TSRA THESIS

TSRA is primarily a semiconductor intellectual property (IP) firm with 2 additional divisions that are currently unprofitable (imaging/optics IP and silent air cooling for laptops). 

TSRA is trading at a massive discount to other semiconductor IP firms given what we believe is an incorrect perception that the company's royalty stream will end starting in mid 2012 as existing contracts begin to expire.  There are several catalysts outlined below that could result in a significant increase in valuation.  In particular, based on our due diligence, we believe TSRA's key customers will renew their licenses (see "Bear Case and Counter Argument" below).

• TSRA trades at 2.8x 2010 EBITDA (after adding back stock-based comp and intangible amortization) and 4.9x P/E (excluding $8.75/share in net cash) and 9.9x overall P/E. FCF yield is 10% levered and 20% unlevered.
  • Semi IP comps like RMBS, ARMH, MIPS trade at 10x-20x EBITDA. TSRA traded in this range after its IPO, so if the revenues are viewed as more sustainable, it is possible TSRA regains some of its multiple.
  • TSRA has a great balance sheet with roughly 50% of market cap in cash and no debt.
  • At 20% unlevered FCF yield, TSRA just needs 4-5 years of cash flow to generate the current market cap. Also, with 15%-20% royalty growth/year, even if the top 2 customers (25% of sales) were cut in half in 2012, TSRA would still be at $1.50 in cash EPS.
• Well over 100% of EBITDA comes from semi packaging royalty and even though other 2 segments are losing money, they have positive value.
  • Imaging/Optics division lost ~$45M of EBIT in 2009 vs. ~$100M of core GAAP EBIT for TSRA.
  • Recently, TSRA withdrew $100M sales guidance for 2011 vs. ~$30M currently. However, imaging sales will continue to ramp and mangement is confident they can reach several hundred million in sales over time.
  • TSRA spent about $240M in acquisitions for Imaging/Optics products vs. ~$425M in current total EV.
• Potential Catalysts.
  • Royalty Renewals. TSRA had 4 Japanese customer contracts that were/are due to expire in 2010. While these are small customers, and thus, not meaningful in and of themselves, they are an indication of how future contract renewals may go. From our diligence we expect that these customers will renew and that the larger customers with contracts that expire further in the future will renew as well.
    • 1 of the 4 small customers, Fujitsu, just renewed on 9/21/10 on the original terms.
    • Earlier this year, TSRA also signed a contract with Nanium, a former customer at the bankrupt Qimonda.
    • Numonyx acquisition by Micron starts to pay royalties and adds 10% to current EPS (not in our numbers). Micron recently acquired flash memory maker Numonyx (closed on May 7). We do not believe Numonyx is in analyst estimates as it is not totally clear yet if MU will pay royalties on Numonyx.
  • Image/Optics. As sales ramp, profitability will improve. Even getting this division just to break-even vs. its 2009 loss of $45M of EBIT would substantially add to TSRA's EBIT.
  • Silent Air Cooling. Design win with Apple for silent air cooling technology in 2H-2010. Management believes a design win is imminent. There is a large market potential here: current fans cost $5-$20/unit (silent air cooling would be at a premium) and Apple sold 7.2M laptops in 2009. The total notebook market is ~50M units.

BUSINESS

TSRA's primary source of revenue is royalties from the packaging of semiconductors (this is over 100% of EBIT).  Importantly, TSRA is paid based on chip volumes which have grown at a 21% CAGR for the past 5 years.

• Customer contracts are tied to numerous Tessera patents so the expiration date of any one patent is supposedly not important. Samsung (largest customer) contract with TSRA lists ~150 patents.
  • The main packaging technology is micro BGA (ball grid array). Packaging technologies do not change often and the prior packaging technology TSOP (thin small outline packages) is still in use. Flip chip is a competing technology (currently used primarily in high-end chips like INTC microprocessors b/c of the high cost).
• Revenue Model: Royalties are based on volumes shipped (not revenues/ASP). Volumes are much more stable (and other semi IP companies receive royalties based on sales like ARMH, MIPS and RMBS to an extent).
  • Over the past 5 years, DRAM unit volumes have grown at a 21% CAGR and did not decline over that period.
  • TSRA receives $0.0005/pin (and there are dozens of pins per chip). Analyst estimates of royalty vary, but at IPO in 2004, TSRA IR said royalty was disclosed at ~2c/chip.
• High margin & return business with 45%-55% EBIT margins.
• Customers: We estimate ~70% of TSRA royalties are from the DRAM industry (~85%-90% of DRAM industry are licensees). TSRA receives royalties from 70 customers including in non-DRAM spaces like TXN (for DSP chips for cell phones), Philips, INTC, etc. TSRA's 2 largest customers are likely DRAM companies (Samsung at 15% of 2009sales and likely Micron at 10%).

