As of year end 2007, they had 133mm diluted shares outstanding, 25mm of a 5.45% convertible bond due in Sept 2010, and a cash balance of 34mm.  Breakeven should be 11mm per quarter going forward.

 

Cherb405 wrote TXCC up many times in the past and those writeups are worth reading.  His analysis was top down, as it had to be at that time, but now we can go into a more bottoms up calculation to view how the company might look in the near future.  In a year’s time, 70mm is a realistic annualized revenue rate.  With a 60%+ gross margin, TranSwitch will produce large amounts of cash relative to its recent share price.  While Cherb405 postulated significant increases in revenues, the precise source of these revenues was necessarily an abstraction at the time.  While it may not be obvious at all from either the share price or the income statement, the company has come a long way since his write-up.  It is now possible to discuss revenues in terms of specific, funded deployments that are set to ramp over the next few years.  While there are still many abstractions with respect to revenue growth, it is possible to model significant revenue growth based upon tangible and defined opportunities  

 

Currently, legacy revenues are running at $12 million per year.  These are remnants of design wins from the 1990’s.  It’s a declining business but interesting in that it illustrates how long the sales from certain products can last.   Next generation revenues have been around $12 million a year.  TXCC acquired Mysticom in January 2006 for $5mm of TXCC stock (more on this later).  Today, Mysticom accounts for $4 million a year in revenues.  Together, this is $28 million.  In 2007, the company produced $32 million in revenues so this is a slight decline. 

 

Below, I’m going to discuss announced, concrete potential business, although, as I mentioned above, I believe these projects are merely indicative of a larger trend at work.  They will demonstrate that there is enough of a margin of safety simply using announced, quantifiable revenue opportunities to comfortably establish a position in TXCC while waiting for additional opportunities to make themselves known.

 

Revenue Opportunities:

 

British Telecom has announced it’s upgrading its network to a Next Generation architecture in a project called 21CN over the next 4 years.  BT expected to spend a little less than $20B which it expects to produce cost savings of $2B per year. If successful, BT will serve as a model for other telecom deployments going forward.  In fact, BT has started a consulting unit to take advantage of the knowledge they have gained in the process.

 

This single project could drive TXCC’s income statement from a net loser to profitability.  TXCC will benefit via design wins with Fujitsu, and specifically within Multiservice Access Nodes (MSANs).  TXCC has publicly stated that they expect to see $100 million in revenues from this project, but that was based on a 55% market share in MSANs for Fujitsu, and 45% to Huawei.  While unconfirmed, there are rumors that Huawei has been running late and Fujitsu could see a larger market share.  For now, I will model $25 million per year in revenues for this project, although if the rumors are true, the company could see an additional $30-40mm of revenues in this portion of the project.  In addition to Fujitsu design wins, TXCC will benefit from the ADVA and Redback platforms that will be included in 21CN as well – together, these could add 2-4mm of sales.

 

At this point, 21CN has been running late, but BT has announced that they intend to start launching on April 30^th.   Indications are that the buildout will start slow, but ramp through 2008.  There is risk that this project gets delayed but as we get closer to the launch date, no news = good news.

 

Korea Telecom is also scheduled to upgrade its transport network and this project could be in full ramp by late 2008 (part of a larger upgrade).  While there isn’t as much published information as in BT’s 21CN, the company projects 40mm of revenue from this project and I am depending on the company’s estimate.  TXCC has meaningful content with five of the six likely vendors including all four of the Korean vendors that are expected to be awarded the bulk of the contract.  Assuming the project takes 4 years, that equates to $10mm per year.

 

 

Bharat Sanchar Nigam Limited (BSNL) is India’s leading telecommunications provider.  Originally, BSNL tendered for 45mm lines.  However, after that tender was completed,  Union Minister for Communications and Information Technology A. Raja, in an irrational display of power, put the brakes on the tender as it stood to negotiate lower contact terms.  Now that the bureaucracy has finally (or at least temporarily) subsided,  BSNL is building out 23mm cellular lines. 

