Another Look At The Low-Cost Building Blocks Of Storage - Part 1

| About: Western Digital (WDC)
This article is now exclusive for PRO subscribers.

Shares of Western Digital (NYSE:WDC) were sold for investors accounts on August 13th, 2012 at $44. As of close of market on Friday Oct. 19th, the shares were being traded for a mere 34.88. A decline of a full $9.12 or almost 21% in barely 9 weeks. Is the value of the company really less than it was, when the shares recently traded at $44?

wdc chart 10-22-2012

IMG 1327-copyRecently the opportunity to go traveling presented itself again, and couldn't be passed up. This time, the trip would include Switzerland and parts of Italy. For a change the cell phone camera was left behind in favor of a new Canon (NYSE:CAJ) T4i. Canon is a remarkable company that produces remarkable products and that incidentally (its stock) is also on sale. This device, like so many others generated exponentially more data / media than its predecessor. There was a sense of guilt splurging on a new camera, but after returning triple digits to investors for some time, perhaps it could be forgiven (especially since no fee or commission had ever been charged, although the profits for this handful of individual investors had already climbed into the millions).

Human nature can be curious. When Amvona was building its photo equipment and online business, great care was taken to keep everything confidential - despite this the ideas were often copied and the strategies that Amvona pioneered in those days can still be found on in other businesses large and small to this day.

When Amvona was resurrected as a site that publishes (amongst other things) securities analysis as a means to discern sound investments virtually all of the ideas (that time permitted) were published and made freely available to the public. More than two years later, and despite consistent triple digit returns, hardly anyone seemed to notice or care. Even intelligent businessmen, with under-performing portfolios, but who are otherwise successful have turned down help (offered entirely for free), and although an abundance of evidence exists, cast doubt on the entire operation. From this a great lesson regarding human nature (and parenthetically) the markets has been learned - that is, you can show people to the truth (supported by all the data and evidence in the world), and often it is still rejected, particularly if it threatens an existing world view.

IMG 2017-copyGetting back to the T4i - it's a whole new camera, the 18 Megapixel CMOS sensor is incredible, as its the ability to shoot multiple images at great speed - when this is paired to the radically faster i/o speeds of the latest flash cards a new meaning arises from the term "data storage."

Upon returning home it seemed there must be over a thousand (high res.) images on that tiny card. When the images were transferred to the hard drive on the desktop there were actually over 2200 (representing 12 days of travel).

Is there a misunderstanding?

Generally it seems the market feels that somehow SSD's will replace HDD's. This is not correct, at least for the time being. They are at the very least complimentary - indeed the more high capacity, high speed SSD's which come into existence, the greater the need for HDD's. SSD's are temporal storage for data that is intended to remain in perpetuity, but if this media is to be truly "stored", it will be for the foreseeable future on mechanical hard drives. However it is not enough to recognize the complimentary nature of SSD's and HDD's, but also to understand the long term changing dynamics of data storage and retrieval, particularly as it relates to three concepts:

1. The creation and growth of Cold Storage (e.g. Amazon Glacier)

2. The long tail of data storage (mathematical eventualities)

3. The potential in data which is not yet stored.

One type of media storage is dying, it is called "Tape," and it will be widely replaced for archival storage purposes, but not by SSD's which are altogether unnecessary for such an application. HDD's have a much lower cost per GB than tape, yet much faster access times. Therefore, drives specifically designed for this purpose, needless to say, have the potential to capture "Exabytes" of existing market share from tape.

Jay Parikh, VP of infrastructure engineering at Facebook (NASDAQ:FB) recently made comments which appear to support the HDD over Tape thesis, although his comments were subtle, and difficult to interpret given how he delicately circumvents the issue, nonetheless they appear to mean that HDD's might well be preferable in next compute challenge - to quote:

"… Facebook is rethinking the infrastructure for how it stores huge repositories or photos and videos in a way that's accessible and convenient but also cost-effective. Unlike a business that might store records on tape, Facebook can't afford to let users wait that long to access something, nor can it afford to build data centers that keep photos in caches next to the servers (those Fusion-io machines aren't cheap!).

