition of an oligopolist is a player with a large enough market share to influence market conditions without being able to, unlike a monopoly, dictate them. That's why there are "antitrust" laws.

Being large enough to influence price by your actions and seeking to maximise profit isn't price-fixing, nor is actually doing so. Colluding with others to do so is, bar those exceptions I mentiond.