Description
Short Recommendation: Research in Motion (NASDAQ: RIMM). RIMM has not been able to make money during a time when it had the wireless email market to itself; new competition and price deflation will force the company out of business. Consider:
1.Accelerating Insider Sales – RIMM’s top two executives sold significant percentages (6.5% and 11%) of their holdings during 2003, more than twice their total sales since the company’s IPO in 1999. These sales occurred when RIMM's share price was the lowest since the IPO. Increasing insider sales begs the question: what do they know that we don’t? The following analysis is our attempt to answer this question.
2.Weak Patents – RIMM recently lost a patent infringement case to a patent holding company, NTP. The implications of the NTP patent infringement suit are devastating. If RIMM wins on appeal, the architecture of wireless email as we know it will be in the public domain, leaving RIMM vulnerable to the armada of competitors lining up against it. If RIMM loses on appeal, they will continue to pay a 5.7% royalty on revenue generated in the US (US=83% of RIMM’s revenue) plus a court ordered enhancement of 2.85% - total payments to NTP going forward: 8.55% of US revenue. The trial transcripts offer a look at RIMM’s intellectual property (IP). Bottom line: they don’t have enough IP to stave-off mounting competition.
3.Deteriorating Business Model – Over the last 18 months, RIMM was forced to re-organize their sales force around wireless carriers. Their software business is at risk of becoming a feature of Microsoft Exchange Server. RIMM is being pushed into a pure- play hardware manufacturer which sells handhelds through wireless carriers, a tough business dominated by entrenched competitors e.g., Nokia, Motorola, Samsung, Kyocera.
4.Poor Disclosure – RIMM’s disclosure is poor e.g., analysts are left guessing at customer churn, royalty payments are labeled restricted cash and accrued litigation expense, even determining insider sales is difficult. We have uncovered and documented one case of dubious accounting (please see section on Poor Disclosure below) and would not be surprised if more exist.
5. Fat Valuation - Burning cash, RIMM sells for 3.4x book value and 7.8x revenue.
6.Not a takeover candidate - The takeover rumors involving RIMM are unfounded; please consider:
•As mentioned above, RIMM does not control key aspects of wireless email enabling intellectual property. The company will be defending itself for the foreseeable future against NTP and Xerox.
•RIMM lacks scale compared to its competitors.
•As mentioned above, RIMM’s valuation is rich by any measure.
Background
RIMM, founded in 1984 in Waterloo, Ontario, created and sells a service which allows its customers to read and respond to emails through a wireless handheld PDA. The components of the service are: 1) a server which sits next to the customer’s email server 2) RIMM’s relay servers in Waterloo, Ontario 2) a wireless network, and 4) the BlackBerry handheld. RIMM enjoys the following positive attributes:
1.First and foremost, RIMM developed and proved the market for a new business service: enterprise wireless email which is integrated with customers’ existing email.
2.Operationally, over the last year, RIMM management has increased sales while it reduced capital spending, and reduced working capital requirements. Skeptically speaking, it is not entirely clear how they managed this. (Please see below for a discussion on poor disclosure and a detailed example of questionable accounting).
3.RIMM’s assets include $497 million of cash and about $110 million of tax loss carry forwards and investment credits. The tax loss carry forwards are no longer reported on the balance sheet because there is “a significant degree of uncertainty regarding the realization of the future tax assets.” (RIMM’s 2003 AR, pp 26-27.)
Accelerating Insider Sales
Insider share sales are accelerating as RIMM’s shares depreciate. A look at RIMM’s proxy statements over the last few years indicates that Jim Balsillie, the company’s chairman and co-CEO sold 11% of his holdings last year (at the lowest average price since the company has been public). Mike Lazaridis, the company’s president and co-CEO sold 6.5% of his holdings last year. While both executives retain significant ownership in the company, their sales last year were greater than their total combined share sales since the company went public.
Weak Patents
Recent events have given interested observers a peak under the hood, and it appears RIMM’s intellectual property is weak. The company is much more vulnerable to competitive threats than it has led investors to believe.
We have reviewed the transcript (1,849 pages) from the NTP vs. RIMM patent infringement trial. We will first summarize RIMM’s defense during the trial, and then quote the judge’s opinion of RIMM’s defence. Finally, we will draw some inferences from the large amount of information disclosed during the trial.
