At the end of last week the large-cap stock market, as measured by the S&P 500 Index, was within 1/10 of 1% of where it began the year. However, even though there has been very little net change in the overall stock market, there has been a lot of price movement occurring in individual stocks and sectors.

For example, the S&P 500 is flat for the year, but the tech-heavy NASDAQ Composite Index is up nearly 7%. That’s not surprising given the inherent advantage that tech sector have over every other sector; it isn’t reliant on a single input value such as oil that can swing wildly in price, it isn’t dependent on another sector for demand for its output such as basic materials, and it isn’t subject to the whims of societal trends that affect so many consumer discretionary companies.

For those reasons, there are always “good” tech stocks to buy if you can figure out exactly which ones they are. And that’s exactly what we try to do every week by utilizing three distinct approaches – and portfolios to match – in evaluating the tech sector. We’ve had a lot of success with that approach, and will continue with it into 2016.

However, we are also always on the lookout for new ways to identify future stock market winners, and this year we discovered an approach that is quite different from anything else I have seen before in my more than 30 years of stock market experience. It comes to us courtesy of Linda McDonough, a former hedge fund analyst who has written for this publication many times in the past.

Beginning next month Linda will begin managing a new service for us based on her proprietary methodology for finding what we like to refer to as “Launchpad stocks”, so named for their propensity to take off shortly after they display the unique set of attributes she is looking for. Fortunately for us, she is sharing one of her very first picks with us below. And if you’d like to know more about other stocks she has her eye on, you can click here to find out more about those in the weeks to come.

By Jim Pearce

Supercomputers

By Linda McDonough

Bring up the subject of supercomputers and images of giant reels spinning and machines the size of refrigerator come to mind.  Those of us far from the world of high performance computing (HPC) have been trained to believe that the evolution of technology delivers smaller and smaller packages.  Yet the geeks of the supercomputing world are knee deep in large, unspeakably fast supercomputers.  And customers are lining up to buy them.  

International Data Corporation (IDC), the IT market research guru, expects the high performance computing (HPC) market to grow significantly faster than other hardware markets over the next five years.  Growth, expected to be 10% annually, is being driven by the need to solve increasingly complex problems with reams of data.

Importantly, IDC notes that smaller players will grow significantly faster than the market.  The top three vendors, Hewlett Packard, IBM and Dell, will see smaller increases as they lose market share to rising stars. Cray, who came in at number five, is one of those stars enjoying bubbling growth.  In the first nine months of this year it grew revenue 53%!

As the excitement in high performance computing rarely makes front page news we likely would not have noticed its resurgence except for some outsized trading volume in Cray Inc. (Nasdaq: CRAY) in late October.  That trading was our first clue that Cray might qualify as a “Launchpad” stock. Launchpad stocks are stocks undergoing a significant shift in fundamentals.  They are first identified by a surge in trading volume. The stocks are then weeded through to confirm that the volume was instigated by a new event that will improve earnings. The next step is to analyze each company’s financial statements to ensure fiscal health.  The final phase is to determine if the stock offers a good value to investors. Investing Daily will be introducing a newsletter based on the Launchpad system early next year.

While the increase in trading volume for Cray was related to its commercial customers growing as a percent of total revenue, secular trends in high speed computing are boosting demand for many vendors.  Standalone computers do not have the processing power to operate multiple processors at once, a requirement when analyzing different data sets simultaneously.

The proliferation of data from networks all over the world needs to be managed. Files monitoring our tax payments, highway travel (think EZ Pass) and possibly, cell phone records are collected by the government. Mouse clicks, Google searches and online purchases are tracked by marketing firms.  Even our “likes” on Facebook are fed into a giant data repository. The force of ocean surges is measured by remote equipment and sent to meteorological centers.  Daily stats monitoring the health of lab animals on new drugs are collected by biotech research firms.

Some of Cray’s new customers are energy companies using the company’s XC40 supercomputers to process seismic data to improve drilling accuracy and for reservoir simulation, a process used to predict the flow of liquids through the ground.  Financial customers are using Cray’s computers to scan for credit card fraud and to enforce cybersecurity on their networks. Manufacturers are using them for predictive modeling to forecast the strength of different materials under various stress scenarios.

While these may sound like problems that can be handled by off the shelf servers, supercomputers are needed to maintain speed and accuracy when the volume of data being fed into mathematical models increases geometrically.  These computers are not your ordinary machines. Cray’s Trinity supercomputer, being used by Los Alamos National Laboratory, is comprised of five rows of twelve cabinets linked together.  Each cabinet is roughly six feet tall, two feet deep and two feet wide resulting in a total footprint of almost 1,500 square feet.  Some rooms housing fleets of supercomputers are the size of basketball courts.

Of course other component makers should flourish with the increased demand for these computing goliaths.  Intel’s second generation Xeon Phi processor is used in many of these machines.  Nvidia’s graphics processing unit (GPU) named Tesla, works in conjunction with central processing units (CPUs) to deliver lightning fast processing.  CPU’s are effective but work by processing one unit of data at a time.  GPU’s can simultaneously process thousands of pieces of data, increasing the velocity of analysis.  AMD’s Firepro GPU also competes in this space.

As new commercial applications for high powered computing grow, demand from academic and government agencies remains high.  The Titan, a supercomputer housed at the Oak Ridge National Laboratory, is the second fastest computer in the world and is owned by the U.S. Department of Energy.  It is used to study global warming patterns and to design super-efficient solar panels.  The Department of Defense recently purchased a supercomputer from Cray to study the force of blasts on combat vehicles, soldier protection in heavy war zones and cybersecurity network modeling.

Although Nvidia, Intel and AMD will earn additional profits from this growth in supercomputers, the new business is not large enough to move the needle on such large bases of revenue. Cray, our favorite name in the group, is a pure play on high performance computing. Nearly all of its revenue is generated by supercomputer hardware sales or service revenue.  The stock trades at a P/E of 20 despite estimates that predict 36% earnings growth next year. Its pristine balance sheet has $5 per share in cash and zero debt.

Investors looking to cash in on the next wave of big computing should jump on the back of this data devourer.

NOTE: If you would like to learn more about Linda’s new service, please click here to sign up for a “sneak peak” at what she has in store for 2016.

NASDAQ Composite Index:                                                                       

Thursday, December 24 = 5,047.49                                                  

Year to Date = + 6.8%                                        

Trailing 4 Weeks = – 1.8%

Trailing 7 Days = + 2.5%