Why we are short Rubicon Technologies, Here’s the NEWS !
Rubicon Technology is the #2-3 manufacturer/grower (behind Kyocera & Namiki) of sapphire wafers for LED substrates. Sapphire is one of two commercial technologies for growing wafers. Silicon Carbide is the second type of technology, and CREE Industries(CREE) is the leading player there.
The success for RBCN has been in promoting the opportunity for growth in LED.
LED is one of the next gen technologies going into Televisions(TV’s) and Wall Street always like a clean and simple story. We want to point out we are not anti-LED’s or even Wall Street’s affinity to the LED story. We recognize that company’s like VECO and AIXG, who make capital equipment for LED’s, and are the only two players in the space, have seen their revenues grow and backlog explode. For those two specific reasons their stock prices have been rewarded and we are not short them.
We address RBCN because unlike VECO and AIXG, RBCN’s revenues have been declining in the face of this growth market, something we will address later, and because unlike being one of only two players, RBCN is one of 19 players just in the sapphire market, not including Silicon Carbide. List of other companies.
– Atlas (Ru)
– Crystal GmbH (NYSE:D)
– Crystal Systems (US)
– Crystal Wise (TW)
– CrystalQ (NYSE:N)
– Epistone (NYSEARCA:CN)
– Honeywell (US)
– Interoptec/Fujimi (J)
– Kyocera (J)
– Monocrystal PLC, Inc. (Ru)
– Namiki (J)
– Rubicon Technology (US)
– Saifei (CN)
– Saint-Gobain Crystals (F & US)
– Shinkosha (J)
– SilianTech (CN)
– Sino American Silicon (TW)
– Sumitomo Metal Fine Technology (J)
– Tera Xtal Technology (TW)
Of course nothing is quite so simple. How RBCN distinguish itself is related to the claim of novel technology and ability to manufacture larger wafers better than the competition. Currently the majority of wafers are grown on 2 and 3 inch wafers. The thesis on LED wafers is that like in the semiconductor industry, manufacturing efficencies increase the larger the wafer, so naturally the industry will eventually go to larger wafers.
It is these last two claims, novel technology and larger wafer manufacturing that we investigate. We are not wafer engineers, but it was our luck that we found a fantastic source to describe how RBCN’s technology wasn’t novel but a commodity process imported from 1970s Russian technology. That source was none other than Rubicon’s founder, Radion Mogilevsky. A man who, while no longer with the company, actually still owns the equipement that Rubicon uses to manufacture wafers. Rubicon pays him rent, b/c they are using his facility as their Franklin Park, Illinois headquarters. All this information can all be found in his lawsuit against the company filed ONLY 4 MONTHS ago. (don’t try to find it in RBCN’s filings, they don’t list it under legal proceedings).
Case 1:09-cv-02554 Northern District of Illinois
Cash 09-CH-016116 Circuit Court Cook County
The summation of the case, Mr. Mogilevsky goes into much greater detail, relates to RBCN’s method, the Kyropoulos, the history of this method’s creation in Russia, how it came to RBCN, and his ultimate point as he states is that RBCN’s technology is neither “unique” nor of “the highest available purity”.
The CEO of RBCN on the other hand wants the street to believe they are not a commodity. He told Semiconductor today the following.
“If the difficulties associated with uniform epiwafer production at those larger diameters can be ironed out by then, Parvez believes that Rubicon's “ES2” crystal manufacturing process will result in a significant competitive advantage over rivals using the more traditional vertical-gradient freeze (VGF) and Czochralski techniques.
“To get respectable yields and throughput [with 6-inch production], you need ES2,” he said. The technological barrier with the other techniques is significant.” http://www.semiconductor-today.com/news_items/2009/JULY/RUBICON_290709.htm
The only problem is, they don’t have new technology, it’s just the old technology with a new name.
Prior to the use of ES2, RUBICON TECHNOLOGY, INC. referred to the same method as Kyr-plus.
See “Rubicon Technology, Crystal Clear,” by Jennifer Monroe, November, 2002,
www.themanufacturer.com/us/profile/704/R... (April 18, 2009)
So there is not new technology, and it’s not novel. It’s 1970’s Russian technology, slightly updated and repackaged.
Slow Movement to Large Wafer:
The second point from a technology side is wafer size. Just like with semiconductors, to achieve lower costs and increased yields the goal is to reduce edge effect and process more die in a singe process step. In theory we see where the industry is going. While LED’s are much smaller than semi chips, we have no issue with the idea. Below we have provided a link to both VECO and AIXG’s views on 6 inch wafer processing. We would point to pages 4 & 5 of the VECO. In summary, VECO has issues currently with 6 inch yields and specifically cost to manufacture a 6 inch wafer that has the necessary thickness vs. 2” & 4”. Remember, conformity of yield is the most important issue.
"Judging from the uniformity of the 18,000 1x1 mm Vertical LED on Metal Alloy chips from one single wafer, we can clearly see the yield advantage of this larger wafer MOCVD process. One of the contributing factors to yield enhancement is the significantly reduced edge area compared to the area equivalent of nine 2-inch wafers.”
Since I’ve bored you with the information, the conclusion is pretty clear. Yes, wafers will over time go to 6 inch. They are going to 4 inch now, up from 2 inch. But they aren’t going quickly and wait, because it gets better, who cares. After investigating this claim, a majority of companies, those already listed above, all offer larger wafers (see link to Kyocera, the worlds largest manufacture). Most companies state that 6 inch wafers are not used commercially and are limited to custom orders. A far stretch from RBCN’s claim of having unique ability to manufacture them specifically.
