"Fads are the kiss of death. When the fad goes away, you go with it."
- Conway Twitty
Proto Labs Inc. (PRLB) operates a very easy-to-understand business. The company is a low-volume manufacturer of injection molded and machined parts. Approximately 75% of its revenue is generated in the United States.
Injection molding is simply producing parts by injecting material into a mold. Machining refers to various processes in which a piece of raw material (e.g. metal) has material removed in a controlled manner until an end part is produced.
Injection molding is not what we would consider "new" technology. The first patent on injection molding was filed in 1872:
You will not be shocked to learn that over the past 141 years the technology has become well established and is used in a large variety of manufacturing activities.
Similarly, machining and machine tools have been around since the 1700s, with more "modern" CNC, or "Computerized Numerical Control" machines used since the 1960s. Since the 1970s CAD, or "Computer Aided Design" software programs have been used to design products and produce the information needed by CNC machines to produce produces. Wikipedia provides a decent overview of the process:
In modern CNC systems, end-to-end component design is highly automated using computer-aided design (CAD) and computer-aided manufacturing (CAM) programs. The programs produce a computer file that is interpreted to extract the commands needed to operate a particular machine via a postprocessor, and then loaded into the CNC machines for production. Since any particular component might require the use of a number of different tools - drills, saws, etc., modern machines often combine multiple tools into a single "cell." In other cases, a number of different machines are used with an external controller and human or robotic operators that move the component from machine to machine. In either case, the complex series of steps needed to produce any part is highly automated and produces a part that closely matches the original CAD design.
So What Exactly Does Proto Labs Do?
Proto Labs refers to itself as "the world's fastest provider of CNC machined and injection molded parts." I checked with the folks at Guinness World Records and unfortunately they do not track this as a category so we will have to accept the company's word for it.
Essentially the company buys and operates CNC and injection molding machines and specializes in low-volume or one-off production runs. Customers send in their computerized CAD files, Proto Labs machines or molds the part as required, and ships it out to the customer, all for a fee.
What Makes Proto Labs Different?
Small machine shops and injection molding companies are literally everywhere. As an example, below is a Google Maps search for "machine shop Chicago." Try to count the dots, or even better replicate the search in your own city.
The prospectus confirms this fact, noting "there are thousands of machine shops and plastic injection molding suppliers worldwide" and lists these firms, as well as captive in-house services of larger companies, and alternative manufacturing vendors such as 3-D printers as competition (more on 3-D printing later).
So what makes Proto Labs different from the thousands of small, medium, and large machine shop and injection molding services out there?
The company promises "Real Parts, Really Fast" (unlike those virtual parts we all hear about) and has trademarked the term. The company lists five "Competitive Advantages" in its IPO prospectus:
- Advanced Proprietary Technology
- Turnaround Speed
- Operations Designed for Low-Volume Manufacturing
- Marketing and Sales Strength
- Deep Industry Knowledge
I don't know about you, but when I read the above list apart from "Advanced Proprietary Technology" (which we will get into later) it seems like a pretty lame list that some junior flunky at an investment bank cooked up to fill out a page.
I am going to dismiss "Marketing and Sales Strength" and "Deep Industry Knowledge" as virtually meaningless bumph (for example the company touts "quickly following up on leads and quotation requests") and focus on the other 3 more meaningful stated advantages.
"Turnaround Speed" is important for many customers. Proto Labs claims that "our competitors often require several days just to generate a price quotation." However it seems that speed is a focus for many of Proto Labs' competitors also. A Google Search for "rapid prototype service" yields over 5 million hits, and legions of companies promising fast turnaround times.
"Operations Designed For Low-Volume Manufacturing" isn't really a competitive advantage, as the company doesn't even compete in the high-volume/production line manufacturing market. It's fair to say that every company competing in the low-volume markets has "operations designed" for that market. Next.
That leaves "Advanced Proprietary Technology" as the linchpin of PRLB's potential moat. So what do they mean by this? Again to the prospectus:
Our proprietary software largely automates the areas of manufacturability analysis and feedback, price quotation, order submission, mold design, tool path generation, mold or part manufacture and production management.
The company provides this visual for its "technology-enabled processes:"
The top half of the above chart really boils down to a web-based quotation and order tool. Customers can upload any standard 3-D CAD file, it is analyzed (e.g. by complexity or number of cuts) and based on that analysis a quote is spit out. This doesn't strike me as anything particularly novel or not easily replicated. In fact, a quick tour of some other "rapid prototyping" companies confirms this - below is from "Quickparts," a private Proto Labs competitor that was the first Google hit on the earlier search:
QuickQuote is the patented geometric analysis and instant quoting engine behind Quickparts (Patent #7305367). Traditional methods of manually quoting custom manufactured parts take too much time, leave too much room for error, and are often a bigger hassle than they're worth.
QuickQuote streamlines this process by analyzing your 3-D CAD data (based on volume and geometry), then producing an instant customized quote in the process and quantity you specify. Quoting and buying can take as little as 2 minutes!
Change your quote to suit your needs, and see the project's price change instantly. QuickQuote makes it easy.
The bottom half of the "technology-enabled process" chart basically tells the machines what to do. A simple search for "CAD CNC Software" and "Injection Molding CAD Software" yields a vast array of software solutions - not surprising for technology that has been around for decades. The software category is known as Computer Aided Manufacturing ("CAM"), with CAD and CAM often referred together as "CAD/CAM." Basically CAM software converts a CAD file into industry standard "g-code" for that CNC machines understand. For a deeper dive on the type of software that is available off the shelf in this market a comprehensive list of vendors and summary of their capabilities can be found here.
As a final point of validation for my opinion that the Proto Labs "technology" is really some streamlining and process automation, I had a look at the company's patent portfolio. Interestingly, the company's IPO prospectus states that Proto Labs has been issued 12 patents (15 per the latest 10-K). However a search at the U.S. Patent And Trademark Office only lists 9 patents:
You can see that the majority of the patents are related to "automated quoting." A look through some of the claims yields such technological breakthroughs as "receiving a CAD file over a global communications network" and "generating a quotation based upon the identified and located material removal stems."
My personal conclusion is that Proto Labs is simply an old-timey metal bashing and molding company with a web front-end and some process optimization.
Summary of financial performance of Proto Labs is provided in the following table:
As you can see, the financial performance has been good. While the business stagnated in the 2008/2009 period, it has enjoyed a few solid years of top line and bottom line growth. Revenue increased at a compound annual growth rate of 29.8% from 2008 to 2012.
IPO and Share Price Performance
Proto Labs originally filed an IPO prospectus on July 22, 2011.
The IPO attempt seemed to be a dud, with the rest of 2011 passing without completion. Then a wonderful thing occurred - the stock market started going mad for a technology known as "3-D Printing" with its poster child, 3D Systems Corporation (DDD) taking off in early 2012:
Below is a chart from Google Trends that shows Google searches for "3-D Printing" and how the hype was just getting rolling as Proto Labs was attempting to go public.
Even though Proto Labs has nothing to do with 3-D Printing (an "additive" technology that allows virtually anyone to "print" a solid 3-D object at very low cost for both the machine and material), and actually lists 3-D printing as a competitive threat, PRLB was lumped in with 3-D Printing companies such as DDD and the IPO was ultimately concluded in February 2012 at a price of $16 per share at a healthy 21x P/E multiple. However despite the $16 price (up from an expected range of $13-$15), the stock soared 81% on the first day of trading to close at $29.
The company seems to send out mixed signals on 3-D printing. In the prospectus, PRLB refers to 3-D printing as a competitive threat:
"We compete for customers with a wide variety of custom parts manufacturers and methods. Some of our current and potential competitors include captive in-house services, other custom manufacturers, and alternative manufacturing vendors such as those utilizing stereolithography, selective laser sintering, fused deposition modeling and 3-D printing."
Yet at the Needham & Company, LLC Growth Conference in January PRLB's CEO Bradley Cleveland stated:
"We believe we are very complementary and very synergistic with 3-D printing."
Perhaps the juicy valuations ascribed to 3-D printing companies (for now) have Mr. Cleveland excited about some of those "synergies" wearing off on PRLB's stock price.
So what has happened to PRLB since going public just over a year ago at $16? The stock has skyrocket on the back of the 3-D printing mania, tripling to approximately $48 today. The screenshot from Motley Fool below demonstrates how the market is incorrectly lumping Proto Labs in with 3-D Printing companies (which are arguably in their own bubble, but that is another article...)
Here is a chart for PRLB since the Feb 2012 IPO:
Note the chart above starts at $29 (the closing price on the day of the $16 IPO), making it look slightly less outrageous.