 

TXCC will benefit from design wins in Tejas Networks and United Telecoms (UTL) in the backhaul network.  Coincidentally, the delays have been good for TranSwitch in that due to the contract restructuring, more of their silicon content will be used.  Previously, TXCC had estimated potential revenues at 3.5mm per year but I believe that actual revenues will be twice that amount for 7mm for the first 23mm lines.  Importantly, BSNL is falling behind because of delays while Indian cellular subscribers are growing at clip of 6mm per month – to maintain their competitive position, they will have to aggressively build out additional lines.  It is likely that the local companies Tejas and UTL will win the subsequent business and if on the same terms as the original 23mm lines, they will benefit at the rate of approximately 0.30 per line.  Revenues from this project, while not as large as BT, have the potential to carry on for quite some time.

 

Alcatel will be an important customer for TranSwitch over the next few years.   On conference calls, TranSwitch has estimated $8mm of sales annually to Alcatel.  Part of the recent decline in TXCC shares is due to a lack of recent Alcatel orders.  The company blamed revenue shortfalls last year on expected orders that did not come through.  This, in combination with a delay in the expected onset of orders from BT21 (above), led to a selloff in TXCC shares.  TranSwitch has many design wins with Alcatel and specifically, some meaningful design wins with their ET-3 and ET-48 in Alcatel’s 1626, 1692, and 1696 next generation WDM platforms which have a leading market share.  Alcatel probably had too much inventory late last year, and as they burn through that inventory, orders to TranSwitch will begin again, and perhaps as early as the second quarter.

 

In China, TranSwitch’s relationship with Alcatel will produce at least $12mm of revenues over the next two years from announced contracts, and perhaps double or triple that.  Alcatel has signed a package of agreements with China Unicom and China Mobile to supply network equipment valued at $1.1B using platforms listed above.  This is on top of the 8mm estimate described earlier, and Alcatel’s contract will be implemented over the next 2-3 years.

 

In the US, TranSwitch should benefit from wins in Tellabs in Verizon, and potentially Motorola in Verizon.  For now, I’m modeling these at zero, although I do believe TranSwitch will see revenues from these wins.

 

Also in the US, AT&T and Verizon have both made press releases recently that indicate Ethernet over Sonet (EoS) is here.

 

 

 

 

While it’s not clear when this will begin to ramp and impact TranSwitch, it is clear that TXCC will benefit.  TXCC has over 115 design wins for the ET-3 in Customer Premises Equipment (CPE) that will be necessary for this service.  In late 2004, Cherb405 estimated 15-35mm of revenues for the ET-3 in 2006 with the expection that EoS would begin to ramp earlier.  While he was incorrect in timing, his estimates for revenue size have yet to be tested.  For now, I will include 5mm, but I believe this area has significant potential for growth within the next few years.  This source of revenues is one of the wildcards that could propel TranSwitch shares into the high single digits if all goes well.

 

I believe there will be other sources of revenues that will occur as the industry begins to ramp.  Because I want to demonstrate a margin of safety in this write-up, I’m going to use a base case of 0, and a high case of 5mm.  I think both of those numbers will prove to be too low.  I also hope that by putting low values in this section that I do not mislead a reader into believing that this area is insignificant when in fact, I believe just the opposite.  Perhaps the most significant piece of my thesis is that the “Other” category will produce large amounts of revenue over the upcoming decade.  Just a few years ago, many of the discrete revenue scenarios that I presented above would have fallen under the “Other” category because they weren’t announced.  In a sense, the format in which I’ve presented this thesis is awkward – I’ve spent a lot of time outlining specific deployments (the tip of the iceberg) to demonstrate how much lies below.  This is a particularly important point because the market won’t attribute a high multiple to a stock that has a single set of four year projects – but when the market realizes that this is just a subset of projects that validates a much larger and longer trend, I believe that TranSwitch shares will benefit from not only from increased revenues/profits but the valuation multiples that investors will afford TXCC will increase substantially as well.

