Warren Buffet has said that it’s seldom possible to buy a great company for a bargain price; about the best you hope for is to buy a great company at a fair price. This is the present situation with NATI where a business with a tremendous technology franchise is currently trading at 17.4x  and 15.8x my 2008 and 2009 FCF estimates, respectively.  The downside risk to the stock appears to be pretty close to zero.  It has occasionally dropped below $25.00 (down 9% from here) over the last five years, but those episodes have been infrequent and brief.  The upside, however, is substantial since reasonable expectations suggest a share price that is 40% higher two years from now and 100% higher five years from now.  When I get into the numbers, I’ll show why those estimates may be conservative.

 

Engineers and scientists make frequent use of test and measurement (T&M) equipment for myriad tasks such as conducting experiments, testing new design concepts, validating prototypes, testing end products for quality assurance, and automating processes.  A few years ago, I visited a factory that made superconducting wire and noticed that they were using NATI T&M equipment to inspect the wire as it moved through the fabrication process.  In that same facility, I saw engineers using other NATI equipment to test and diagnose defective portions of wire.  On a separate occasion, I saw a test station where an engineer was using T&M equipment to measure the impact resistance of motorcycle helmets.  In yet another situation, I saw an automated microscope that a scientist had programmed to count the number of cell cultures in a Petri dish.  These are a few examples of the tens of thousands of diverse applications for T&M equipment. 

 

Historically, T&M equipment has come in the form of relatively large “boxes”—steel enclosures that are filled with electronics and have displays, buttons and knobs on the front.  These instruments have only been able to perform vendor-defined measurements—ones that the manufacturer thought its users would find helpful and are preprogrammed into the box.  It is analogous to financial calculators.  These calculators can perform calculations like NPV’s, but the user interfaces are unintuitive and users are limited to the functionality and methodology already embedded in the device; you can’t customize calculations the way you could in a spreadsheet.  In addition, performing different types of measurements usually requires buying different equipment for each type of measurement, stacking them up in racks, and configuring a system that will make the instruments work together.  (The resulting systems are often referred to as “rack and stack” instrumentation.)  Such systems have been the standard approach for decades, but they are expensive, inflexible, and take a long time to configure.  These systems are also sometimes subject to rapid obsolescence in cases where technology standards are changing.

 

NATI was founded in 1976 and originally produced General Purpose Interface Boards (GPIB’s) that allowed computers to interface with T&M equipment.  In 1986, they developed a new software product called LabVIEW (Laboratory Virtual Instrumentation Work Bench) that represented a revolutionary approach to T&M and would eventually become a world-class technology franchise.  LabVIEW is a programming language that scientists and engineers use to create their own “Virtual Instruments” (VI’s).  Instead of creating a “rack and stack” system, customers buy peripheral equipment such as sensors, plug them into a personal computer, and then build their own virtual instrument on the PC.  LabVIEW comes with preprogrammed VI’s for users who want to make standard measurements, but the beauty of the technology is that it lets users modify and customize their instruments to make measurements exactly the way they want to—they’re not limited to vendor-defined measurements.  This is analogous to the difference between a financial calculator and a spreadsheet.  A spreadsheet offers incredible flexibility so users can customize their calculations and present them just the way they want to.  VI’s made through LabVIEW look like a virtual-version of a traditional instrument.  The PC monitor will show any “displays,” “knobs,” or “buttons” that the user “built” into the VI, hence the name “virtual instrumentation.”

 

Importantly, LabVIEW is a graphical programming language.  Users drag and drop icons, create lines between the icons to represent the data flows, and add in any necessary algorithms to create their virtual instrument.  This intuitive approach has made LabVIEW extremely accessible since it doesn’t require its users to know traditional programming languages such as C++.  This accessibility has proven very empowering to scientists and engineers who often don’t know how to code, but would like to have instrumentation that fits their unique needs.  As a result, LabVIEW has become one of the most widely used programming languages in the world.  (LabVIEW is so accessible that Lego uses a simplified version of it in their Mindstorms product, a Lego kit that allows kids to build and program their own robots.)

 

This ease of use makes it much faster to set up instrumentation, and the resulting systems are usually far less expensive than traditional “rack and stack” systems.  Being software based also makes the instrumentation easier to change which is very attractive in areas where technological standards are changing.  For example, NATI’s RF test equipment can be updated through software upgrades, whereas traditional RF test equipment has to be replaced.  Given these advantages, LabVIEW has attracted a loyal customer base.  I spoke to one scientist who said he will not buy equipment unless it comes with LabVIEW drivers and can thus be easily integrated into LabVIEW virtual instruments.

