SGI Sees Revenue Growth in 2012

Silicon Graphics International (SGI) sells data processing systems with a focus on technical computing. Their high-end systems are used to solve some of the toughest computational problems encountered by government and industry. They also provide datacenter equipment for cloud computing to companies like Amazon.

The new SGI is basically a combination of the old Rackable Systems, which specialized in datacenter server systems, and the old, bankrupt SGI, which specialized in high-performance science computing. I had my doubts when the two companies merged, but the new company turned out to be better at executing than either of the old companies.

The transformation of SGI is about technology, sales, and profit margins. In last Thursday's report on fiscal Q4 2011 (ending June 24) we saw record revenues for the typically slow Q4. Revenues were $195.5 million, up 36% sequentially from $143.7 million and up 92% from $101.6 million year-earlier. While GAAP net income was negative $12.1 million, non-GAAP net income was $3.9 million. Of course I would prefer to see GAAP net income in the black too. In this case the difference is largely due to non-cash operating expenses, restructuring, and software revenue recognition rules (because the computer systems have software bundled with the hardware). As a check on the merits of GAAP vs. non-GAAP, the cash balance was up $9.4 million in the quarter.

SGI is debt-free and had a cash and equivalents balance of $143 million at quarter's end. They can fund a strong R&D effort and could make acquisitions if needed.

The new Altix UV supercomputer line has no real direct competitors. It has a memory and processor model that make it very attractive to high-end technical users. Because of that profit margins are good. The rackable systems for server farms and cloud computing continue to offer innovative designs, but margins have been improved there as well.

The company is now truly international, which is important when your key products are supercomputers. Service revenue is also a key factor in the new, profitable business model.

In response to questions from analysts management went into some detail, and speculation, about technologies they are developing. They are working with Microsoft to expand the capabilities of SQL Server. They believe that for certain types of computing they will be able to deliver performance equivalent to Oracle's Exadata systems at about one-third of Oracle's current price. Considering how successful Exadata has been, both in terms of compute ability and revenue generation, that could be highly significant in 2012.

Analysts also speculated that with budgets thin, SGI might have difficulty selling its computers to government agencies. However, SGI has little or no exposure to state and local governments in the U.S. Federal agencies still seem eager to decrease their other costs by upgrading their compute capabilities. For industry, the total cost of the design process is decreased by buying more computational power.

While not giving guidance by quarter, for fiscal year 2012 (running to June 2012) revenue is expected between $740 and $780 million, up to 24% over fiscal 2011.GAAP EPS is estimated between $0.15 and $0.30. Non-GAAP EPS expected between $0.60 and $0.80.

For more details on quarter results, see my SGI Q4 fiscal 2011 analyst call summary.

Disclaimer: I am long SGI.

The usual risks apply, so keep diversified.

See also: http://www.sgi.com/

Oracle (ORCL) Sun Shines

Yesterday Oracle (ORCL) issued stunningly good results for its 4th fiscal quarter (Q4) 2011 ending May 31st. The price of the stock promptly took a major dive in after-hours trading, and is still down substantially as I write.

Usually when this sort of thing happens, a good report followed by a price fall, it is because either: (1) investors expectations were even higher or (2) guidance for the next quarter is below analyst expectations.

This time the issue is something far more specific: the revenue derived from a specific segment, computer hardware, that was acquired with Sun Microsystems last year. Before reading my commentary below, you might want to first familiarize yourself with Oracle quarter results with my notes on Oracle Q4 2011 Analyst Call.

Despite an overall increase in Oracle revenues of 13% from year earlier, hardware systems product revenues were $1.16 billion, down from $1.23 billion in the year-earlier quarter. That raised some flags with some analysts. Which mostly shows these guys have too many stocks to cover, or were hired for their connections, not their analytic ability. Oracle management explained their strategy before, and they explained it again during the question and answer session.

I'm with management on this one: Sun was poorly run from a business perspective, and Oracle is not obligated to repeat Sun's business mistakes. Rather than looking at specific products, Oracle looked at the overall Sun problem: low margins. Fortunately, those low margins were not all of a kind.