Now let's move on to PRLB's valuation:
It's obvious from the data above that investors are paying exceedingly high multiples (over 50x earnings, 9x sales, and a whopping 49x free cash flow) on the expectation of future growth.
So how much growth is baked into the shares of Proto Labs at the current price? To illustrate, I ran a relatively simple model assuming that 10 years from now PRLB will trade at a market multiple of 15x, giving a very long growth runway, and discounting that future value back at various discount rates to arrive at a current value. The table below shows what kind of annual growth in earnings needs to occur compounded for the next 10 years just to justify the current price, using discount rates ranging from 7% to 13%:
Only the numbers highlighted in green show a value above the current share price. As highlighted even a Herculean 10-year CAGR of 15-20% yields valuations in the $16-20 range, or 60-70% below the current share price. Using a 10% discount rate, the shares at $48 are currently discounting a 10-year earnings CAGR of approximately 38%.
How reasonable are such growth rates? Below is a graphic illustration of PRLB's Net Income over the next 10 years at a 38% CAGR:
Put simply, PRLB must deliver the above performance over the next decade just for investors (paying $48) to earn a modest (e.g. 10%) return.
How realistic are those growth rates over the next decade? In my view they are ludicrous. As a point of reference, according to Bloomberg, Google (GOOG) has increased its earnings over the past 5 years at a 20% rate. Facebook (FB) is expected to grow earnings by 37% next year (2014) per consensus estimates. Proto Labs is not Facebook.
But don't take my word for it. In the recently filed 10-K Proto Labs management flat out tells you that it can't sustain these growth rates:
Although our revenue has grown from $44.4 million for the year ended December 31, 2008 to $126.0 million for the year ended December 31, 2012, we likely will not be able to maintain our historical rate of revenue growth.
That 2008-2012 growth rate is 29.8%, and management is telling you that it will come down. Yet I calculate the market is discounting very long-term growth rates far in excess of what management is telling you they can't deliver. You can see in the table above that even if they could meet those growth rates the stock could suffer a 40% swoon.
There is evidence that growth is already slowing. According to Bloomberg Q1 consensus revenue estimates are for $36.9M in revenue, which would only be a 23% increase over Q1 2012 and consistent with management's cautionary comment in the 10-K about growth rates coming down.
I also believe that the 2008-2012 growth rate is exaggerated as 2008 was in the depths of the financial crisis when industrial production was falling off a cliff. The cyclical rebound has been a nice tailwind but now that production is back to 2007 the growth rate should slow from here - coupled with a struggling Europe = growth headwinds for companies like Proto Labs. Any meaningful traction in 3-D printing may only add to the growth pressure as a competitive substitute for a meaningful subset of PRLB's business.
Markets do not like uncertainty and variability. Unfortunately Proto Labs, offering fast turnarounds on customer orders, has limited backlog or revenue visibility, and thus is highly vulnerable to a growth disappointment (especially when the market is pricing in 10-year EPS CAGR of nearly 40%). A hint of this exposure can be found deep in the 10-K where the company disclosed that for 2012, sales to existing customers in Europe actually declined by $3.7 million.
Link Between Growth and Reporting Changes?
Interestingly, up until Q3 2012, Proto Labs broke its revenue into 3 geographic locations: United States (79% of Q3 Revenue), Europe (18% of Q3 Revenue), and "Other" (3% of Q3 revenue). However 3 months later in the 10-K filing the company changed its reporting and lumps Europe and "Other" into a catch-all bucket they call "International." The company also broke out "reportable segments" in its financial statements up until Q3, but the term has now been dropped:
New Reporting (as of Dec 31 2012 10-K)
This is interesting. Recall the note above that sales to existing customers in Europe declined by $3.7M in FY 2012. If we go back to the Q3 2012 10-Q filing, we see that number for Europe as:
a period-over-period decline of approximately $1.7 million in sales attributable to approximately 825 existing customer companies
Interesting. For 9 months in Europe revenue from existing customers declined by $1.7M, throw in Q4 and that number more than doubles to $3.7M. Meaning that for FY 2012 over half of the decline from existing customers occurred in the final quarter.