 

While LabVIEW software is NATI’s technological franchise, it is part of a much larger T&M ecosystem that includes sensors, industrial computers, data acquisition devices, and a wide array of other hardware.  This hardware is manufactured by NATI as well as third parties.  Engineers pick out the components they will need for their particular applications in much the same way that residential contractors pick out items at Home Depot to complete their various projects.  In some cases, a given piece of equipment will be manufactured by both NATI and a competing vendor such as GE’s Fanuc division.  In these cases, however, NATI has a distinct advantage because of the seamless integration between their hardware and LabVIEW.  As one engineer explained to me, he could use GE Fanuc hardware if he wanted to, but it is faster and easier to build a system out of NATI components because they are perfectly integrated with LabVIEW.

 

A final point about the LabVIEW’s franchise is that as a software product, it has a learning curve which creates switching costs and thus a competitive advantage.  NATI itself faces this when they are trying to win new customers.  Alex Davern, NATI’s CFO, explained that when a company is considering buying LabVIEW, they aren’t thinking about the $2,000 they’ll spend for the license.  They’re thinking about short-term productivity of the $80,000 / year engineer who will have to spend time learning the new software before putting it to use.  Once this engineer has learned LabVIEW, however, NATI is well positioned since users typically like working with LabVIEW and its tight integration with NATI hardware gives NATI an advantage in winning follow-on hardware sales

 

While revolutionary, LabVIEW was far ahead of its time, and it’s initial adoption was limited by a host of factors.  The first is that it ran only on Macintosh computers which confined it to a relatively small group of users.  Even when the PC version came out, however, LabVIEW’s power was limited to the processing power of the conventional PC’s that it was running on, which made it less capable than traditional instrumentation which had custom hardware.  While this was a limiting factor in the early years, it has proved to be a very attractive strategic position as NATI’s instruments have piggy-backed on the $BB’s of R&D spent by hardware companies like Intel and Dell as they have advanced the processing power and capabilities of conventional PC’s over the years.  Each generation of new hardware from these companies has made NATI’s instruments progressively more powerful, and this will continue for the foreseeable future.  (The capability of a test system can be thought in terms of the precision of the measurements being taken and the frequency with which the measurements are being made.  This capability is constrained by a host of factors such as the processors themselves and the BUS’s that transfer data from one device to another.  NATI’s instruments lagged in capability for many years, but in the last two years have reached parity with traditional T&M instruments for all but ultra precise measurements taken at low frequency.) 

 

In the beginning, NATI’s products were primarily used in the lab by customers who valued LabVIEW’s flexibility and didn’t require the high-end capabilities offered by traditional instruments.  As mentioned previously, improvements in PC’s extended the competitive reach of NATI’s products over time.  Concurrent to this, management was filling out the company’s product line to allow NATI to address new needs and enter new markets. 

 

Many of the measurements used by NATI’s customers in the lab were also used on the factory floor for things like quality assurance and machine control.  Factory applications, however, required a much higher degree of reliability than LabVIEW could offer because LabVIEW ran on Windows OS and was thus susceptible to Window’s reliability issues (remember “Blue Death”?).  In order to meet the reliability requirements of these applications, NATI developed its own OS, LabVIEW Real-Time.  They also began to develop “control” functionality into their products so measurements could be translated into decisions.  A simple example would be an assembly line that makes electric motors.  Before the motors are shipped, they need be tested to measure things like vibration and temperature.  If they fail the test, a decision has to be made (i.e. this motor is defective) and an action has to be taken (i.e. actuate the arm that will kick this motor into the reject bin).  NATI also began to ruggedize their products so they could withstand harsh environments in the factory and outdoors. 

 

Adding control functionality began to make NATI products more suitable for “embedded” applications.  As technology has progressed, an increasing number of traditionally simple products have begun to incorporate electronics to make their operation more variable and sophisticated.  Going back to the motor example, motors used to be made to operate continuously regardless of how hot they became.  Today, however, even simple products like vacuum cleaners and papers shredders often incorporate electronics to monitor the motor’s temperature and disable the motor if it gets too hot and is at risk of burning out.  This is referred to as “embedded” design since embedded software is incorporated into a product’s electronics.  Five years ago, one of my aerospace companies recognized how important embedded software was becoming and acquired a company that specialized in it.  Two years ago, I spoke with an auto supplier who explained to me that an ideal engine would have “variable everything,” so all of the components changed their operation as the conditions changed.  This requires embedded software.

