The Real Internet Boom
Cockeysville, Maryland
* * * Still more telecom deals…
* * * Microsoft says it will begin to distribute pay-per-view video over its Xbox Live network. Mr. Softy bought 1,000 hours of content from various studios (and the rights to distribute the Ultimate Fighting Championships) and plans to offer these to video gamers for something like $1.99 per episode.
Xbox Live is set up to operate in "store and forward" mode, like your TiVo. In plain English, your Xbox will save the video content to its hard drive before showing it on the screen, causing some time delay – some latency – before the programming begins to play. This store-and-forward model eliminates transmission problems by first building a memory cache, then playing without interruption. I don’t think people will care whether or not it takes 30 seconds for the show to start, as long as there aren’t any commercials or other interruptions once the program begins.
You won’t be able to watch the Xbox Live network without a hard-drive equipped Xbox. Currently, Xboxes are equipped with a 20-gigabyte hard drive. But an hour-long television show in high-definition (HD) format takes up at least 2 gigabytes. Storing HD videos will fuel demand for ever-larger hard drives. The same is true with TiVos.
Last Friday, I recommended the stock of a company that supplies the key material in hard drives to the world’s largest hard-drive makers. These companies manufacture 75% of the world’s hard drives, including those used in the Xbox. If you agree with me that store-and-forward networks, such as TiVo and now Xbox Live, will be the most desirable way of buying video content, this is a stock pick you shouldn’t miss.
* * * Video soaks up a lot of bandwidth. Level 3 (LVLT) is the network provider for Xbox Live. Its stock is up 5% today… and nearly 100% over the last 12 months.
* * * More video on the Internet: Verizon is in advanced talks with YouTube in order to bring the Internet videos to Verizon subscribers. This would be a major move to bring videos to cell phones. Verizon also will offer YouTube as an OnDemand feature of a TV service that’s part of the company’s fiber-to-the-home rollout. The deal should close in a matter of weeks.
* * * Today, China’s holdings of U.S. Treasury obligations will exceed $1 trillion. I don’t know what this means, but it doesn’t feel good knowing that we owe them so much. I can’t think of one thing I own from China that I really need.
* * * We stuck our foot in our mouth yesterday in regards to our new quantitative investment service. While the data on our housing stock indicator was correct, an editing mistake led us to publish incorrect figures regarding the recent price history of the home builder’s index. The point we were making was still valid – homebuilder stocks soared and then crashed – but our numbers were off. Here’s what we meant to say:
– From August ‘04 to August ‘05, homebuilders gained 92.7%
– From August ‘05 to November ‘06, homebuilders lost -40.7%
In regards to Ian’s very contrarian view, Toll Brothers released truly horrible news today. New orders are down 56% from last year. Even worse, the company will write-off as much as $100 million in land options in the next quarter. Nobody in his right mind would buy this stock… but it’s up for the day.
Maybe Ian’s on to something.
* * * Reader TL Lune says we’re the loony ones: "You have all now officially lost your minds… The poor people of this country are worse off by virtually every measure available in the time since Wal-Mart became a national institution – but that doesn’t stop Dan for a second… Stop telling the rest of us that we do not have the right to feel the way we do, or that you are somehow morally superior to the rest of us."
Porter Comment: We specialize in facts, not feelings. And, having known our most of our editors for long time, I can assure you that none of us would hold the moral high ground for long. Except for maybe Sjug. That guy really is a Boy Scout.
* * * Alliance member Jim Pursley makes a good point about Wal-Mart and creative destruction: "…I found it interesting that Dan had to argue more than the bull case for the stock. His arguments had merit – in fact I wrote a piece in my client letter the following week which mused that creative destruction in tech is OK but in retailing it isn’t."
* * * Finally… I hope you will pardon a personal note of congratulations. My good Italian friend, Michele Castelli, completed the New York Marathon last weekend in three hours and 26 minutes. His pace was a respectable 7:53 per mile. He came in 720th for his age group.
Most marathoners aren’t much for mid-race conversation. Usually, nobody talks during an entire race. But Michele amazingly talks nonstop when he runs. He made more friends during his race than anyone else, telling everyone how "grey-ot" (great) they were doing and about the New Year’s Eve party in Miami a few years ago when "I kissa de man by mistake… I thought he was wooman." In terms of character, charm, and personality, Michele came in first by a record margin of victory. Michele, congratulations on your successful race. Salute!
* * * * * * * * * * * * * * * * *
You remember Moore’s Law, don’t you?
Moore’s Law, as it was popularly repeated during the 1990s, predicted microchips would become 100% more efficient every 18 months. Thus, the amount of computer power you could buy today for $1,000 would double in 18 months.