'The current data center and hardware design is actually very sub-optimal for that problem," Parikh says. "People say to us we should just use tape, but I'd rather poke my eyes out with chopsticks.' "

"Facebook's next compute challenge is cold storage" (Gigaom Oct. 3rd, 2012)

Even if the message is not clear, one thing is certain, in order to avoid a potential medical emergency Jay ought to avoid Chinese food at all costs.

IMG 2378-copyPeople scoff. The hard disk as we know it is "old school," and it's death has been predicted for a long time. SSD's are cool, small fast, etc., companies that make them even have cool names like "SanDisk" (SNDK), companies that don't have boring names like "Western Digital" which sounds like it's from the transistor era (It is). However, the market may be misunderstanding the relationship of the two.

For example many types of engines have emerged since the dawn of powered flight, the progress is astonishing. However, the basic shape of an airfoil (the very thing which produces lift) has not changed nearly as much, nor the idea of how to control the air which flows over its surface (what the Wright brothers initially called "wing warping" and later became "ailerons"). The basic function of lift and control some hundred years later is still intact. The hard drive is somewhat newer, and although greatly improved, the basic concept is still intact because it works.

For this reason, even exciting new companies, with far greater premiums on their market price, are endeavoring to find new ways to work with the HDD - for example, Prinevile, Oregon (amongst other towns) will soon have close to one million sq. feet of cold storage space dedicated to the HDD as indicated in the following:

"The plan is to use the building to house a brand-new type of low-power deep-storage device that Facebook engineers will cook up over the next six to nine months. They're designing a hard-disk storage server that powers off when it's not in use, says Tom Furlong, vice president of site operations at Facebook. 'It's going to sit in a dedicated building that is optimized to support this device that we don't need to access very often.' "

Facebook Will Ice Backup Data in 'Sub-Zero' Building (Wired Magazine August 18th, 2012)

To be clear the reference is to a "hard-disk storage server," not an SSD storage server.

Facebook even after its epic stock price decline still trades at a P/E ratio of 65 and a P/B ratio of 3.05. With the above article it is evident that at least part of the billions raised from the victims of the FB IPO will be used to build large storage facilities for Hard Drives - and even though the storage may be in a way "free" to the end user, the drives are not free to FB or any other operator endeavor to build such a facility - of which there are many.

There was one other part of the article that seemed to have at least minor importance, namely:

"Facebook doesn't use tape backups - they're too kludgy and expensive for the social media company. But it does save two backup copies of all of its data: one that is there to be used whenever there's a problem on the live server, and a second copy for emergency use only."

Facebook Will Ice Backup Data in 'Sub-Zero' Building (Wired Magazine August 18th, 2012)

It is interesting to consider what might happen if an article or two was published that spoke of millions of square feet of storage space being built to house SSD's, and that companies building these facilities were going to require two back-up copies of everything (on SSD's of course) - and in fact that whole new types of servers would be built specifically for these SSD's - what would happen to companies such as SanDisk and their stock price (which is already trading at roughly 3x the forward P/E of WDC, although WDC is more profitable)?

The market appears to like what is "technologically exciting" much more than what is "financially exciting" (WDC has better margins that SNDK for instance), even though few probably appreciate just how sophisticated these devices are. According to Pantelis Alexopoulos, executive the director of Singapore's Data Storage Institute:

"The customer doesn't know it's a marvel of engineering." A drive, he says, "is more sophisticated than a Boeing 747."

Pantelis Alexopoulos's Push for Slimmer Disk Drives (Businessweek September 27th, 2012)

Boeing 747s are still being produced with hundred year old technology (see above). That having been said, even modern jets have more than two primary suppliers - the same cannot be said of HDD's.

Although it might at first seem a bit of a revelation, apparently there is over 4 times more content created than is actually stored on digital media. According to WDC, capturing only 5% of this "unstored data" would equal all of the Exabytes shipped on enterprise disk and tape combined (See Driving Datacenter Change: Storage Opportunities in the Cloud [Mike Cordano September 13th, 2012] )

IMG 2436-copyIt is worth contemplating if perhaps the greatest market opportunities (not to be confused with rate of growth) in storage are not in the smallest and fastest (e.g. PCIe solutions - High performance SSD / NVM, etc.), but rather these so called "Cold Storage" Solutions with their commensurate low cost per TB and TCO advantages. That having been said, it seems also more than likely that WDC will pursue both ends of the spectrum aggressively although they are not equal.