Oracle decided to stop selling low margin computer hardware products whenever possible, with the exception of those that margins could be raised over time. So a year later, they are selling slightly less product, but making far more profits on it. Not only that, but they are gearing up to sell a lot more product, in particular their Exadata and Exalogic lines. What is there not to like about that?

Oracle is a rapidly growing, undervalued technology stock. We are not in a tech bubble. A gold bubble, yes. A bond bubble, yes. Idiots are holding bonds at this point. Bonds are trash, in most cases not bringing in enough returns to keep even with inflation. Their principal is just as much at risk as it would be if invested (wisely) in quality stocks.

I don't own Oracle stock at this point, but with the price drop and a bright future, I could buy ORCL in my next buying round. The competition is mainly small cap tech and biotech stocks that may grow even faster, and yet are also cheap right now. [See what I own at William Meyers current stock portfolio]

One aspect of the situation not mentioned was Oracle's current war with HP over Intel Itanium based computers. I don't see much downside for Oracle here. Oracle software can run on many types of hardware. Itanium was always a bad idea, and Oracle should not be obligated to support two different Intel chip architectures.

Oracle, Databases, Verticals and Business Intelligence

Oracle (ORCL) reported on its third fiscals quarter ending February 28, 2010 last Thursday. Along with other technology companies with fiscal calendars and some guidance changes, all indications are that the quarter ending December 31, 2009 was not an anomaly for the industry. Both consumers and businesses are resuming their purchases of hardware and software. While some areas of the world are now lagging, particularly Europe, on the whole the increase in demand is broad based. The release an adoption of Microsoft Windows 7 as well as new PC CPUs from Intel and AMD is making the refresh cycle attractive for servers, desktops, and laptops. The high-end cell phone craze is just adding to the rally. With some exceptions, technology stocks are undervalued, but probably will remain in that state until more money comes out of bonds and into stocks.

I'd buy Oracle right now for myself if I had a larger portfolio and wanted another relatively large cap stock. I believe Oracle is going to continue to eat SAP's and IBM's lunch, and probably steal some of Microsoft's afternoon snack as well. Now that Sun has been swallowed by the great white whale, or shark, companies that make high-end database server hardware need to fear Oracle as well. In Thursday's analyst conference Oracle management explained that they are already tearing out the unprofitable parts of Sun, like commodity hardware, and focusing on ramping the profitable parts. Sun has impressive hardware; combining it with Oracle's databases and other enterprise software application packages is going to allow companies to get bundles of value for their most demanding applications like transaction management. The Oracle hardware system for this is called Exadata, and it may prove to be a giant killer. Coming soon is the integration of Oracles software packages in Fusion, with will have built-in, across the board Business Intelligence (BI).

Microsoft, of course, is a far more diverse company than Oracle. It has a gravy train in its Windows operating system that is still under no real threat from Linux. Its database system for businesses, SQL Server, is competitive at every level from the home office to the enterprise. But at the enterprise level, SQL Server's main advantage is ease of use and integration into the legion of Windows PCs on corporate networks. For truly big enterprises, Oracle has two major advantages apart from its database's quality. Oracle is able to target verticals, which are industry segments like oil, financial services, and biotechnology, in a way Microsoft cannot. It also has a broader and probably better set of applications that are enterprise-specific such as ERP (enterprise resource planning). Oracle does not need to compete with Microsoft Office, and Office is just not that big of an edge when an enterprise is looking at something like a transactional database to deal with Internet content provision and commerce.

See also my Oracle analyst conference summary for fiscal Q3 2010 and www.oracle.com

As I write, according to Nasdaq, Oracle stock is trading for $25.36 cents, giving it a (non-GAAP) P/E of 16.6 and a forward P/E of 13.6. Which equates to annual earnings on investment of 6% and 7.35%. Which, however well it compares to other technology stocks, is way, way better than being in bonds, which are particularly unsafe now since any increase in interest rates will cause the sale value of past-issued bonds to fall.

But do Keep Diversified!