On May 23, 2003 the court found that RIMM infringed on a series of patents filed in 1991 by a team led by Thomas Campana; at the time, Mr. Campana was an engineer and principal of a paging company. There were three elements to RIMM’s defense:
1.Some of the co-inventors were not included on the original Campana patent applications.
2.The Campana patents did not disclose the best way to carry-out the inventions described in the patents.
3.The US Patent office did not have access to full information about prior art when it approved the Campana patents.
RIMM attorneys even went so far as to allege that what BlackBerry users send and receive on their handhelds are not in fact emails; needless to say, this line of reasoning didn’t fly with the court. We are not in a position to opine on the validity of patents, but RIMM’s defense strikes us as one based on legal technicalities. Here is what the judge had to say:
“[RIMM] offered no real defense to NTP’s infringement case at trial.” (Memorandum Opinion, United States District Court for the Eastern District of Virginia, Richmond Division, May 23, 2003, p.11).
Other IP Facts and Inferences: In addition to losing the case, there are a number of interesting facts and inferences which are included in the trial transcript. Taken as a whole, one is left with the impression that RIMM’s intellectual property is not a significant barrier to competitors wishing to offer RIMM like services. For instance, RIMM consistently touts its push architecture as a key innovation. While it is true RIMM handhelds and the BlackBerry service employ push architecture, RIMM does not hold an exclusive claim to this architecture. In fact, here is how Robert Karhl, one of RIMM’s attorneys, described push architecture at trial: “Every pager, from the time paging was first thought of in the 1950s, used push technology.”( Transcript from the NTP vs. Research in Motion, United States District Court for the Eastern District of Virginia, Richmond Division, November 4, 2002, p. 120.) Mr. Karhl goes on, but you get the point. Ask a RIMM salesman and you will get a different story; we did and were told that RIMM has patented “push” technology.
The court found RIMM’s behavior during the trial so disagreeable that it increased the royalty awarded to NTP from 5.77% to 8.55%. In addition, the court ordered RIMM to pay 80% of NTP’s legal fees. One attorney who read the court’s findings at our request commented:
The defendants [RIMM] and their attorneys really pissed-off the Court, thus explaining the judgment enhancement and award of attorney's fees. Basically the Court came to the conclusion that the defendant and their counsel were far from following the rules that govern patent rights.
Consider the likely range of outcomes if RIMM decides to appeal the NTP case. From RIMM’s perspective, the best possible outcome is that the patents are upheld, but the appeals court lowers the royalty that RIMM has to pay. What is the likely lower limit? RIMM’s expert testimony during the initial trial suggest a 1.5% royalty would be fair (Transcript from the NTP vs. Research in Motion, United States District Court for the Eastern District of Virginia, Richmond Division, November 18, 2002, p. 1732.); adding to this the court’s 50% enhancement, the lower limit comes to a 2.25% royalty.
Another possible outcome is that the court overturns the case completely or that the U.S. Patent office invalidates the Campana patents. Either way, this outcome is less desirable, because RIMM will find it very difficult to prevent others from providing competitive wireless email services i.e., wireless email as we know it would not be patentable. RIMM is already facing challenges from upstarts such as Danger (consumer market) and Good Technology (enterprise market) and several interested incumbents including Microsoft and Nokia. Finally, the appeals court may let the lower court ruling stand, in which case RIMM is stuck paying royalties and penalties equal to 8.55% of its US revenue.
Deteriorating Business Model
RIMM’s service depends on the integration of four elements:
•Wireless Network
•Handhelds
•Email Software
•Relay Servers
RIMM is at a disadvantage vis-à-vis competitors in three of the four areas. The following discussion describes how RIMM has been shut-out of being an integrated service supplier with direct customer contact and forced into being a handheld manufacturer selling PDA’s through wireless carriers. We will then make the case that this is not a tenable position.
Wireless Network: RIMM does not own a wireless network. Originally, RIMM bought bandwidth from paging companies which it bundled (with its handhelds, software and relay server service) and resold to its customers. While this was a reasonable strategy to get the service started, it didn’t last. Observe that bandwidth suppliers are fixed cost businesses and that RIMM’s costs are primarily variable (handhelds, resale of bandwidth). Also observe that 2.5 G technology allowed the wireless cell carriers to offer packet switched data services e.g., wireless email. Unlike the paging companies, which were desperate for business, the cell carriers are not content to allow RIMM to purchase time on their networks for resale: the cell carriers want to own the customers. Guess who won this battle?
In the last 18 months, RIMM has been forced to re-organize its sales force around individual cell phone companies i.e. RIMM sales personnel are assigned to one of five wireless carriers who resell RIMM in the US. RIMM is now fielding a sales force which makes calls on behalf of the wireless carriers. There is duplication, not to mention RIMM is picking up the tab to sell someone else’s service. In the New York market, for example, RIMM has sales personnel assigned to AT&T, Verizon Wireless, Nextel and T-Mobile. As a result, RIMM is being forced into a business model reliant on sales of handhelds through wireless carriers.
Handhelds – Size Matters: As outlined above, RIMM is being pushed into becoming a company dependent on selling handhelds through wireless carriers; this is a tough business dominated by a few very large companies. The economies of scale in this business include the obvious - R&D and manufacturing, as well as the slightly less obvious - logistics and distribution. For example, Verizon Wireless purchases something north of 10 million handsets annually. Note that Nokia makes 140 million handsets anually and offers a full range of designs; RIMM made 352,000 handhelds in the last 12 months: who is in a better position to negotiate with Verizon? For the 33 months ending 5/30/03, RIMM shipped 1.1 million handhelds. For comparison purposes, the following table shows the number of handsets some of the dominant players who sell through wireless carriers sold in the quarter ending 3/31/03.
Handset shipments,Q1 ‘03
Company-------------------------------3 Months ending 3/31/03
Nokia--------------------------------------39,100,000
Motorola-----------------------------------15,200,000
Samsung------------------------------------12,000,000
Siemens-------------------------------------7,500,000
Source: Strategy Analytics
RIMM sells an order of magnitude less units than the dominant players in the industry. While one may point out that the tables are not apples to apples e.g., RIMM is selling handhelds while Nokia is selling phones, this is a false distinction. Note that all of RIMM’s new handhelds include a cell phone. Also note that Nokia (to take one example) is currently introducing the model 6800 into the US market at a price of $100 to $150. This is at a time when RIMM’s 6750 is being sold by Verizon Wireless for $530 with a one year contract. The 6800 is a cell phone which is email capable. It includes a QWERTY keyboard, color screen and a number of other features.
Will the 6800 sell well? It is hard to say. The 6800 is just one example of an industry giant encroaching on RIMM’s turf; or more accurately, RIMM is being forced into competing on Nokia’s turf.
RIMM is quick to point out that they are licensing technology to Nokia for use on the 6800. What does Nokia need from RIMM? It turns out that RIMM is licensing its protocol stack which will allow the 6800 to communicate with BlackBerry Enterprise Server software. This is not an offensive maneuver on RIMM’s part. What this means is that current BlackBerry users may upgrade to the Nokia 6800 without swapping-out their BlackBerry Enterprise Server software for a competing brand e.g., Good Technologies. And what is Nokia paying for this protocol stack? RIMM hasn’t disclosed the terms, but suffice it to say, not much. In fact, as the market matures, it is possible that RIMM may have to pay Nokia for the privilege of including the RIMM protocol stack as an option on Nokia phones just as AOL has to pay Gateway to include its software on Gateway computers.
One final point: unlike most of the other handheld manufacturers, RIMM is vertically integrated (making their handhelds themselves) rather than outsourcing the manufacturing to companies like Flextronics, Sanmina-SCI, Solectron, Celestica, JabilCircuit, etc. It remains to be seen if vertical integration is a prudent strategy, especially with RIMM’s lack of scale.
Email Software – Game Over: If you think the handset business is likely to be tough going for RIMM, consider that there are two companies that control 100% of the enterprise email software market: Microsoft (MS Exchange Server) and IBM (Lotus Domino Server). What are the odds that Microsoft includes BlackBerry Exchange Server functionality in its MS Exchange Server software? If history is a guide (remember Netscape Navigator) it is likely. To make maters worse, the NTP patent case indicates that RIMM does not control the key elements of the intellectual property underlying Blackberry Exchange Server. In addition, as wireless networks become faster, there is less processing that needs to happen before routing the email to the appropriate wireless carrier, reducing the role of the Blackberry Exchanger Server software.