 

In many embedded applications, however, processing speeds must be incredibly fast (i.e. measured in nanoseconds) which is beyond the reach of software that runs on a general processor like those found in PC’s.  Achieving these speeds requires putting the logic right onto the microprocessor itself, as through an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array).  LabVIEW’s flexibility and accessibility makes it uniquely suited for engineers who are trying to add “intelligence” to their designs, but NATI needed to find a way to put VI’s onto the microprocessor.  In order to meet this need, they developed LabVIEW FPGA which would translate LabVIEW VI’s into code on a FPGA.  (This improved the accessibility of FPGA’s since the language required to program FPGA’s is a particularly difficult one.)  NATI also developed a new product platform called Compact RIO (Compact Reconfigurable Input Output) which is a small, rugged industrial computer that uses an FPGA to perform very fast measurement and control calculations.  Compact RIO has proven to have many applications, but one example that I saw was a custom motorcycle where Compact RIO functioned as the electronic brain that controlled all of the engine’s functions, such as timing.

 

Compact RIO is today being used in factory settings and on prototypes, but it is usually spec’d out at the Beta unit or production stage due to its relatively high variable cost relative to something like an ASIC (which has high fixed costs but low variable costs).  There are some instances, however, where cRIO is getting OEM’d into end products such as sophisticated machines that are made in low volumes and thus can’t justify the high fixed costs associated with an ASIC.  There are structural dynamics in the chip industry that are gradually increasing the number of applications that can be addressed by FPGA’s (and thus cRIO), but those are beyond the scope of this report.

 

Like most technology companies, NATI experienced tremendous revenue growth during the late 1990’s, but this came to a halt during the industrial recession of 2001 when revenue declined 6.1% and then grew only 1.4% in 2002.  This was a relatively shallow decline compared to the industry, but it was NATI’s first revenue decline in its 24 year history.  Believing that the industrial economy would remain slow for many years, management wanted to change the company’s model so it could introduce a continual stream of new products and thereby grow independently of the economic cycle.  Since LabVIEW had been introduced in 1986, management had come across users who were using the software in creative and non-conventional ways which demonstrated NATI’s opportunity to expand within and beyond their traditional T&M market.  Concurrent to this desire to increase their development efforts, the tech wreck created a sea change in the availability of technical talent which dramatically improved NATI’s ability to recruit engineers and enter new areas.  Agilent, for example closed its RF effort in Santa Rosa, CA and moved that initiative to Malaysia.  NATI seized on this opportunity, hiring many of these engineers to begin its own RF development center. 

 

In addition to increasing NATI’s investment in R&D, management choose to begin augmenting NATI’s S&M efforts.  As a percentage of sales, these expenses ramped from 2000 to 2003 and increased from a combined 49.6% of sales to 54.3% of sales.  This was concurrent to a 320 b.p. decline in gross margin resulting from a mix shift in revenue.  Together, these factors pushed operating margins from 18.6% in 2000 down to 11.3% in 2003.  It’s important to note that most of this decline was deliberate and stemmed from senior management’s goal of dramatically increasing NATI’s reach over the long-run.  Such long running sacrifices were made possible in part by the fact that Dr. James Truchard, NATI’s founder, president, and chairman owns 22% of the company’s stock.  (His co-founder, Jeffrey Kodosky, holds 5% of the company’s shares, though Mr. Kodosky no longer works full-time at the company.)  Dr. Truchard is an effective manager, but he is also a visionary and a long-term thinker.  (When he founded NATI, he wrote a 100 year business plan.) 

 

I have attended NATI’s analyst days since 2002, and each year NATI’s CFO has talked about how these investments would translate into revenue growth and margin expansion.  Beginning in 2003, revenue did begin to pick up, but the company was increasing their R&D and S&M expenses so quickly that margin gains remained modest through 2005.  I think a lot of investors just gave up after each year’s projection of margin expansion fell short of expectations, especially since NATI’s stock was expensive at that time.

 

One the 4Q05 conference call, however, something changed.  Dr. Truchard said, “I will focus on increasing our revenue growth and maintaining budget discipline while positioning the company for long-term success.”  “Budget discipline?”  This was coming from a company that had just spent the last five years hiring nearly every talented engineer that walked in the door.  Then on the 1Q06 CC, he said, that NI had “solid expense management” during the quarter.  Dr. Truchard had begun to accelerate the margin improvement that Alex Davern had been talking about for four years.  NATI’s financial history is shown below:

 

 

 

 

 

 

 

 

 

 

 

(In retrospect, I think there was a huge problem of “time compression”—that phenomenon where managements announce a new development and investors then expect its impact to be imminent.  NATI’s long-term focus, and the audacious nature of their plans amplified this disconnect.  Investors like myself dramatically underestimated the amount of time over which this progress would occur.  NATI wasn’t just developing new products, they were developing new architectures and entirely new platforms.  Once launched, the product lines around those platforms had to be filled out and matured.  Early adopters had to be won over, and the products had to progressively prove themselves in the marketplace.)

