The Hard-Disk Duopoly Is Really Cheap, But For How Long? [View article]
Hard disk drives volumes are growing at 11% CAGR and the SSD "revolution" has done nothing to change that. WDC,STX and Toshiba will ship about 450-500 Exabytes of capacity this year. about 180 of that is notebooks. Total NAND Flash production is maybe 25-30 Exabytes a year at a stretch, and a lot of that cannot go to SSDs because it is going into phones and tablets etc. What's more it still costs ~10x as much as magnetic storage and some manufacturers are losing money at that price. The standard ultrabook architecture within 12 months will be a ~30GB NAND Flash mSATA drive with a 500GB 5mm magnetic drive sitting right next to it.
Notebooks will not convert away from magnetic storage to SSD. It raises the cost enormously and has a capacity penalty while the new architecture matches pure SSD in performance and (eventually) power consumption at a much lower cost. SSDs are shock resistant, but they do have a limit to their useful life that magnetic storage does not have. To my mind reliability is a wash and will not change people's purchasing decisions either way.
Then you have the whole cloud phenomenon which is growing magnetic storage at ~23% CAGR by units shipped (off a low base). One of the trends we are seeing right now is the movement of capacity from the device to the cloud.
SSD has a bright future in enterprise analytics engines and other high-speed database applications. But who sells these SSDs? Sandisk, Seagate and Western Digital (now. Was Hitachi).
The biggest threat to STX and WDC is the potential for collapse in the notebook/ultrabook market in favour of tablets. I expect we will see tablets with magnetic storage eventually, but there will be a dip in the curve because notebooks are such a huge chunk of the business today.
The other threat is, as always, discounting. It will be really interesting to see the first fully-consolidated quarter of WDC numbers next month. Current ASP is around $67-$68/unit and at that level it is very profitable. Before the floods it got as low as $42. If the ASP can be kept above $55-$60 given current softness in the PC market, I would see that as very bullish for WDC and STX.
The Elephant In The Room For Seagate Technology [View article]
If every bit of flash from every fab in the world were made into SSDs, then it would equal about 4% of HDD market share. And except for Samsung's newest fab, those fabs are losing money.
The decline in PC shipments is a threat to the HDD manufacturers. SSDs are not.
Can Hybrids Extend Life Of Seagate And Western Digital? [View article]
The question is really this: Will people be happy with a small about of storage onboard and use the cloud for the rest, or will they want want large amounts of local storage, whether onboard or in home LAN storage?
If the former turns out to be true, then the spinning hard drive will go the way of the dinosaur. (Yes. Spinning platters will still be needed in the cloud, but the revenue from that is nothing like the revenue from noteboks and PCs.) The big problem with SSD is that there is not enough Fab capacity in the world to populate all notebooks with SSD.
If the latter turns out to be true, then hybrids will do well and STX and WDC are pretty-much the only game in town.
I have over a terabyte of photos alone. Solid-state storage will never be the full solution for me, and there is no cloud service that can meet my needs at a reasonable price. Also, in much of the world, cloud storage just isn't fast enough or economic to use as primary storage because of Internet speed and volume caps/pricing.
I converted my main desktop PC's system disk to hybrid a couple of months ago. Bootup speed is only very slightly slower than with an SSD. Application loading is better than a traditional hard drive but not as good as an SSD drive. But I suspect that is because second-gen hybrids still have not quite enough flash onboard at just 16GB. It is more stable than the pure SSD solution. For some reason my mobo doesn't like SSD drives. If that were not the case, by far the best solution would be a 128GB SSD cache paired with a 3TB spinning drive.
On the other hand, if hybrid technology fulfils its promise we will probably see tablets with hybrid drives in them.
Edit: The decline in notebooks and desktop PCs is obviously a headwind in the short to medium term. It will take a while for cloud, hybrid and NAS to pick up the slack.
Why Natural Gas Has Not Lifted Coal Stocks [View article]
"In between, the most common sources of cycling capacity are natural gas combined cycle generation turbine (CCGT) plants. They burn gas to make steam to turn turbines to generate electricity and they also recycle heat to boost plant efficiency."
Minor Quibble: Combined Cycle generation turbine gas plants actually burn the gas in the same way as a jet engine (Brayton Cycle) and then use the very hot exhaust gases to create steam for a steam turbine (Rankine Cycle). The Brayton cycle operates at about 50% thermal efficiency and the Rankine cycle converts about 25-30% of the waste heat into electricity, giving a combined thermal efficiency of about 60-65%. Compare this to a coal plant that typically manages to convert coal to electricity at about 33% efficiency depending on a bunch of factors.
Disagree with you on the investment/consumption thing. You will need shelter all your life. The house you live in is comsumption, pure and simple. It is a liability, not an asset. It will never earn you money. It will cost you money forever, just like your car.
But the cost of consuming a house you own vs. the cost of consuming a house you rent needs to be compared. You have no choice but to consume, one way or the other. And if you are saying that a cashflow analysis shows that you are usually far better off to consume by owning (essentially to have less other consumption now in order to be able to have more other consumption later) then I agree fully. It works even with a long-term decline in the value of the house you buy, which would make it a very bad choice as an investment but still a good choice for consumption.
Wall Street Breakfast: Must-Know News [View article]
My theory is that the head of YUM China is in the process of pulling off the biggest sandbagging coup of the century. He got his quota down by 25% and is already exceeding target by 8%. Next: promotions, which stopped at the end of last year, will be reinstated.
Go to the top of the class for being the only analyst I have read to date who picked up on the real story in the last week's data. Industrial production up 9.2% + Exports up only 1% == Domestic consumption up very robustly, which is exactly what the doctor ordered for China's economy.
China is not running its economy for the benefit of western commodities exporters. For some time now there have been quite targeted moves to transition the economy: a) The construction industry has been actively suppressed for nearly 2 years now, and this is ongoing. b) Lower-value-add manufacturing is no longer an "encouraged industry." No wonder we are seeing textiles moving to Vietnam and other places, just as it moved from Hong Kong to the mainland in the 90s, and moved from Europe to Asia in the 80s. c) On the other hand I can get all sorts of tax concessions for starting a service-sector company in Beijing at the moment. d) The minimum wage has gone up by more than 50% in the last 2 years, and factory job salaries have doubled - all aimed at promoting consumption.
The guys in Beijing are not dumb. They know that the faster growth is, the harder it is to transition from investment-based to consumption-led growth. Expect growth in China to continue to be suppressed as they work out how much growth is really required to stave off unemployment.
But Cisco is only American in that its CEO sits in California and it is listed on the NASDAQ. It's ownership, customers, R&D, workforce and manufacturing are all international. Why should it pay tax in the US on income that had nothing to do with the US?
Sounds like what Shanghai taxi drivers tell you when they try to tell you the fare is RMB100 from Hongqiao airport to the Bund.
Food prices have been volatile - for example the price of pork tripled and then went back down about 6 years ago, and vegetables go up and down in price as well. It's all about seasonality and supply. As of this week, food is cheap. Except for shallots for some reason. They are up 3 times over what they should be.
Inflation has been up and down recently, but it has never hit 10%. I have lived in 10+% inflation and I know what it is like. China has not had that in the past 14 years that I have lived here. But there have been one or two times during which poor families temporarily could not eat meat.
Last Call For Vietnam: Buy Now Or Be Sorry [View article]
I've been pursuing business (sales side, not investment side) in Asia for about 14 years, and have to second the cautionary notes on Viet Nam. On Corruption: For those who criticised the link to Transparency International for it's measuring perceptions rather than reality, I'd point out that (a) it's hard to measure the reality because there are no official statistics, almost by definition. The perrceptions do, however, come from businesspeople who experience the reality every day. TI indexes correlate closely with my experience. I prefer to think of corruption as a tax that happens on every transaction - dollars that go illicitly into the pockets of people who have no contribution to the transaction, whether it be in producing goods and services, buying those goods and services, providing liquidity and risk minimisation, information that helps parties accurately price and analyse the transaction, etc etc. Every country has that tax. In " squeaky clean" countries it might be as low as 2%, but it is still there. For example, I would rate USA at about 3.5%, Australia about 2.5-3%, France 5%, etc. China was about 10-13% back in 1998, improved to perhaps 8% in 2002/3 but has been rising ever since Hu Jintao and Wen Jiabao took over from Jiang Zemin and Zhu Rongji. Right now it is probably back to 11%. Indonesia has been consistently over 10% but is probably falling. India swings between about 10% and more than 15% depending on where they are in the election cycle. Thailand used to be at 8-9% and falling, but Thaksin spiked it to well north of 15% and reduced the share going to the Army and the Monarchy, which may explain the coup. Vietnam is certainly of the order of 15% right now, putting it very much in the top bracket.
