There has been no shortage of "bubble callers" during the stock market boom of 2013. Many of those Chicken Littles are either naïve or missing the relative value of equities when compared to other asset classes. However, 3D printing stocks appear to be in a bubble, driven by Wall Street hype, good stories, and a scarcity of perceived growth stocks. Earlier this year, a range of financial publications, from Barron's to Reuters to Morningstar, highlighted the exuberance of 3D stocks. The market leaders, Stratasys (SSYS) and 3D Systems (DDD) are up 97% and 216% respectively over the last 12 months, despite earnings outlooks that are relatively unchanged. But it isn't the leaders that market observers should focus on to identify a bubble. It is the "me-too" companies that have flimsy fundamentals, relying instead on stories created by bankers and analysts to capture the minds of speculators. Amazon.com was overvalued in 1999, but it has undeniably maintained its leadership role transforming e-commerce. Webvan and Pets.com are much more indicative of the low quality "me-too" companies that come public at asinine valuations in the latter stages of a bubble, only to crash when fundamentals matter again. We believe recent IPO Voxeljet (VJET) will be seen over time as the Webvan of the 3D printing bubble. While the valuation of Voxeljet is beyond any rational framework, our negative stance is driven predominantly by the misrepresentations of its selling model and other forensic issues. It is our belief that Voxeljet has at least 75% downside to $16.79 per share.
Unlike the dot.com blow-ups, which were in many cases IPOs based on unproven PowerPoints, Voxeljet actually has historical results indicative of an unimpressive financial and operating model. With a market cap comfortably over $1 billion, and 2013 sales guidance of only €11 million, Voxeljet trades at a dot.com-like 70x revenue. Voxeljet's stock is up nearly 500% since pricing its IPO last month and has doubled in November alone. At its most recent price of $65.00, it is the most expensive 3D printing stock, despite possessing the slowest growth profile of its comparable set, and a myriad of questionable practices we will discuss in this report.
To be clear, we do appreciate the 3D hype and in no way are we suggesting the potential for 3D manufacturing is overstated. Stratasys is a fine company, while 3D Systems has built impressive scale with 40+ acquisitions over the last five years. Stratasys and 3D Systems will probably be around in a few years should the 3D "revolution" materialize. However, Voxeljet, like Webvan thirteen years ago, appears to be nothing more than a story. In the words of 3D Printing Industry, the leading global resource for 3D printing, Voxeljet appears to be a bubble, with significant financial risk for traders and investors.
Below we discuss many topics, including our belief that Voxeljet completely misrepresented its third quarter printer sales. Based on disclosures found in its 11/14/13 earnings 6K, we believe that Voxeljet did not sell one printer in the third quarter at arm's length. While not fraudulent per se, Voxeljet management failed to disclose on its earnings call that all three printers sold in Q3'13 were either financed by Voxeljet, or booked as revenue with future payments to be made in services. The unorthodox sales all appear to have occurred despite correspondence just one month earlier with the SEC stating Voxeljet does not rely on vendor financed sales. Voxeljet management also seemed to misrepresent its change in other operating expense on its Q3'13 call in response to a question from a sellside analyst. Given the potential severity of these issues, we emailed management after its earnings call but have yet to hear back with nearly a week having passed (we tried calling IR as well). Should we hear back from management, we will gladly provide the response in a new post or InstaBlog.
We believe Voxeljet has extremely limited intellectual property and R&D, as evidenced by expenditures that do not seem to reconcile with publicly disclosed headcounts. Further, it appears competitor ExOne (XONE) can impose restrictions on Voxeljet manufacturing based on a 2003 patent agreement, which the SEC requested Voxeljet disclose in its Registration statement. We also believe that Voxeljet's unusual past payments to its CFO's company need to be clarified, while other transactions are more reminiscent of the U.S.-listed Chinese debacles of the last five years (including paying the CFO for his art collection at Voxeljet's headquarters and purchasing materials from the CEO's brother). We believe it is only a matter of time before the retail-driven momentum runs its course. When Voxeljet finally trades on fundamentals, it will likely trade at a steep discount to its 3D comparables based on many of the issues we discuss herein. Should Voxeljet simply trade in line with its 3D comparables, the stock price would be $16.79, representing 75% downside. If Voxeljet traded at a more likely 25% discount to its comparable group, the stock would face 79% downside.
1) Voxeljet is a 3rd tier manufacturer
Voxeljet has very little mind share in the real world of 3D printing. According to a Discern survey released on 11/15/13, Voxeljet was not even one of the top eighteen 3D printing companies listed with an Intent to Buy. We would encourage anybody interested in the 3D printing space to contact Discern directly. Survey participants were asked a number of questions, including which printers/manufacturers and service bureau/paid parts have they used, as well as questions about materials. Discern's survey was quite extensive and the results serve to confirm our belief that Voxeljet is not relevant to the 3D printing landscape. Per Discern's survey, respondents in the 3D printing community listed the following 3D companies they intend to use:
Stratasys, 3D Systems, Materialise, Blue Printer, Solidoodle, Botmill, ExOne, Fortus , MakerBot , Solidscape, EnvisionTEC, Ultimaker, Printrbot, UP!, Shapeways, Ponoko, Solid Concepts, and Sculpteo.
We counted eighteen different 3D printing companies/service bureaus, yet Voxeljet did not appear one time in the published survey.
It is not just prospective buyers ignoring Voxeljet. The publicly-traded 3D competitors also seem to have very little concern for Voxeljet. In the most recent annual reports from Stratasys, 3D Systems, ExOne and Arcam, there were twelve companies listed as competitors across the four filings (several have since been acquired by SSYS and DDD):
· 3D Systems Corporation
· EOS Optronics GmbH
· EnvisionTEC GmbH
· Z Corporation
· Stratasys Inc
· Solidscape, Inc.
· Objet Ltd.
· Solid Model Ltd.
· Concept Laser
Voxeljet was not listed as a competitor in one filing from any of the four publicly-traded 3D printing companies. Perhaps this is because Voxeljet does not look, or act, like a 3D printing company. In the most recently completed calendar year, Voxeljet only grew its revenues by 20%. Voxeljet's growth significantly lagged the 30% growth experienced by the 3D industry, as reported by Wohler's Associates in its oft-cited 2012 3D Industry Report. This paltry growth is even more disappointing considering the law of small numbers. In 2012, the company only registered €8.7 million of revenue. The number of printers Volxejet sold in 2012 (6), as well as the lack of growth in printers sold (0%), is completely incongruous with a dynamic 3D printing company. In 2011, Voxeljet sold 6 printers. In 2012, Voxeljet also sold just 6 printers. Through the first 9 months of 2013, Voxeljet has again only sold 6 printers (including one used printer). Further, we will show later that at least three of the printers sold in 2013 required Voxeljet to finance the purchases or accept services in lieu of cash. We were able to obtain a picture of Voxeljet's facility in Friedberg, Germany. It is worth noting that their facility is only 16,000 square feet, which is not much larger than the typical floor of a New York high-rise. As can be seen below, the printing facility illustrates the extreme disconnect between a $1 billion market-cap company and Voxeljet's 16,000 square foot warehouse. The picture below is reminiscent of Telestone (TSTC) and AutoChina (AUTC), two of the Chinese RTO frauds we wrote about in the past (TSTC has declined 98%, while AUTC is being sued by the SEC for fraud).
2) Voxeljet appears to have sold ZERO 3D printers in Q3'13 at arm's length?
Based on analysis of Voxeljet's public filings, we believe there is a dirty secret that has been suppressed by management, and conveniently ignored by the sellside analysts at the firms that banked the Voxeljet IPO. Unlike other public 3D printing companies, we believe Voxeljet had to provide financing to sell its printers. We also believe Voxeljet management mislead the Securities and Exchange Commission approximately one week prior to its IPO.
On page 51 of Voxeljet's Registration statement is the brief disclosure, "On two occasions, we have provided loans to customers to cover the purchase price of a 3D printer. These loans have a grace period of six to nine months after which they are to be repaid on a monthly basis." If Voxeljet has indeed been supply constrained as management stated on its third quarter earnings call, then why exactly would they need to sell printers to customer with effectively zero down for six to nine months? Further, the company lists a handful of multi-national customers in its Registration statement, such as BMW, Daimler, Ford, and Porsche. In fact, several of the sellside initiations highlighted these customers as evidence of Voxeljet's relevance. We believe customers of this nature, which definitely look very good in a Prospectus, would not require vendor financing to buy a 3D printer from Voxeljet. Apparently, the SEC shared our concern. While we do not have access to Voxeljet's confidential Registration statement under the JOBS Act, we do know the original filings resulted in a series of questions from the SEC on its use of vendor financing. In the September 6, 2013 correspondence, the SEC remarked:
We note your disclosures on page F-16 and page 51 regarding two loans you have granted to customers. We note your discussion on page 51 that you concluded that the conditions had been fulfilled and that recognition of amounts as revenue was appropriate. Please tell us more about the terms of these loans. In this regard, we note that on page 51, you state that these loans provide a grace period of six to nine months after which they are to be repaid on a monthly basis. However, you state on page F-16 that repayment of the 2010 loan did not commence until November 2012. Please clarify for us the repayment terms of these loans, including when repayment begins. Discuss how you considered the delay between the time you granted the loan and the time payments under the loan began in concluding that it was probable that the economic benefits associated with the transactions would flow to you.
It appears the original disclosures relating to the loans were removed from Voxeljet's final Registration statement. We have significant concerns that Voxeljet is using its financing programs to potentially overstate revenue (an issue we will address below as it pertained to the third quarter). In its September 17th response to the SEC, Voxeljet confessed that one of the loans was actually not repaid in full. Instead, Voxeljet repurchased the printer, refurbished it, and then sold it to a bank - only to lease it back from the bank. The printer was then subleased to another customer - generating revenue again. We can't emphasize enough that these types of shenanigans are inconsistent with a healthy company. Further, the detailed explanation below can only be found in the SEC correspondence, because the background in the Registration statement has been truncated to just two sentences in a note on page F-41:
In December 2010, the company granted a loan in the amount of €340,000 to a customer. The terms of the loan required payments to begin in October 2011, with monthly payments of interest and principal thereafter. The final payment of the remaining balance on this loan was to be due in September 2016. In February 2013, as a result of the individual who managed this customer's business accepting an offer of employment with the company, and, thus, deciding to discontinue the customer's business, the company identified a new customer for the printer and repurchased the 3D printer subject to the loan at a price equal to the then-outstanding unamortized loan balance and extinguished the loan. The 3D printer was then refurbished and sold to a bank in a sale and leaseback transaction and subsequently subleased by the company to the new customer under the terms of an operating lease. The company does not believe the subsequent repurchase of the 3D printer calls into question its initial assessment that it was probable that the economic benefit associated with the sale to the customer would flow to the company. The company discloses the repurchase of this 3D printer in Note 3 to its interim financial statements on page F-41 of the Prospectus.
While vendor financing may be costly for a company with minimal capital stability (Voxeljet prior to its IPO), there is nothing inherently wrong with giving away product for some period of time before a customer needs to make payments. However, we believe based on its third quarter results, Voxeljet may have misled investors and the SEC regarding its propensity for providing vendor financing. In its final correspondence with the SEC prior to the IPO, Voxeljet management stated unequivocally that it "does not ordinarily offer loans to current or prospective customers to cover the purchase of its 3D printers." The language of the response from October 7, 2013 states the loan program is indeed out of the ordinary.
We believe Voxeljet was highly misleading in its response letter, because in the quarter completed September 30, 2013 (Q3'13), Voxeljet appears to have provided direct financing for 1/3 of its printer sales and indirect financing for the other 2/3 of its sales. In the body of Voxeljet's Q3'13 earnings release, there is no reference to the financing program, nor was there any disclosure in its Q3'13 earnings call. However, two unassuming sentences in the detailed footnotes suggest Voxeljet was unable to sell any printers in Q3'13 without providing assistance. The two sentences are (emphasis ours):
· In September 2013, another loan was granted to a customer for the purpose of financing their acquisition of a 3D printer in the amount of kEUR 678.
· In September 2013, the company recognized revenue of kEUR 868 on the sale of two new 3D printers to a customer in exchange for consideration consisting of kEUR 630 cash and kEUR 238 in research services to be received.
In the third quarter, it appears Voxeljet sold one printer via an on-balance sheet loan, and two other printers with 27% of the purchase price representing future research services the purchaser agreed to provide in the future (presumably in lieu of cash payments). Further, it appears all three of Voxeljet's printers were sold in the last month of Q3'13 (September), and without such sales, Voxeljet would have missed Q3'13 by approximately 50%. As such, Voxeljet sold ZERO printers in the third quarter for cash at its list price, despite declaring to the SEC that this process is not "ordinary." We have been unable to find any disclosure regarding the nature of the "services to be received," but find this transaction extremely troubling in light of related party transactions discussed later.
3) Voxeljet's low-tech profile in a high-tech industry
We believe Voxeljet is a bottom-tier company with minimal mind share in the 3D printing industry. Not only is this thesis supported by the Discern survey and lack of one reference as competition from other 3D names, but the company's own financials support the thesis. Voxeljet disclosed in its public filings that it only had "24 employees working in our research and development department." While this is a razor-thin number of engineers, we believe the quality of the engineering staff may be well below average based on compensation estimates.
In the third quarter just reported, Voxeljet only spent €447,000 on research and development. In 2012, Voxeljet disclosed 44% of COGS were personnel costs. Presumably we can use a similar ratio to determine what percentage of operating expense is attributable to compensation (vs. non-comp costs). The Q3'13 R&D expense would equate to an annual run-rate of €1,788,000. Assuming the 44% compensation ratio holds, then Voxeljet is only paying its 24 engineers €32,780 on average per year. According to Salary Explorer, the average engineering salary in Germany is €44,616, meaning Voxeljet pays its engineers 27% less than the average engineering salary in Germany. Comparing Voxeljet's R&D expense to its purported 3D peers is even more striking. ExOne is often considered the small, third-tier U.S. 3D printing company. Yet XONE spends nearly 3x more on R&D than Voxeljet. Further, Voxeljet spends just 0.18% of its market capitalization on R&D, well below the peer group. Therefore, Voxeljet is significantly under-spending its peer group, or its stock is wildly overvalued (or most likely both).
The infinitesimal R&D spend also fits with Voxeljet's poor IP positioning, which has forced the company to license critical intellectual property from competitors. Based on discussions with ExOne, it is our understanding that ExOne can place restrictions on Voxeljet's ability to manufacture machines utilizing certain processes and materials. The SEC deemed its IP business risk material enough to ask Voxeljet for more disclosure.
Source - SEC correspondence disclosures
In fact, Voxeljet seems to acknowledge this business risk, although we are unsure whether it was the SEC's request that led to the disclosures (Z Corp was bought by 3D systems):
We have license agreements with respect to certain intellectual property that is important to our business with both Z Corp and The ExOne Company, or ExOne, that impose restrictions on our use of certain intellectual property.
Under this agreement, we are permitted to use machines and provide services relating to these technologies, but not to make or sell machines utilizing these technologies without ExOne's consent, although ExOne has an obligation to consent if the machines do not compete with products engineered, manufactured or sold by ExOne or its affiliates.
While our rights are restricted regarding use of certain binders and sand-based casting methods in 3D printers under our agreements with ExOne, we believe these restrictions will not materially impact the growth of our business.
Illustrating the inefficiency in the market, ExOne will do 300% more revenue than Voxeljet, grows significantly faster and owns critical IP upon which Voxeljet is dependent. Despite these factors, Voxeljet's enterprise is currently valued at a 25% premium to ExOne (as a side note, we are not making a positive argument on ExOne - it too is obscenely valued and has missed and/or guided down every quarter since it came public).
4) The Use of Sale Leasebacks and a Not-So-White Fib
Prior to its IPO, Voxeljet had been in such a precarious position that the company has been forced to populate its service center with printers owned by third parties. Simply, Voxeljet would sell a 3D printer to a bank (or other party) and then lease the printer back from the bank for service center usage. Like so many of the flags in the Voxeljet story, we have been unable to find sufficient discussion in sellside reports of this tactic. Fortunately, Voxeljet was required to disclose the imputed interest rate, as well as minimum lease payments in its Registration statement. Based on the unamortized interest expense, as a percentage of the PV of minimum future lease payments, we can compute an effective interest rate. Based on our analysis, Voxeljet is paying an effective 23% aggregated interest rate for the right to use its own printers via a sales/leaseback transaction.
While the usage of sale/leasebacks is typically associated with weaker capital structures, it is perfectly legal and acceptable. However, management misrepresenting the nature of sale/leasebacks when responding to a question on an earnings call is not acceptable. On its third quarter earnings call, management was asked by a sellside banker (we cannot tell if he is a banker or sellside analyst) about its increased other operating expense.
Analyst: Within operating expenses, you have a swing from a small income to a small expense in the other operating expenses not the financial result but the one that's above operating income. I was just wondering what caused the swing?
Rudolph Franz (CFO Voxeljet): It was mainly driven by let's say the increase in expenses by establishing the organization for the next growth okay. This is mainly an investment in that area and I think all the rest is I think it will grow with the revenue but on a very low basis that's our plan actually.
We believe Mr. Franz's answer is a half-truth at best. Voxeljet clearly states in its Registration statement that the impact from sale/leasebacks is accounted for in other operating expenses. Mr. Franz failed to include this detail in his answer, which we believe is indicative of several other questions we would raise relating to Voxeljet's management.
5) Unconventional Related Party Transactions
We are not willing to accept related party transactions as benign, simply because they are disclosed. As it relates to Voxeljet, we have found several related party dealings that may be indicative of a culture inconsistent with acceptable U.S. corporate governance.
First, Voxeljet paid €150,574 to a company called Franz Industriebeteiligungen AG, which was an affiliate of Voxeljet owned by Mr. Franz. We have been unable to find additional information other than payment was remitted to a company owned by Mr. Franz and members of his family for "advisory services." According to public filings, Mr. Franz was paid this fee prior to becoming CFO of Voxeljet. The SEC also found discrepancies between payment amounts to its CEO that were originally provided by Voxeljet. These discrepancies have since been remedied according to Voxeljet's response, and the erroneous confidential filings on which the SEC is focused were not released to the public.
Additionally, Voxeljet has been paying its CFO, Mr. Franz, for personal paintings that have been on display at Voxeljet facilities. While the remitted payments do not appear to be significant, nevertheless Voxeljet is paying its CFO for his personal paintings.
Finally, we are extremely troubled by the fact Voxeljet has been buying materials directly from Schlosserei und Metallbau Ederer, a metal workings company owned by Robin Ederer, the brother of Voxeljet's CEO. A Google search for the Schlosserei und Metallbau Ederer facility responsible for Voxeljet's COGS shows what appears to be a small lot in a residential neighborhood.
Source - Company filings and Google Maps
6) Relative Fair Value is $16.79 and perhaps much lower
It is hard to fault the sellside analysts for promoting Voxeljet. It is their job to help banking, even if it means creating unprecedented methods to value Voxeljet. For example, Cowen uses a DCF that is anchored by an unheard of 10% terminal growth rate, as well as a target price based upon a multiple of 100x 2017 earnings. We believe it would be wise for retail investors to proceed with caution if they are relying on any of the fundamental work of the covering analysts. For example, the Citi analyst has a Taiepi address and covers only three stocks, none of which are based in Asia Pacific. The Cowen analyst covers an eclectic mix of banking clients, including Uni-Pixel (UNXL) which he initiated around the Cowen secondary when UNXL was $36.20 (his $46 price target looks a bit aggressive today). He also covers questionable biofuel stocks such as KiOR (KIOR), Solazyme (SZYM), Amyris (AMRS), Gevo (GEVO), as well as bankrupt SunTech (OTCPK:STPFQ) and Ascent Solar (ASTI). The Stephens analyst covers XONE and Voxeljet, but does not cover the two most important companies in the 3D space, SSYS and DDD. Again, this is not a criticism of the analysts. We simply believe they face incredible challenges presenting a balanced picture.
Who gets left holding the bag when Voxeljet ultimately reflects its third-tier status? Voxeljet management sold 900,000 shares less than one month ago. In fact, Voxeljet management not only sold 900,000 shares on the IPO, but their insider shares also represented 100% of the 975,000 greenshoe that was sold. Their sale price on 1,875,000 shares was $12.09 net of the underwriter's discount.
So our guess is the retail investor recklessly bidding up Voxeljet gets stuck holding the bag. According to Bloomberg, for the week ending 11/15/13, in which Voxeljet's stock rose almost 50%, over half of the advertised volume was attributable to three heavily retail-oriented brokers (NITE, UBS - via Schwab, and Fidelity Capital Markets).
We believe Voxeljet shares have at least 75% downside in the near-to-intermediate term. Should we begrudgingly assume the other publicly-traded 3D stocks are fairly valued, and Voxeljet trades in line with this peer group, then Voxeljet shares would be worth approximately $16.79, which is 75% below its recent share price. However, this assumes Voxeljet trades in line with its faster growing comp set, AND it assumes the comp set is not overvalued. Should Voxeljet trade at a discount to its comp set, or should the comp set correct, then Voxeljet could see downside well in excess of 75%. For example, if Voxeljet were to trade at a 25% discount to its peer group (which would still be 11.6x sales), the stock price of VJET would decline by 79% to $13.34 per share. We believe the damage is a matter of when… not if.