 

 

NATI’s stock price has been stagnant over the eight years.  Coming into this decade, NATI had a very high valuation like many technology companies at that time.  Management’s aggressive investments over the next several years served to depress earnings and keep this multiple high.  Since earnings began to grow again in 2004, the firm has grown into its stock price and the multiple has consolidated as shown below:

 

 

 

 

While it is difficult to know just how much a multiple can compress, I believe that NATI is fairly priced at $27.44.  This represents 17.4x 2008 EPS, ex. SBC.  I’m projecting FCF of $1.49 and $1.63 for 2008 and 2009, which translates into Net P/FCF multiples of 17.4x and 15.8x, respectively.  While not “cheap” in an absolute sense, it is a good deal for a company with these kind of returns on capital (20% ex. cash), visionary and capable management, and the opportunity to continue taking share in existing markets and entering new markets.  This is a disruptive company with a tremendous track record that is changing the way that scientist and engineers go about their work.  Most importantly, I believe that we are in the final innings of a 9 year investment cycle that will lead to accelerating revenue growth and margin expansion, the latter of which might be in excess of management’s current guidance.  There are a number of forces coming together that support these assumptions.

 

NATI’s organic revenue has been steadily growing in the low double digits for the last ten quarters despite an industrial economy that has been decelerating for the last six quarters and declined in 1Q08.  So low double digit growth is a starting point, sustained by ongoing marketshare gains and the maturation of existing products such as cRIO which had triple digit revenue growth in 4Q07, four years after its launch.  New products and new application should also support this low double digit revenue growth.

 

In August 2007, NATI announced that they intended to double their field salesforce by 2010.  NATI’s traditional business has primarily been a book-and-ship business with average order sizes in the $2,000 range.  The field sales force, however, has proven instrumental in leveraging NATI’s new platforms like cRIO into “system level sales” that have average order sizes of $20,00 and above.  There is significant opportunity to win more system level sales, but these require a more focused sales effort which is why NATI is aggressively expanding its field sales force. 

 

The sales department recognized the potential of these new platforms years ago and told senior management that they could sell more product if they had more resources.  Dr. Truchard, however, had the foresight to know it would be best to postpone a large increase in the field salesforce until the new platforms were more mature.  He also appreciated the degree to which he was already stretching his shareholders through the large R&D increases.  So the aggressive ramp in field sales signals that NATI’s new products are ready for prime time.  I am expecting the doubling of the field sales force and an eventual improvement in the industrial economy to lead to a moderate acceleration in NATI’s revenue growth in coming years.  This could prove conservative.  A hint of the company’s potential came on the 1Q08 conference call when Alex Davern casually mentioned that 1Q08’s revenue growth of 12.4% in such a weak environment was a milestone on their path to becoming a “multi-billion-dollar company.”  They had previously never talked about a goal beyond $1BB.

 

For the last six years, management has set the company’s long-term operating margin goal at 18%.  As they have made tremendous progress towards this goal over the last three years, it has become evident that this goal was not only attainable, but that they could probably exceed it if they moderated R&D and S&M growth.  In August, 2007 at their analyst day, however, management indicated that they didn’t expect OM to exceed 18% through 2010.  Given the new plans for the sales force, I interpreted this to mean that they were going to keep investing aggressively through 2010, but that OM expansion beyond 18% after that was possible.  I am modeling modest margin expansion beyond that, but you’ll notice that keeping R&D% and S&M% near their current levels implies enormous absolute increases in these expenses relative to what NATI has done in the past.  This seems unlikely given their mere size and the cadence of productivity gains that should be playing out at NATI over this time.

 

NATI has a unique corporate culture and the company regularly appears on Fortune’s list of “The 100 Best Places to Work.”  One of the ways that management maintains this culture is by hiring engineers directly out of college.  One drawback to this, however, is that these inexperienced engineers are only marginally productive during their first year of work, and they don’t hit their stride until their fifth year at the company.  NATI began to crank up R&D headcount in 2001 which means that the earliest of those new hires really started to come into their own in 2006.  It also suggests that the productivity of their R&D effort should continue improving which could offer the opportunity for operating leverage in this area. 

 

Field salespeople, by contrast, have a much quicker payback and become productive in 12-18 months.  NATI has begun moving a crop of 2007’s application engineers into field sales in 2008, which implies an initial payback in 2009-2010.  This could continue through 2011 as each year of new field sales people become more productive. 

 

My model is shown below: