By Daniel Holland
Since mid-December, many 3D printer stocks have been on a tear, with 3D Systems (DDD) up 30% and Stratasys (SSYS) up by as much as 26% before falling in the last few weeks. Still, both companies are pricing significantly ahead of our fair value estimates. Consensus earnings per share estimates for 2012 and 2013 have drifted higher for both names as investors have been quick to embrace the technology's potential but have also been willing to overlook some of the hurdles to growth. At present valuations, the market assumes an additional 7-9 percentage points of annual growth over our projections for both companies for the next 10 years (market growth assumption for Stratasys is 32%; for 3D Systems, 30%). In our opinion, little has happened during the past month to justify increasing our growth assumptions.
Soaring Valuations Based More on Hope Than Fundamentals
We trace the beginning of the recent run-up to a mid-December analyst meeting where General Electric CEO Jeff Immelt singled out additive manufacturing as one of the ways GE is using technology to discover cheaper ways to manufacture goods. While it is encouraging to hear the leader of a major global manufacturer speak to the merits of the technology, our recent discussions with GE's management team as well as other large manufacturers indicate that 3D printing is still several product generations away from being a viable input into select manufacturing processes. Scalability continues to hold the technology back, and Stratasys' management has admitted this is the core focus of the company's current research and development efforts at the moment. That said, Immelt's recognition of the technology adds an air of credibility to manufacturing as one of the three addressable markets; rapid prototyping and personal printing are the other two.
While the broader growth story has merit, we hesitate to assign a higher valuation for either name. While we like Stratasys' combination with Objet as it broadens the product portfolio, former Stratasys shareholders now own only 55% of the combined company, with Objet owning 45%. Stratasys shareholders were forced to give up a lot to obtain Objet, and we believe the equity dilution will stunt EPS growth for the next several quarters. Even after factoring in the equity dilution, we are still concerned about Stratasys' ability to drive the $10 million-$12 million per year of planned operating synergies. Corporate presentations given at the time of the announcement mentioned back-office synergies, which is in some contradiction to the company's actual decisions and execution since then (for example, the company intends to maintain two separate headquarters). Given Stratasys' lack of acquisition experience, we think it will be difficult to pull off a merger of this size without a hiccup, which may add volatility to earnings over the next few quarters.
3D Systems has recovered well from the negative attention it received in November from analyst reports questioning the company's true organic growth rate. The company identifies acquisition growth and organic growth, giving investors some insight into the composition of growth. After a newly acquired firm has been with 3D Systems for a year, 3D Systems rolls the associated revenue into the base revenue calculation. Research published by a firm with a short position in 3D Systems asserted that this presentation inflated true organic revenue growth. While we agree that it is difficult to gauge the source of revenue growth after the initial year, this is a complexity in analyzing any acquisitive company's growth and is not necessarily misleading. As a result, we see no problem with how the company reports organic growth, as it appears consistent with how other acquisitive firms communicate acquired versus organic growth.
The company's latest round of acquisitions appears aimed at improving development software, a key hurdle to broader consumer acceptance, in our opinion, because of the technical nature of most 3D design software platforms. While we think the current positioning of the company and its growth prospects are strong, the growth strategy to date still seems contingent on successful acquisitions and expanding the addressable market, which we need to see play out further before increasing our fair value estimate.
Rising valuations give us pause on both names, as the market's growth expectations call for growth that few companies have been able to deliver consistently. When we back into the market's growth rate using our discounted cash flow model, we find the market is assuming roughly 30% annual revenue growth for 3D Systems and 32% growth for Stratasys over the next 10 years. Having this strong growth directly fall to earnings is a challenging task particularly for Stratasys, in our opinion, given its higher gross margins than 3D Systems: 52.0% for Stratasys compared with 50.7% for 3D Systems. While we agree with the market's assumption that Stratasys has a better growth profile than 3D Systems because of its more concentrated exposure to the rapid prototyping market and more mature business model, achieving a compound annual growth rate north of 30% over 10 years is difficult to rationalize at this stage of the business cycle.
We link the growth rate of 3D printers to the broader $30 billion product lifecycle management industry, of which 3D Systems and Stratasys combine to make up a 1.2% share. Consistent with our expectations for other PLM firms like Dassault (OTC:DASTY) and Autodesk (ADSK), we expand the market just under 10%, though Stratasys and 3D Systems benefit from share going from 1.2% to 3.0% over 10 years. We think the companies can achieve share gains through new product introductions and higher machine utilization, leading to more consumable revenue.
Current share prices imply market share expands closer to 9% for 3D and Stratasys together under the same market growth assumptions in our model. If correct, the market assumes these businesses will be roughly 18 times bigger than they are today. While this is a possibility, we think it's a tall order for two firms just now finding their footing after a couple of decades of fairly benign growth. As a point of reference, Autodesk and Dassault are roughly 7% and 8% of the PLM market. While we are encouraged by the market's reception of the technology, it is too early to assign such expectations to the manufacturers. In our opinion, scalability barriers and the steep learning curve will impede broader market adoption.
Working Capital Growing, but Not Alarming Yet
Working capital management is one of our key concerns for companies experiencing strong end market growth. The cash operating cycle has improved from eight quarters ago at both 3D Systems and Stratasys, and we're encouraged that working capital is increasing in concert with sales, given the rapid growth of the top line of both companies. However, at Stratasys the bulk of the improvement has come from extending payable terms, and we caution that the cash-conversion cycle has started to move higher sequentially for both firms after steadily declining over the previous six quarters. After dropping from 80 days to 71 days, 3D Systems' cash cycle extended back to 81 days as of the third quarter of 2012, primarily a result of payable days dropping from 76 to 69 (although inventory and receivables have collectively grown 3.5 days over the same period as well). Continued difficulty damping inventory needs or extending payable terms could lead to outsize cash use for the company as it increases revenue further.
During the third quarter of 2012, Stratasys increased total inventory on hand more than 40% sequentially, to $32 million. The company opened distribution outlets in Asia and Europe and elected to fill both centers with finished goods inventory. Given that Stratasys has been operating out of a single distribution center in the United States and the imminent merger with Objet, management sought to develop its global distribution strategy to match growing demand. The company ramped production to build inventory, which had a positive effect on gross margins of around 2 percentage points in the third quarter thanks to better fixed-cost absorption. This benefit helped to offset some margin pressure coming from higher printer sales, which have lower margins than the corporate average.
In addition to the inventory build, the company also incurred $0.8 million of overhead costs in the quarter associated with the distribution outlets. While we recognize this as a cost of the global distribution strategy, we expected Stratasys to leverage Objet's global distribution to accomplish this. Continuing to operate from two headquarters on separate continents is probably not optimal for driving the $10 million-$12 million of synergies the companies cited when announcing the deal.
Sales Growth Not Quite Dropping Down to Cash Flow for Stratasys
The steady deterioration of Stratasys' cash margins also gives us pause at current valuation levels. The inventory build in the third quarter reduced cash margins roughly 5 percentage points; after adjusting for that use of cash, margins have still steadily marched down from 19% in the fourth quarter of 2010 to an adjusted 12% while 3D Systems' margins have remained roughly level, around 15%. The rapid growth in sales has yet to expand cash flow, as Stratasys continues to increase corporate expenses in line with sales while gross margins have deteriorated. To the extent that this trend persists, we think the market's expectations will adjust downward, since we believe the market is currently pricing stellar performance into the stock.
MakerBot Continues to Pose a Threat in Consumer Markets
MakerBot manufactures lower-priced printers and continues to invest in its product portfolio. After the successful launch of the $2,199 Replicator 2 printer in mid-2012, the company is coming out with the Replicator 2X in the first quarter of 2013. We recently visited MakerBot's New York storefront and viewed the printers and finished products. The Replicator 2 is a big step up aesthetically from earlier printers and features a build volume of 410 cubic inches compared with 167 cubic inches for 3D Systems' Cube and 125 cubic inches for Stratasys' Mojo. While the larger build volume is helpful in making larger objects, the Replicator is restricted to only one plastic type (and hence a single color) per build, consistent with the Cube and Mojo. Even though there are enough distinctions among the different models to make it difficult to draw comparisons and individual customer constraints may affect the decision, we see the Replicator 2 as a viable competitor even considering the price.