With respect to manufacturing competitiveness with China: Vietnam certainly has a cheap workforce, but it suffers from poor connection to the supply chain, dramatically lower availability of highly skilled people such as industrial engneers, and significantly higher sovereign risk. As such it can compete for the lower end of the manufacturing dollar, especially in those industries that China is starting to move away from but I don't see Hon Hai opening up an Ipad factory in Vietnam any time soon.
Long term Vietnam will prosper. Short term you may be able to profit from a rise in Vietnam stocks. Any time frame in between, I would be vewy vewy careful. It's not a buy and hold proposition.
The Sole Apple-Beater: Samsung Electronics [View article]
It all depends where you go and who you are. Take my local computer bazaar in Beijing. 12 months ago when you walked in, it was all Apple. There were 15 shops selling Apple, and I had to go right down to the back to find the one place selling Samsung tablets and phones. Now when you walk in the door, they have equal billing and equal numbers of retailers. Yesterday the iPhone 5 16G was going for RMB5400 and the Galaxy Note II for RMB5100. None of the local brands were much in evidence. On the other hand, if you go to YaShow which is the discount consumer bazaar you will be inundated with Lenovo, Huawei, HTC, Xiaomi etc Android 2.3 devices selling for around the RMB1000-1500 mark.
If you go into the average office, Android will outnumber IOS by 5 to 1, but they will mostly be the low-end devices. Xiaomi seems especially popular among the office ladies.
Annaly Capital (NLY) is reiterated a Sell with price target cut to $13.50 from $16 at FBR as its proposed purchase of CreXus is admittance its business model of Agency MBS purchases is toast in the current environment. Isn't the purchase of CreXus a step in changing the business model though? [View news story]
Hmm. Should one be taking financial advice from someone who doesn't know the difference between admittance and admission?
The Hard-Disk Duopoly Is Really Cheap, But For How Long? [View article]
Yes. I have tried it. And I like it. My desktop has a 256GB SSD right now, along with 2x1.5TB magnetic drives. I also enjoy having Photoshop open so much faster. But I can't buy an SSD drive big enough to fit W7 and Adobe CS for my notebook. So on my notebook I still need my 500GB rotating drive. The new genaration of Ultrabooks will have both an mSATA SSD *and* a 5mm magnetic drive. Check the latest on what Intel is proposing in this space. You get the best of both worlds at a much lower price.
Most people *could* manage with a 256GB SSD drive, but why *would* they when there is a superior and cheaper solution available either with 2 drives or with a hybrid drive. The standards are already there, and the products are coming out now. You will plug drives into your notebook/ultrabook like you plug RAM in now.
The alternatives are interesting. All of them are years away from being produced in commercial quantities, except for FRAM, the old chestnut with which Ramtron have been fleecing gullible investors since the 1980s. It's the orbital engine of electronics.
Unlikely to crash. Still a lot fewer notebooks sold with SSD drives than with rotating strage, and we need a couple of generations of SSD before that can change. SSD storage is still significantly more than 5x the price/GB of rotating storage but with a significant increase in performance of course. Also, consumer SSDs just aren't big enough for most people's needs. This is not going to be helped by Win8. Every time SSD size doubles we turn around and need more storage.
In the meantime, enterprise storage applications and NAS are growing at 20+% per year and hybrid drives and motherboards with flash cacheing technology are in the process of to significantly narrowing the performance (but not so much energy usage) gap. I expect hybrid magnetic/flash storage to match SSD storage in performance within a year or so, still being at a disadvantage in power usage, but having a solid advantage in price and capacity.
It's a bit of a technology race in the notebook, and the other side of it is that those tablets and ultrabooks, because they are so storage-starved, are driving storage growth in backup servers, external hard drives, NAS and the cloud. So the calculation is can STX and WDC make as much money by selling smaller numbers (but higher total storge amounts) of higher-end disks in NAS and cloud - where are the lines going to cross, and what's the profitability at that crossing point? One would expect ASP to stay solid, but see some softness in the growth in the total number of units until the cloud/NAS/external drive units catch up. (Currently client compute is more than twice the number units of the rest combined.)
To me the biggest threat is not share of storage units in notebooks - that will be robust for some time, but an overall reduction in the number of notebooks, which is definitely on the cards. So far there seems to be no sure sign of this, but tablets are getting more capable on a semi-annual basis.Next quarter's numbers when both Seagate and Western Digital are firing on all cylinders are going to be very interesting. It might be the highwater mark for client compute. (Desktop+Notebook+Ultr... or at least (notebook+ultrabook).
In the meantime, expect Seagate and WDC to enter the consumer SSD market. They, along with Sandisk, are already the main players in the (small) Enterprise SSD market. Building SSDs is relatively low-tech and low-margin compared to magnetic storage unless you are Toshiba or Sandisk and are vertically integrated.
The Hard-Disk Duopoly Is Really Cheap, But For How Long? [View article]
WDC,STX and Toshiba will ship about 450-500 Exabytes of capacity this year. about 180 of that is notebooks. Total NAND Flash production is maybe 25-30 Exabytes a year at a stretch, and a lot of that cannot go to SSDs because it is going into phones and tablets etc. What's more it still costs ~10x as much as magnetic storage and some manufacturers are losing money at that price. The standard ultrabook architecture within 12 months will be a ~30GB NAND Flash mSATA drive with a 500GB 5mm magnetic drive sitting right next to it.
Notebooks will not convert away from magnetic storage to SSD. It raises the cost enormously and has a capacity penalty while the new architecture matches pure SSD in performance and (eventually) power consumption at a much lower cost. SSDs are shock resistant, but they do have a limit to their useful life that magnetic storage does not have. To my mind reliability is a wash and will not change people's purchasing decisions either way.
Then you have the whole cloud phenomenon which is growing magnetic storage at ~23% CAGR by units shipped (off a low base). One of the trends we are seeing right now is the movement of capacity from the device to the cloud.
SSD has a bright future in enterprise analytics engines and other high-speed database applications. But who sells these SSDs? Sandisk, Seagate and Western Digital (now. Was Hitachi).
The biggest threat to STX and WDC is the potential for collapse in the notebook/ultrabook market in favour of tablets. I expect we will see tablets with magnetic storage eventually, but there will be a dip in the curve because notebooks are such a huge chunk of the business today.
The other threat is, as always, discounting. It will be really interesting to see the first fully-consolidated quarter of WDC numbers next month. Current ASP is around $67-$68/unit and at that level it is very profitable. Before the floods it got as low as $42. If the ASP can be kept above $55-$60 given current softness in the PC market, I would see that as very bullish for WDC and STX.
The Elephant In The Room For Seagate Technology [View article]
The decline in PC shipments is a threat to the HDD manufacturers. SSDs are not.
Can Hybrids Extend Life Of Seagate And Western Digital? [View article]
If the former turns out to be true, then the spinning hard drive will go the way of the dinosaur. (Yes. Spinning platters will still be needed in the cloud, but the revenue from that is nothing like the revenue from noteboks and PCs.) The big problem with SSD is that there is not enough Fab capacity in the world to populate all notebooks with SSD.
If the latter turns out to be true, then hybrids will do well and STX and WDC are pretty-much the only game in town.
I have over a terabyte of photos alone. Solid-state storage will never be the full solution for me, and there is no cloud service that can meet my needs at a reasonable price. Also, in much of the world, cloud storage just isn't fast enough or economic to use as primary storage because of Internet speed and volume caps/pricing.
I converted my main desktop PC's system disk to hybrid a couple of months ago. Bootup speed is only very slightly slower than with an SSD. Application loading is better than a traditional hard drive but not as good as an SSD drive. But I suspect that is because second-gen hybrids still have not quite enough flash onboard at just 16GB. It is more stable than the pure SSD solution. For some reason my mobo doesn't like SSD drives. If that were not the case, by far the best solution would be a 128GB SSD cache paired with a 3TB spinning drive.
On the other hand, if hybrid technology fulfils its promise we will probably see tablets with hybrid drives in them.
Edit: The decline in notebooks and desktop PCs is obviously a headwind in the short to medium term. It will take a while for cloud, hybrid and NAS to pick up the slack.
Why Natural Gas Has Not Lifted Coal Stocks [View article]
Minor Quibble: Combined Cycle generation turbine gas plants actually burn the gas in the same way as a jet engine (Brayton Cycle) and then use the very hot exhaust gases to create steam for a steam turbine (Rankine Cycle). The Brayton cycle operates at about 50% thermal efficiency and the Rankine cycle converts about 25-30% of the waste heat into electricity, giving a combined thermal efficiency of about 60-65%. Compare this to a coal plant that typically manages to convert coal to electricity at about 33% efficiency depending on a bunch of factors.
Should You Buy A House Now? [View article]
But the cost of consuming a house you own vs. the cost of consuming a house you rent needs to be compared. You have no choice but to consume, one way or the other. And if you are saying that a cashflow analysis shows that you are usually far better off to consume by owning (essentially to have less other consumption now in order to be able to have more other consumption later) then I agree fully. It works even with a long-term decline in the value of the house you buy, which would make it a very bad choice as an investment but still a good choice for consumption.
Wall Street Breakfast: Must-Know News [View article]
China's Transition Continues [View article]
Go to the top of the class for being the only analyst I have read to date who picked up on the real story in the last week's data. Industrial production up 9.2% + Exports up only 1% == Domestic consumption up very robustly, which is exactly what the doctor ordered for China's economy.
China is not running its economy for the benefit of western commodities exporters. For some time now there have been quite targeted moves to transition the economy:
a) The construction industry has been actively suppressed for nearly 2 years now, and this is ongoing.
b) Lower-value-add manufacturing is no longer an "encouraged industry." No wonder we are seeing textiles moving to Vietnam and other places, just as it moved from Hong Kong to the mainland in the 90s, and moved from Europe to Asia in the 80s.
c) On the other hand I can get all sorts of tax concessions for starting a service-sector company in Beijing at the moment.
d) The minimum wage has gone up by more than 50% in the last 2 years, and factory job salaries have doubled - all aimed at promoting consumption.
The guys in Beijing are not dumb. They know that the faster growth is, the harder it is to transition from investment-based to consumption-led growth. Expect growth in China to continue to be suppressed as they work out how much growth is really required to stave off unemployment.
Cisco, Left For Dead [View article]
Amazing Cisco Blunder [View article]
China's Transition Continues [View article]
Food prices have been volatile - for example the price of pork tripled and then went back down about 6 years ago, and vegetables go up and down in price as well. It's all about seasonality and supply. As of this week, food is cheap. Except for shallots for some reason. They are up 3 times over what they should be.
Inflation has been up and down recently, but it has never hit 10%. I have lived in 10+% inflation and I know what it is like. China has not had that in the past 14 years that I have lived here. But there have been one or two times during which poor families temporarily could not eat meat.