From Kyocera: “We offer a wide range of products including standard 2”, 3” and 4” wafer diameters in various crystal orientations such as C-, R-, A- and M-planes.
Custom orders for larger wafers (greater than 6”), super-flat wafers or special requests to improve LED brightness are also welcome.” http://global.kyocera.com/prdct/fc/sapphire/index.html
So to conclude on technology, it’s neither novel, nor does RBCN have an advantage to growing larger wafers.
Financial Results & Outlook:
But let’s get to the point. What makes stocks go up and down? Technology is important, but what about revenues. A small piece of a huge pie is good enough right. That’s at least what’s driving some of the bulls here. Let’s examine revenues.
2009 2008 Chg
2Q LED revenues $2.5M $6.9m, down -63.7%.
3Q LED revenues $4.7m $6.78m down -30.6%.
So in a growing space, RBCN’s revenues are declining.
If the story isn’t about the past, what about full year? From RBCN’s last conference call, the company has stated their capacity to be $30-$35m, which is down from the $50m in capacity that the company claimed on the 1Q09 conference call. So the number is declining. If you follow RBCN closely you will know that some of that has to do with pricing, now if you follow the semiconductor industry closely, and technology in general, you know that pricing is supply and demand weighted. With 18 other players and a commodity technology, pricing should over time continue to decline.
The other revenue issue RBCN wants you to focus on, and allows this write up to come full circle is the % of wafers produced above 3 inches. Basically why try to convince someone you have a better mouse trap, when the only thing people are buying is the basic mouse trap. I’ll let the data do the talking.
2nd Quarter % of LED Revenues 3 inch or greater 24% 40%
3rd Quarter % of LED Revenues 3 inch or greater 33% 47%
I hate to make a joke of this, but these are % of LED revenues, and as you might recall, LED revs y/y are down. So the amount of LED sales in the category that we question RBCN has an advantage in, is down y/y in both % and dollar terms.
The last comment we will make in regard to RBCN’s LED revenues revolve around RBCN’s largest customers. RBCN’s top accounts are not the suppliers to the LG’s and Samsungs of the TV world. They are Perigrine, Shinkosa, Crystalwise. So for all the talk of huge opportunities here, currently, RBCN is not accessing these large end customers. Does a raising tide lift all boats, it might, but watch out if your ship is named the S.S. Minnow.
The final bull component to the story is the push into new business opportunities in polishing. Rubicon has stated recently that they would like to spend their cash to build capacity so they can polish the wafers, in addition to growing them, which is their current business. The reason of course is to get a high ASP and not lose margin to a middleman. Again, in theory, we get the back of the envelope math. We ask two simple questions. If the manufacturing and growing of sapphire wafers for LED is so HUGE, why need another business, especially when the company only did $4m in revs last quarter. The other and more important is, your customers are the polishers, and processors, if you are going to compete with your customers, with 17 other options, good luck, especially since you’re selling a commodity product.
So what have I missed? How can a company that did $4.7m in LED rev’s last quarter and have capacity for $30m in LED revenues be worth $362m or over 10x these LED revs. The bulls will of course say huge growth. (That’s assuming a bull even gets this far down before using some expletive about a short seller). And the bulls are correct in that LED will grow, but if we are being intellectually honest with ourselves this makes this a thematic trade, and as there are few institutional investment opportunities in the LED space, it increases the likelihood of a company, like RBCN, to get bid up artificially for a time, as people buy the space not the company.
So is it cheap then? First we looked at VECO and AIXG, as they are riding the LED wave. They are also capital equipment manufactures, so its not even close, RBCN is much more expensive. So maybe CREE, who is a manufacturer of LED. Amazingly, as expensive as CREE is, RBCN is more expensive on a forward PE, EV/Sales, EV/EBITDA, Price to book multiple, and that’s even with a 3 year rev CAGR that’s below CREE’s.
We even went all the way back to 1998-2000 and looked at AXTI. AXTI used a proprietary crystal growth technique to produce semiconductor substrates. SOUND FAMILIAR. For a period of time AXTI was the sole supplier and even then, with zero competition vs. RBCN’s 18, RBCN is still more expensive on every metric.
So what else could it be? Following on the theme Wall Street analysts have jumped on board with one analyst after the next upgrading to a buy and increasing their price targets. But what comes next? We don’t want to be there when this disappoints. In addition, the street’s estimates are far too high. As the street often does, in order to justify a price target, the street needs a certain level of earnings and EBITDA. The current 2011 estimates, the very same estimates that the street is banking on for their buy ratings, require a corporate margin that this company can not support. RBCN currently has a negative EBITDA margin, and negative earnings, for these future street numbers to come close, RBCN would have to achieve greater margins than VECO, AIGX, AXTI, and a multitude of other similar companies. Unless RBCN gets into another business, these estimates are wrong. Our conclusion is they will seriously MISS these estimates and the stock will come crashing down.
LED is hot, RBCN has had a fantastic run. We could say this about “.com’s” or solar, or a number of other thematic plays that were hot, but then became commoditized. As we see from VECO, AIXG, AXTI (in 1999) and in our opinion RBCN, the early money has been made. Now RBCN has to grow into its 10x LED sales multiple. We think RBCN is a commodity company in a hot space, nothing more. Lower pricing will continue to impede the growth investors are looking for. Pat yourself on the back, and sell now. If you want to wait it out, take a look at companies like Global Crossing, or Evergreen Solar (ESLR). Both got the trend right; the web would grow exponentially and solar energy powers a significantly greater portion of the grid than it did even 3 years ago. The trend was right, the stocks, dead wrong. RBCN is a trend right, stock wrong play.
Disclosure: Author is Short RBCN