It does not take much calculus to understand that the faster a high performance memory card functions, the greater the number of low end drives are ultimately needed. The relationship is hardly adversarial, in fact it is more than complimentary, it could even be said that the high performance SSD card "needs" the HDD for its very existence. In other words, there would be limits on its speed, if there were nowhere to store all of the media over the long term at a very low cost. This is a significant and fundamental paradigm shift in the very nature of how data is stored and utilized.

Even if images from a family trip to Europe are just of average quality, they will still be viewed many times on Facebook or other photo sharing sites - but not for very long. After a matter of days, they will be all but forgotten in the online world that has become our collective consciousness. That is to say, content on social media sites is characterized by high media access when it is at first new, then dramatically lower accesses as it ages even slightly. This is the very nature of the new world we live in, content may have great breadth and even depth, but at once is met by the short attention span of a changing society. However, the HDD's the media is ultimately stored on (unlike the SSD it is created on) will never be completely idle, because the content remains, and must be available at any time for access.

It is good if a technology company makes its first priority financial performance. Many technology companies with brilliant ideas have failed because there is a certain romance, even seduction to pure technology, it burns in the mind of the true technologist - like true innovators, these folks struggle to build successful enterprises. However, WDC is not run by a true technologist (although they employ thousands), they are run by guys interested first and foremost in sound management and superior capital allocation - and to that end, there is no management team greater in technology or otherwise.

IMG 3056-copyConsider for a moment the fact that both FCF and GM% over the last ten years have increased in tandem, despite intermittent declines (as the current instance) in the price of the shares. In fact, prior to the HGST acquisition WDC's gross margins averaged 18-23%, however the company now projects margins in the range of 27-32%, an increase of about 50%. The earrings announcements and guides provided since March (when the transaction closed), seem to prove the new altered, and higher margin business model. Given that OPEX, CAPEX and the Tax Rate are set to remain about the same, it seems WDC has entered a new phase of profitability.

How does he know?

Friday, October 12, 12:58 PM More on Citi's downgrade of Seagate (STX -0.7%) and Western Digital (WDC -2.4%): Analyst Joe Yoo thinks Q4 hard drive sales could total just 140M units, below Seagate's lowered guidance of 150M-155M. Moreover, he thinks Windows 8 will "largely be a non-event for traditional PCs;" and that high-margin enterprise drive sales (Seagate has more exposure here) may have fallen 27% Q/Q in Q3. But he's still bullish on the industry long-term, viewing it as a play on general data storage growth.


WDC indicated at their September 13, 2012 investor conference that the company expected the TAM to be 140 M units - so it is not correct to say that this is what "Analyst Joe Yoo thinks" but more correctly, this is what "Analyst Joe Yoo learned." And it is not that they "could" but rather, that they "will" ship ~140 M units.

With the formalities out of the way, it is curious how Joe Yoo could possibly know that "Windows 8 will 'largely be a non-event…' "

If he is correct, then it can be inferred that Mr. Yoo is in fact more insightful and perhaps intelligent than a plenitude of CEOs who have invested heavily in the success of Windows 8. It's possible he is. However if one often changes opinions, engages in sudden upgrades and downgrades, and generally appears to go with the prevailing "spirit of the times" there are possibilities other than intelligence and the gift of foresight.

It is always possible that one fears moving against the crowd, and prefers the comfort found in numbers - in the multitude. However, is it not always the crowd who lose on Wall Street? When in the history of the stock exchanges did the majority win? Wealth could not be created for the few if this were to occur. It is easy to say "the PC is dead" or "Windows 8 is no good" - after all who is not doing that? It is hard to actually critically analyze companies and historical performance - and to ask the opposite question in the later instance - who is doing that?

Microsoft has never enjoyed the popularity in the minds of consumers that say Apple (NASDAQ:AAPL) or Krispy Kreme Donuts (KKD) has. Despite this by every financial metric, it is a force to be reckoned with.

A few questions were raised about the opinion of analysts in the Amvona Article Silver Platters, Annie Oakley and "The Sidelines" published on May 1st, 2012. It is hard to find websites that accurately track the wisdom of analyst opinions - this must be by design, so going back and reading the above article was enlightening, because it kept the temporal nature of these sort of comments in a historical perspective.