Last Call For Vietnam: Buy Now Or Be Sorry [View article]
On Corruption:
For those who criticised the link to Transparency International for it's measuring perceptions rather than reality, I'd point out that (a) it's hard to measure the reality because there are no official statistics, almost by definition. The perrceptions do, however, come from businesspeople who experience the reality every day. TI indexes correlate closely with my experience. I prefer to think of corruption as a tax that happens on every transaction - dollars that go illicitly into the pockets of people who have no contribution to the transaction, whether it be in producing goods and services, buying those goods and services, providing liquidity and risk minimisation, information that helps parties accurately price and analyse the transaction, etc etc. Every country has that tax. In " squeaky clean" countries it might be as low as 2%, but it is still there. For example, I would rate USA at about 3.5%, Australia about 2.5-3%, France 5%, etc. China was about 10-13% back in 1998, improved to perhaps 8% in 2002/3 but has been rising ever since Hu Jintao and Wen Jiabao took over from Jiang Zemin and Zhu Rongji. Right now it is probably back to 11%. Indonesia has been consistently over 10% but is probably falling. India swings between about 10% and more than 15% depending on where they are in the election cycle. Thailand used to be at 8-9% and falling, but Thaksin spiked it to well north of 15% and reduced the share going to the Army and the Monarchy, which may explain the coup. Vietnam is certainly of the order of 15% right now, putting it very much in the top bracket.
With respect to manufacturing competitiveness with China: Vietnam certainly has a cheap workforce, but it suffers from poor connection to the supply chain, dramatically lower availability of highly skilled people such as industrial engneers, and significantly higher sovereign risk. As such it can compete for the lower end of the manufacturing dollar, especially in those industries that China is starting to move away from but I don't see Hon Hai opening up an Ipad factory in Vietnam any time soon.
Long term Vietnam will prosper. Short term you may be able to profit from a rise in Vietnam stocks. Any time frame in between, I would be vewy vewy careful. It's not a buy and hold proposition.
The Sole Apple-Beater: Samsung Electronics [View article]
If you go into the average office, Android will outnumber IOS by 5 to 1, but they will mostly be the low-end devices. Xiaomi seems especially popular among the office ladies.
Annaly Capital (NLY) is reiterated a Sell with price target cut to $13.50 from $16 at FBR as its proposed purchase of CreXus is admittance its business model of Agency MBS purchases is toast in the current environment. Isn't the purchase of CreXus a step in changing the business model though? [View news story]
The Hard-Disk Duopoly Is Really Cheap, But For How Long? [View article]
Most people *could* manage with a 256GB SSD drive, but why *would* they when there is a superior and cheaper solution available either with 2 drives or with a hybrid drive. The standards are already there, and the products are coming out now. You will plug drives into your notebook/ultrabook like you plug RAM in now.
The alternatives are interesting. All of them are years away from being produced in commercial quantities, except for FRAM, the old chestnut with which Ramtron have been fleecing gullible investors since the 1980s. It's the orbital engine of electronics.
Seagate: The Good And The Bad [View article]
In the meantime, enterprise storage applications and NAS are growing at 20+% per year and hybrid drives and motherboards with flash cacheing technology are in the process of to significantly narrowing the performance (but not so much energy usage) gap. I expect hybrid magnetic/flash storage to match SSD storage in performance within a year or so, still being at a disadvantage in power usage, but having a solid advantage in price and capacity.
It's a bit of a technology race in the notebook, and the other side of it is that those tablets and ultrabooks, because they are so storage-starved, are driving storage growth in backup servers, external hard drives, NAS and the cloud. So the calculation is can STX and WDC make as much money by selling smaller numbers (but higher total storge amounts) of higher-end disks in NAS and cloud - where are the lines going to cross, and what's the profitability at that crossing point? One would expect ASP to stay solid, but see some softness in the growth in the total number of units until the cloud/NAS/external drive units catch up. (Currently client compute is more than twice the number units of the rest combined.)
To me the biggest threat is not share of storage units in notebooks - that will be robust for some time, but an overall reduction in the number of notebooks, which is definitely on the cards. So far there seems to be no sure sign of this, but tablets are getting more capable on a semi-annual basis.Next quarter's numbers when both Seagate and Western Digital are firing on all cylinders are going to be very interesting. It might be the highwater mark for client compute. (Desktop+Notebook+Ultr... or at least (notebook+ultrabook).
In the meantime, expect Seagate and WDC to enter the consumer SSD market. They, along with Sandisk, are already the main players in the (small) Enterprise SSD market. Building SSDs is relatively low-tech and low-margin compared to magnetic storage unless you are Toshiba or Sandisk and are vertically integrated.