Relay Servers: This is a new business and RIMM is the dominant player. Other entrants include Good Technology and Visto. It is not yet clear if there are any barriers to entry to this business; the service amounts to building a server farm which is capable of receiving emails sent from Microsoft Exchanger Server (or Lotus Domino Server), parsing, addressing and routing the messages to the appropriate wireless carrier. Based on the transcripts form the NTP patent trial, it appears that RIMM has little or no patent protection for this service. Also, while this is currently a separate and distinct service, it is possible that it will be included as a combination of feature offered on MS Exchange Server and the gateway interfaces belonging to the wireless carriers.
Poor Disclosure
RIMM’s disclosure is opaque. The company does not disclose basic facts about its business such as customer churn, major components of SG&A nor does it adequately explain the company’s business model. Facts and figures shift from one year to the next. For example, at the risk of nit picking, RIMM’s 2003 Annual Report includes a pie chart on page 13 which indicates that FY 2002 revenue was comprised of 55.4% Handhelds, 30.2% Service and 15.3% Software and other. RIMM’s 2002 Annual Report includes a pie chart on page 10 which indicates that FY 2002 revenue was comprised of 61.3% Hardware, 30.2% Service and 8.5% Software and other. While the difference is probably in the definition of Handheld vs. Hardware, it is frustrating for the analyst trying to observe changes in RIMM’s business. A second example is their changing definition of Companies which use BlackBerry vs. Companies which use BlackBerry Enterprise Server. Over the years, RIMM management has shifted from one metric to the other as they see fit, making it impossible for outsiders to make sense out of the numbers.
Accounting RIMM’s financial statements are no better. For example, during FY 2002 Q2, RIMM management wrote-off $23 million related to the financial problems (and subsequent bankruptcy) of Motient, who at the time was RIMM’s second largest supplier of wireless network access. Motient settled with RIMM and, according to Rick Burnheimer, Motient Vice President and Treasurer, Motient paid RIMM $9.9 during Q3 of FY 2002. Here is what the Motient 10-K has to say about the transaction: “We reached a settlement with [RIMM] whereby we agreed to pay [RIMM] approximately $9.9 million, representing payment in full, plus interest, for all amounts due and payable under the [RIMM] supply and service agreement.”
RIMM accounted for this payment by reducing SG&A by $3.9 million (reducing Q3 ’02 SG&A by 14.6%) and increasing Revenue by $6 million (increasing Q3 ’02 revenue by 10%). The revenue was booked as Non-recurring Engineering Revenue, (NRE). We went back to RIMM armed with the information that Motient had sent them a check for $9.9 million and asked to know how the money was accounted for; we were then given the above explanation.
Fast forward to the most recent financial disclosure, RIMM’s press release dated 6/25/03. Note that the balance sheet has an item labeled Restricted Cash which comes to $6.9 million. If there were a truth in labeling law, this asset would be labeled Court Ordered Penalties Payable to NTP held in Escrow. On the liability side of the balance sheet, note the category Accrued Litigation and Related Expenses. Again, truth in labeling would call this Accrued Royalties in Arrears plus other Court Ordered Penalties. On the income statement, there is an item labeled Litigation and Related Expenses; this would be more accurately labeled Royalty Expense. Note that while there is reference made to the patent case in the text of the release, nowhere in the press release are the balance sheet terms defined. There are more examples, but you get the point.
Fat Valuation
At $24.61, RIMM is trading for 3.4x book value; its net enterprise value is 6.1x revnue and its equity value is 7.8x revenue. The company hasn’t made a meaningful profit since the last millennia. The company has lost a patent infringement case which discredited its professed intellectual property. RIMM’s service depends on wireless carriers who will continue to take advantage of their position to drive the margins out of RIMM’s business. Given these facts, we conclude that the company is grossly overvalued.
-------------Enterprise Val/Rev--Equity Val/Book Val-----Equity Val/Rev
RIMM------------6.1------------------3.4---------------------7.8
PC mfgs---------1.4------------------7.9---------------------1.5
Handheld mfgs---1.0------------------1.6---------------------1.2
Wireless handset mfgs--1.1-----------2.5---------------------1.1
Elect.Mfgs.Serv.-------0.6-----------1.6---------------------0.6
Software cos. ---------4.7-----------5.4---------------------5.2
Catalyst
1. Increased understanding among investors of the implications of the patent case (NTP vs. RIMM) that the company recently lost 2. Growing competition from handset manufacturers e.g., Nokia, Samsung, Kyrocera, Motorola 3. Inclusion of blackberry enterprise server capabilities in existing competitive products e.g. Microsoft Exchange Server, Lotus Domino Server, Sybase