At the time the article was published back in May the concern was over the future of HDD pricing and its supposed erosion - which did not materialize. Now the thesis seems more difficult to prophesy - the outcome of Windows 8, even before it launches.

As late as May 9th, Citi had a "buy" rating on the shares of WDC. In the five months that followed the delta in Price Target issues by the firm have equated to around 10 B in value - or more (by about 20%) than the market cap. of the whole company as of the writing of this article.

Wednesday, May 9, 6:26 PM Western Digital (WDC +2.1%) finished higher on the day after filing its FQ4 10-Q. Citi (Buy) points out WDC disclosed it has repurchased 7.8M shares in FQ1 as of May 8. Moreover, given the company was in a quiet period until April 26, Citi believes all of the buybacks, which cost the company $305M, occurred over the last 2 weeks. Comment! [Tech, On the Move]

The Amvona earnings FQ1 13 estimate for the record is between $2.85 and $2.95 per share (unlike previous articles, it's a rough back of the envelope estimated) - while acknowledging the possibility that the company could miss even on consensus ($2.45-2.55) - it is not acknowledged or accepted that the company should thus be rated a sell. Also if forced to bet, it seems reasonable that Windows 8 will do just fine.

If speculating on the success or failure of Win8 is not enough, hiding in plain sight is this fact:

17 10.8%
Rev. Δ
0.2 4.8%

According to the company the Total Addressable Market (TAM) is set to decline by roughly 10.8% in the current quarter but the related revenue only by about half as much.

According to Wolfgang Nickl, CFO Western Digital:

"We now expect the total available market for hard disk drives in the September quarter to be 140 million units. That is down from 157 that we guided to in the July earnings call. As a result, our revenue will be lower assuming flat share. And our new revenue range is $3.9 billion to $4.0 billion. However, and, here's the important part, we expect that our gross margin on a non-GAAP basis will continue to be at 30% like we guided to in the July call."

The word "assume" is interesting in the context of market share. It is used often through the investor presentation when references are used to market share. The question is what prevents the company from taking further market share? Visits to retailers have affirmed to incredibly important points regarding WDC's consumer business:

1. The Company has more shelf-space than the competition.

2. The Company's products have a superior look and feel (presentation)

3. The Company appears to have far greater pricing power.

Further with regards to the share repo:

"You might recall last quarter, we finished with an ending authorization of $1.3 billion. I can tell you we spent about $150 million quarter-to-date. And with the additional $1.5 billion, we have around about $2.7 billion left, which is 25% of the market capitalization today."

Western Digital Corp. - Analyst/Investor Day

mypassport edgeThese guys are smart. In the prior quarter (Q4 F12) they spent about 600 M on share repurchases, that is to say when the shares were often in the low 30's. However in the current quarter, when the shares were often above 40, they spent a mere 150 M. Now the shares are back in the early 30's, and one has to wonder how busy the company's broker is. The company claims the share repurchases are "linear." If so, this is the finest "linear" repurchasing ever, because it synchronizes beautifully with the company's stock price performance (which has been anything but linear) - and unlike so many other public companies with share repo's, the company is actually buying at highly favorable prices, which is extremely beneficial to ongoing owners, who benefit every time the share price goes down.

And on the Dividend:

"We do see the value of the dividend and therefore, we have decided to initiate a dividend, $0.25 per share on a quarterly basis. The 8-K will say that it will go to shareholders of record that's of September 28, and that's the last day..."

Western Digital Corp. - Analyst/Investor Day

The fact that WDC presently has the ability / authorization to take back 25% of the entire concern should be startling to sellers of the stock because it could mean the company might be taken private or be acquired.

If neither of the above happened, the dividend will attract dividend investors and funds focused only on companies that pay a dividend (i.e. the stock is now attractive to both value-oriented and income / dividend investors). These facts, along with the realization that the HDD is already decoupled from the traditional PC food chain, may well be the catalyst that releases the share price to more normal valuation ratios.

Continue to Part 2

Disclosure: I am long WDC. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it. I have no business relationship with any company whose stock is mentioned in this article.