To date, household printers have been limited to trinket making, and an individual without 3D design skills depends on other designers for content. To facilitate this, MakerBot and 3D Systems have developed online forums such as Thingiverse (MakerBot) and Cubify (3D Systems) for sharing content. While this is a start, we think the primary hurdle to more meaningful adoption is a more user-friendly design interface that does not have as steep of a learning curve as current packages SolidWorks, Rhino, and Autodesk 3ds. 3D Systems appears to be the leader in flattening the learning curve by acquiring a number of software companies and app developers. With several smaller acquisitions over the past few quarters aiding the effort, including recently announced target Geomagic, we think 3D Systems is making the right decisions to develop the consumer market for the product, though it is not quite clear if 3D Systems will be the sole beneficiary of those investments.
On the other end of the spectrum is ExOne (XONE), a pure play in industrial-use metal applications that recently came public. The company's machines are capable of creating production-grade objects and castings out of steel, bronze, glass, and aluminum. With the capability to print and specialize in metals, ExOne distinguishes itself from Stratasys, which only produces plastics. 3D Systems and ExOne presumably compete for the same customers, but ExOne's larger build platform addresses a key scalability concern that 3D Systems has yet to master. Moreover, to our knowledge, ExOne is a bit ahead of the curve in its qualification of materials to use in its printers.
By eschewing the consumer and even some of the more mainstream prototyping markets, ExOne is somewhat removed from the machine pricing pressures starting to affect 3D Systems and Stratasys. ExOne's machines sell for anywhere from $0.4 million to $1.4 million, with customers favoring the more expensive high-end units than the lower-priced units. In our opinion, the key to ExOne's success hinges on staying ahead of competitors on material capabilities and unit manufacturing capacity.
While the company's prospective position in the industrial 3-D printing market is compelling, we were a bit surprised to see the rather low volume that ExOne has shipped to date, with only 13 units sold in 2012 and 4 in 2011. Although it is encouraging to see year-over-year growth, we find it difficult to claim shipping 13 units a success, particularly since the death knell for many 3-D printing companies through the years has been an inability to meet the market's demand. ExOne seeks to address this by tripling capacity by 2015. While the opportunity is great, so is the risk that the company is taking on too much too soon. ExOne has yet to deliver positive earnings, cash flow, or EBITDA, and while the natural response is that more volume would help the company scale its cost base, we are concerned that the company is adding to its fixed-cost base in order to grow, limiting the strength of that argument.
We continue to believe the most accessible end market for 3D printers is rapid prototyping, enabling design engineers to have faster, cheaper iterations in the design cycle. We validated our initial assessment of the economic argument for 3D printers in the prototyping industry by conducting interviews with different prototyping firms. Initial prototypes needed for form, function, and fit verification can run between $15 thousand and $20 thousand per iteration because of the high costs of developing a custom mold. Additionally, the process can take three to four weeks from initial design to receipt of the prototype. During a normal product-development cycle, there are several iterations of custom prototypes to ensure the final product has the intended form, fit, and function. Each iteration incurs the same capital and time investment. The payback period for a 3D printer can be within one project. While a 3D printer still requires materials, the machine vastly improves the timeliness of the development process and also provides it at a lower total development cost.
Prototyping firms expressed a general preference for Objet's ultraviolet-cured process because of its better finishing and the small layer thickness; the lowest-cost Objet printer can create objects with a 28-micron layer thickness. Companies also voiced concerned about ABS plastics being a bit brittle and not as useful for all phases of testing, but did note that they have more than one type of printer at their disposal. This confirms our previous thinking that the distinctions among printer technologies will facilitate more competitors without necessarily causing a drop in profitability. Because the end customer values traits of the finished product more than anything, and because of the short payback period of the printers for professionals, we think the market can support multiple types of printers, each occupying a relevant niche.