 

Finally, I would like to discuss Mysticom and its HDMI opportunity from a top down perspective, which is drastically underappreciated by the market.  If all goes well, this segment of the company could produce more than $100mm in revenues annually.  This alone would produce a market capitalization worth more than the entire current market value of the company.

 

As noted above, TranSwitch acquired Mysticom in early 2006 for its Serializer/Deserializer (SerDes) technology, which converts data between serial data and parallel interfaces in each direction.  Mysticom had invested more than $100mm to develop this and other technologies.  The acquisition price was only $5mm + an additional $10mm to be paid based on meeting certain revenue objectives (those objectives were never met and TranSwitch did not have to pay more than $5mm).  How could TranSwitch purchase a company for $5mm that could potentially produce such large revenues?  On its own, Mysticom isn’t as valuable as it is to TranSwitch.  TranSwitch took Mysticom’s SerDes technology and applied its own technology (similar to how it maps Ethernet to Sonet) to create its HDMI chips.

 

For a background on HDMI, check out this link:

 

 

InStat estimates that more than 325mm HDMI ports will ship annually by 2010, which includes both the transmitters and receivers.  At an average selling price of $3-4 per port, the market size should be around $1B, which will be broken down into both chips and IP cores, with chips accounting for the majority of the market.  TranSwitch estimates that it has a 6 month lead on any of its competitors for its HDMI 1.3 version technology, including Silicon Image Inc. (SIMG) which is the market leader with over 70% market share and a 55% gross margin.  There are not a lot of players in this market, and a successful company can attain a large market share, and potentially very quickly as the cycle is fast.  A criticism often heard about SIMG is that it competes with its own customers to a certain extent.  TranSwitch will not have the disadvantage of this conflict of interest and this alone could be a major selling point.  Additionally, TranSwitch will target DisplayPort which SIMG does not target.  It was developed and adopted by computer and electronic companies that did not want to have to pay license fees to use the HDMI standard.  DisplayPort has the potential to be a $300mm market by 2010.

 

Obviously, this is a large market and simply being in the lead is not enough.  This is a new market for TranSwitch and a major obstacle is the lack of a knowledgeable sales effort.  However, without minimizing the topic, I believe this problem can be solved quickly by hiring an experienced sales team from one of the larger players in the industry to build out this opportunity.

 

Dr. Santanu Das, co-founder and CEO, has thrived in similar developmental situations.  (On a side note, his story is inspiring and illustrative of the leader behind TranSwitch.  He came to the US from Calcutta, India without a dollar to his name.  After attending college and then struggling financially through graduate school, he got married, and began to work for ITT Corporation where he moved quickly up the ranks.  For the sake of brevity, I’ll skip a few steps to when he co-founded TranSwitch.)  Early on in TranSwitch’s history, the company, like most start-ups, was cash strapped.  Dr. Das negotiated free rental space in return for shares of TXCC.  (TranSwitch still rents its current headquarters from that same landlord.)   When the company encountered internal production problems that would have required the company to spend money it did not have, Dr. Das was instrumental in developing a strategic relationship with Texas Instruments where TI would build the integrated circuits for TXCC in return for TXCC sharing some of its knowledge and the promise of future business.  Through another relationship, TranSwitch was able to win the trust of Siemens-Taiwan and gain entrance into the larger parent organization, something that could not have been accomplished otherwise.  And finally, TranSwitch was able to negotiate a deal with Taiwan Semiconductor Manufacturing Company (TSMC) that proved to be very critical for TXCC.  The two companies forged an alliance which saved TranSwitch more than $20mm in upfront fees at an early stage in TranSwitch’s history and TSMC received an entrance into the telecommunications market.  Almost 20 years later, many of the participants in those early relationships are still considered friends and associates by Dr. Das.

 

Evaluating new markets always involves a lot of uncertainty, and this is no different.  The company has received significant solicited and unsolicited interest in their HDMI technology and I believe that Dr. Das has relationships that will benefit TXCC in this venture.  I am hard-pressed to put a PV on this, but my sense is that a 10% market share is achievable.  Regardless, at the current prices, an investor is getting the Mysticom potential for free.

 

 

Potential Income Statement:

 

Here are two potential revenue scenarios using the low and high cases:

 

 

 

Using a 60% blended gross margin and modest fixed expense increases, I expect that an annualized income statement will look similar to this in a year’s time using my low case:

 

Revenues:         85mm

CGS:                -34mm

Gross Margin:   51mm

SGA/R&D:      -35mm

Operating Inc:   16mm

Taxes:              -2mm (they have Net Operating Losses that can be offset against income)

Net Income:     14mm

EPS:                 .10

 

Using the high case yields a net income of 33.5mm and an EPS= .25.

 

I believe the market will apply at least a 25x multiple to a small, and fast growing income stream and that would lead to a share price of $2.50 and up to $6.25 in the high case.

 

However, many investors value similar companies at a 5-7x revenue multiple.  Using that metric TXCC could trade for anywhere from $3.20 to $5.30.

 

Both of these cases are assuming the company has not bought any stock back (see below).

 

Anyway you look at it, the potential is huge.  In 2-3 years time, I believe TranSwitch will far surpass the low case, and probably surpass the high case as well.

 

 

Potential Catalysts:

 

On February 13^th, the company announced a $10mm stock buyback after pressure from some of its largest shareholders.  The company is not required to repurchase shares.  I think the pressure by the shareholders and willingness on the part of the company to initiate a stock buyback means that both groups see value in TranSwitch.

 

TXCC’s shareholder base includes Herbert Chen of Chen Capital Management (14.89%), Michael Steinhardt (9.89%), QVT Financial LP (7.45%), and Brener International Group (5.61%).   Herbert Chen has written many letters to the company that you can see in the public filings.  He has been nominated for a seat on the Board of Directors along with Thomas Baer.  You can read more in the proxy.  The shareholder base has become more active in the last few months, and I believe they want to see a meaningful change in how the company is operated.  I expect this to accelerate in the upcoming months.

 

As discussed above, potential catalysts include the onset of meaningful orders from Fujitsu for use in the 21CN project, orders for use in Korea Telecom, or orders for use in BSNL.

 

Revenue run rate should begin to move higher over the next year.  I believe this will serve as a catalyst.

 

Risks:

 

While there is differentiation between different chips and IP from the various companies, average selling prices (ASP’s) typically decline over time and that could negatively impact revenues.  However, TranSwitch has been able to maintain gross margins of 60%+ over the last few years.  I don’t understand their technology on more than a superficial level so I can’t speak to how substitutable competitors’ products are.  From what I understand, once a design win has been achieved, it’s usually not replaced.  I’m comforted that they have many design wins, and are not dependent on a single platform.

 

Delays have plagued this industry for years and these continued delays have caused TXCC’ share price to languish.  However, it’s not uncommon for a telecom project to run years behind schedule.  BT has published a public schedule that calls for an April 30^th start so I believe that will reduce the risk of another delay for that project.  Although a delay would hurt TXCC, I don’t think it will change the thesis radically because while net cash was less than $10mm, total cash was $34 million as of year end which will give them ample time to ride out a bad market.  While I think it’s prudent to expect more delays in some projects, I also believe that Cap-Ex revenues dedicated to NGN’s will grow at a rate high enough to offset, or at least minimize, the effect of delays to TXCC.

 

Currently, TXCC is unprofitable and consuming cash.  While they have reduced their breakeven, unless revenues increase, it’s to no avail. 

 

Another major risk to this thesis is that many of the revenue estimates for the above projects relies on information from the company itself.  There are many assumptions that go into each estimate.  For what it’s worth, I believe the company has tried to err on the conservative side in its forecasts.

 

 

Disclaimer: I am a shareholder of TranSwitch.  This is not a solicitation to buy or sell securities.  Please do your own individual research and thorough due diligence before buying/selling any shares in TranSwitch.  I may buy or sell TranSwitch in the future and am under no obligation to provide any update or details on VIC (or elsewhere) on my trading activities.