Goodness knows I'm not a skeptic. But if I were I would highlight this comment in the 10-K related to the above segment reclassifications:
"The Company previously reported segments on a geographic basis. In 2012 as the Company's operations have evolved, the CEO, who is the Company's chief operating decision maker, began reviewing operating results only at the consolidated level, which is how the CEO makes decisions to operate the business and allocate resources. As a result, the Company has determined that it has only one operating segment, and has revised its historical segment presentation to align with the current year presentation. "
OK I am a skeptic. I find it incredibly convenient that just as Europe seems to be accelerating to the downside that the company's operations suddenly "evolved" and now the CEO only reviews operating results on a consolidated level.
"I've been working on a plan. During the exam, I'll hide under some coats, and hope that somehow everything will work out."
- Homer J. Simpson
Moving on. Adjusted EBITDA, Adjusted EBITDA, wherefore art thou Adjusted EBITDA?
In the company's IPO Prospectus, the company offers an "Adjusted EBITDA" calculation, which they define as net income + depreciation and amortization, loss on impairment of foreign subsidiary, and stock-based compensation. The company tells you why they provide this information:
"We have included adjusted EBITDA in this prospectus because it is a key measure used by our management and board of directors to understand and evaluate our core operating performance and trends, to prepare and approve our annual budget and to develop short- and long-term operational plans."
Hmmm. Sounds pretty important. But don't go looking for it in the 10-K. It seems it's suddenly no longer important.
Large Opaque Increase In R&D Costs
Another head-scratcher that has popped up is that Research and Development costs were gradually increasing up to the time of the IPO - running at approximately $1.59M in Q4 2011, and $1.66M in Q1 2012. Then the costs gapped up by 45% in a single quarter to $2.4M and have grown modestly again. The chart below depicts the increase:
Digging through the 10-Qs and 10-Ks for some kind of explanation, we see that most of this increase is from "professional services" for "outside development services." From the 10-Q from Q2 2012 (after the big step up):
"Research and Development. Our research and development expense increased $1.2 million, or 96.3%, for the three months ended June 30, 2012 compared to the same period in 2011 due to an increase in headcount resulting in personnel and related cost increases of $0.3 million, operating cost increases of $0.2 million and professional services of $0.7 million for outside consulting service."
Adding $2.4M in annual costs for "outside consulting services" in a single quarter on top of an R&D budget that was only $5.2M in the preceding four quarters seemed like quite a step up and worthy of an explanation. I therefore sent the CFO (listed as the primary IR contact) this email:
"Hello - I am an investor. I had a question regarding the R&D expense in the 10-Qs and 10-K - the R&D cost jumps up in a step fashion coincident with the IPO from the $1.5M range up to the $2.5M range per quarter - this is attributed to "outside services."
Can you please elaborate on these outside services -
are these former employees?
is this related to "Protoworks?"
Looking for some insight as that is a big jump in a quarter and it seems to be a structural move with outside services.
I thought these were pretty good questions. Unfortunately the response I received from the CFO was not particularly helpful:
"I am sorry but the only disclosure I will make regarding these additional expenditures are stated in our current filings with the SEC including our recent Form 10-K filed in March.
Good luck with all your analysis."
My mistake. I thought this was a public company.
So what have company insiders been up to lately? Why dumping enormous amounts of stock of course. Since the IPO lockup has expired insiders have unloaded nearly $200 million worth of stock (note the current market cap is $1.2 Billion) against purchases of zero.
One of the more violent sellers of stock has been CEO Bradley A. Cleveland. As of the IPO Mr. Cleveland owned 1,526,182 shares of PRLB, representing 6.6% of the company after the IPO. In just over a year he has sold approximately 1 million shares bringing his ownership (per the most recently filed proxy as of March 28 adjusted for April 1 sales) to 493,992 and less than 2% of the company. The largest sale, of 500,000 shares at $45.87 occurred just a few weeks ago for proceeds of approximately $23 million.
In summary if you are looking to short a company that is pricing in financial performance that management is telling you they can't deliver, has risen to very loft valuation levels on the back of hype about a technology that could actually eat the company's lunch, and has management and insiders voting with their feet by selling enormous amounts of stock, you may want to look at PRLB.
Similarly, if one is long an actual 3-D printing company like 3D Systems, perhaps a pair trade of shorting PRLB might be worth exploring.
As a parting thought, "good luck with all your analysis."
Additional disclosure: This article reflects my personal views only. I have a short position in PRLB stock. All data and calculations presented are accurate to the best of my knowledge but have not been vetted, checked, proofread, or independently verified. This article should not be relied upon for any purpose other than for entertainment. I welcome comments and or corrections.