It was one of the three big lies of the 1990s tech boom. The other two were: The Internet grows by an order of magnitude each year and "Metcalfe’s Law," which claimed that networks grow more valuable by the coefficient of their nodes.
None of these things are true.
Moore’s Law was first proposed in a 1965 Electronics Magazine article penned by Gordon Moore. At the time, Moore was an engineer at Fairchild Semiconductor, but shortly thereafter he became a co-founder of Intel. Today, he is one of the wealthiest men in the world. He was certainly right about something… but what was it?
In his article, Moore predicted that semiconductors would grow 100% more efficient every two years. It turns out that there’s a big difference between 18 months and two years, compounded over time.
Moore’s prediction was popularized by America’s leading chip designer, Carver Mead and then bastardized by the press and stock promoters.
Even today, most investors still consider Moore’s Law to be the foundation of all value creation in technology.
But that’s wrong too…
Over the last 10 years, the speed of the Internet backbone has increased 20,000 times over, lowering the cost of communications bandwidth by about 80% per year. Hard drives also are getting cheaper at a faster rate than chips. The cost-effectiveness of hard drives – digital storage – grew 7,500-fold over the last 10 years. Lagging behind, like the runt of the litter, microchip efficiencies have increased only about 60-fold over the same period.
As I originally discovered and wrote about as early as 2000, there is a definite link between communications bandwidth and demand for digital storage. I discovered this when I visited Celera’s gene-cracking factory in Rockville, Maryland. Its network architect told me the company had plenty of computer power and plenty of bandwidth. The bottleneck, he said, was storage.
Let me give you a more mundane example of the storage problem. Consider the TiVo digital recorder. TiVos and similar digital recording devices have radically altered the dynamics of television. Easier to use than a VCR, this digital recorder allows you to record 60 hours of video. That’s probably as much storage as any regular user will ever need. However, as the bandwidth of the cable system grows, more and more high-definition (HD) channels will surely be added. The same 60-hour TiVo drive will only hold about 16 hours of HD programming, which isn’t nearly enough to satisfy most users. Consumers obviously will need much larger amounts of digital storage as HD programming becomes more prevalent.
You can see why consumers are going to want a bigger hard drive as the bandwidth into their homes increases.
Now… multiply that demand by the biggest number you can imagine. That’s Google’s demand for additional storage. Google is attempting to catalog and store every publicly available web page. Its demand for storage is directly linked to the growth of the Internet.
Google is also giving storage away for free to its Gmail users.
On April 1, 2004, Google launched the beta version of Gmail, its free online e-mail service. To differentiate itself from the other free e-mail services (Yahoo! Mail and Hotmail in particular), Google provided users with a mind-boggling amount of memory – 1 gigabyte – for free. It also allowed users to send files far larger (up to 10 megabytes) than the other two popular services do.
A megabyte is 8 million bits of data – about enough memory to store a typical novel. A gigabyte is a 1,000-times more storage. One gigabyte of digital memory can hold 18 hours of MP3 encoded music… or about 11 hours of low-resolution video.
This extra memory has proven to be enormously useful to people like me. I have all of my corporate e-mail automatically forwarded (or "mirrored") into my Gmail account. And, because I’ve never deleted a single e-mail from it, I have a record of all my corporate communications, all of my notes, all of my newsletters… In fact, instead of writing down someone’s phone number, I’ve started simply sending e-mails to my Gmail account with the name and number. It’s easy to find, despite the large amount of mail in my inbox, because the same search engine that powers the Google website also will troll through all of my e-mails, finding anything I need almost instantly. Using free software called Gmail Drive (www.viksoe.dk/code/gmail.htm), you can also use the free storage on Gmail to store any kind of file you want.
Gmail started a memory boom.
Soon Yahoo! and Hotmail were providing similar amounts of storage for free. Recently, a new competitor, 30Gigs.com, advertised e-mail with up to 30 gigabytes of storage – for free.
To keep pace with the trend it started, Google is now constantly expanding the total memory available to each Gmail user. If you go to Gmail’s login page, you see a memory counter that’s constantly expanding by about 15 bytes per hour. Total memory available to each user is now more than 2.7 gigabytes.
Google won’t tell me how many Gmail users it has or how much data it’s storing for them – but it’s clear the amount of data is massive and growing rapidly. Google is building a 34,000-square-foot data center on the Columbia River in The Dalles, Oregon. (A 1.8 gigawatt hydroelectric power plant on the Columbia produces some of the world’s cheapest electricity, and data-storage centers are voracious energy consumers. According to tech researcher George Gilder, the top-five search-engine companies, which together have around 2 million servers working for them around the clock, suck up as much power in a year – about 5 gigawatts – as Las Vegas uses on its hottest days.)
You can see Google’s response to the coming need for thousands of times more memory every year in its incredible capital-spending budget. As recently as 2003 (before it gave away free storage), Google was spending less than $200 million a year on capital expenditures (things like data centers). In 2004, the capex budget grew by more than 50%, to more than $300 million. In 2005, Google’s capex budget more than doubled, to more than $800 million. And currently, the company is spending about $700 million per quarter to buy additional servers and digital storage.
To compete with Google, Microsoft says it will spend two dollars for every one dollar Google spends on resources for its Internet search and free e-mail business. Almost all of this money will fund data-storage and data-search hardware.
What’s behind this insatiable need for more memory?
Let’s start with the obvious. Everything you see or hear on the Internet, on your TiVo, or on your iPod has to be stored somewhere on a magnetic disk. Consider that just two years ago, when Google started giving away a gigabyte of storage for free, there were 37 million domain names (websites) in use. Today, there are 65 million domain names in use.
Website growth has been so fast over the last 10 years that the basic operating system of the Internet – Internet protocol – is running out of addresses and will need to be replaced in the next two or three years. The federal government is leading the upgrade, mandating that all agencies must be using Internet protocol version six (the upgrade) by 2008. Other major institutions will follow suit.
The continued rapid expansion of the Internet, along with faster and faster data lines, means more and more data-storage media will be needed. Broadband penetration rates are rapidly approaching a key milestone – the 20% penetration rate. With previous technologies, such as cell phones, VCRs, and fax machines, this amount of penetration triggered changes in marketing and media. Today, we’re seeing the same thing happen with video over the Internet.
It’s obvious that the Internet will continue to grow rapidly in terms of total content. As the amount of content increases – especially video content – people will want faster and faster Internet connections. As these pipes get "fatter," digital storage demand will grow exponentially.
That’s why I’m bullish on digital media storage. I think it will continue to be one of the fastest-growing areas of technology.
And, while there are a handful of hard-drive makers, there’s only one company that specializes in making the aluminum platters that all drive makers use. In short, this company builds the "guts" of a hard drive and every major hard-drive manufacturer relies on it. This makes it easy for us to invest directly in the growth of demand for digital storage.
I detail the company in this month’s issue of my newsletter, Porter Stansberry’s Investment Advisory.
Good investing,
Porter Stansberry
Stansberry & Associates Top 10 Open Recommendations
| Stock | Symbol |
Buy Date |
Total Return |
Publication |
Editor |
| Seabridge |
SA |
7/6/2005 |
417.80% |
Sjug Conf. |
Sjuggerud |
| Exelon |
EXC |
10/1/2002 |
244.56% |
PSIA |
Stansberry |
| Crucell |
CRXL |
3/10/2004 |
245.23% |
Phase 1 |
Fannon |
| Am. Real. Partners |
ACP |
6/10/2004 |
213.18% |
Extreme Value |
Ferris |
| Akamai |
AKAM |
11/1/2005 |
201.39% |
PSIA |
Stansberry |
| Sirna |
RNAI |
1/13/2006 |
196.50% |
Phase 1 | Fannon |
| Humboldt Wedag |
KHDH |
8/8/2003 |
175.02% |
Extreme Value |
Ferris |
| EnCana |
ECA |
5/14/2004 |
148.56% |
Extreme Value | Ferris |
| Cons. Tomoka |
CTO |
9/12/2003 |
145.00% |
Extreme Value |
Ferris |
| Alex. & Baldwin |
ALEX |
10/11/2002 |
130.70% |
Extreme Value |
Ferris |
| Top 10 Totals | ||
|
5 |
Extreme Value | Ferris |
|
2 |
PSIA | Stansberry |
|
2 |
Phase 1 | Fannon |
|
1 |
Sjug. Conf. | Sjuggerud |
Stansberry & Associates Hall of Fame
|
Stock |
Symbol |
Holding Period |
Gain |
Publication |
Editor |
| JDS Uniphase Corp. |
JDSUD |
1 year, 266 days |
592% |
PSIA | Stansberry |
| Medis Technologies |
MDTL |
4 years, 110 days |
333% |
Diligence | Ferris |
| ID Biomedical Corp. |
IDBE |
5 years, 38 days |
331% |
Diligence | Lashmet |
| Texas Instruments |
TXN |
270 days |
301% |
PSIA | Stansberry |
| Cree Inc. |
CREE |
206 days |
271% |
PSIA | Stansberry |
| Celgene |
CELG |
2 years, 113 days |
233% |
PSIA | Stansberry |
| Nuance Comm. Inc. |
NUAN |
326 days |
229% |
Diligence | Lashmet |
| Airspan Networks |
AIRN |
3 years, 241 days |
227% |
Diligence | Stansberry |
| ID Biomedical |
IDBE |
357 days |
215% |
PSIA | Stansberry |
| Elan |
ELN |
331 days |
207% |
PSIA | Stansberry |