• Imaging/Optics. TSRA has a variety of products here:
  • EDOF (Extended Depth of Field). Camera lens and DSP chip that eliminate need for auto-focus. Supposedly saves $2-$3/module vs. auto-focus modules. This is main product going forward and TSRA gets a 25c/unit royalty.
  • Wafer level optics. Makes camera at the wafer level to reduce size/materials/cost of camera. 5c-15c royalty depending on complexity.
  • Wafer level packaging: Packaging for wafer level optics. 2-3c/unit.
  • Current handset market is 1.2B units although TSRA thinks the available market is over 3+B units including uses like auto, security, toys, etc.
• Silent Air Cooling. Fanless cooling technology. Speculation is that AAPL could be a design win in 2H of this year with revenues ramping in 2011/12. Supposedly Jobs hates fans.

 BEAR CASE AND COUNTER ARGUMENT

• The bear case is that TSRA royalties are not sustainable:
  • BGA has been around a long time now, is widely adopted and is not even the best packaging technology any more.
  • Some of TSRA's key patents (used in many of their legal cases) are expiring in 2010. Once this happens, TSRA customers will bail. TSRA has 2 key DRAM customers expiring in mid 2012 (~25% of total sales).
  • TSRA has experienced mix success recently in legal cases at the ITC vs. prior history of uncontested success.
• There are 4 main reasons to think this will not play out:
  • BGA technology will still be around for a long time (will be used in DDR 4 DRAM which is several years away from introduction).
  • TSRA has 100-200 patents in each contract so no one patent is critical. TSRA also has newer patents on BGA and next generation packaging technologies that they excluded from prior contracts.
  • To help in negotiations, TSRA has a very credible threat of an antitrust lawsuit against most of the DRAM players for price fixing. RMBS is pursuing this and recently settled with Samsung for $900M. TSRA is pursuing this vs. Hynix. Micron (MU) and Samsung are concerned about this.
  • Quality diligence at three large customers (Samsung, Micron, Amkor) indicates high likelihood of renewal:
    • A smaller fourth customer, Qimonda, also indicated they would renew (and signed a contract a earlier this year).
    • Reasons include: fear of litigation with high cost and uncertain outcome, fear of DRAM antitrust suit, TSRA is very smart about re-issuing and continuing patents and TSRA has some newer packaging technologies (from Samsung).
    • For Amkor (AMKR), TSRA historically patented new AMKR developments with AMKR as a write-on licensee.
• TSRA has no real other businesses.
  • If optical sales ramp and/or TSRA wins silent-air cooling with Apple, this will be viewed very differently.

LEGAL HISTORY

• Brief Legal History.
  • Before 2008, TSRA had tremendous legal success and most defendants settled even before trial. This includes numerous large companies like TXN, INTC and Samsung. TSRA never had a legal setback/loss.
  • TSRA was involved in 3 active lawsuits at the ITC (small Taiwanese DRAM players, wireless chip manufacturers and semi subcontractors).
  • In Feb 2008, TSRA had 2 large legal setbacks. Defendants in the ITC cases challenged TSRA patents at the PTO (Patent & Trademark Office) and had some early stage victories. This in turn caused the ITC to postpone temporarily the wireless ITC case (which was widely expected to be a quick victory).
    • TSRA share price declined from $40->$14 in Feb 2008 as a result.
    • TSRA ultimately won the wireless ITC case on appeal (although after an initial loss). MOT signed as a licensee in mid 2009 as a result.
• Pending legal cases.
  • Wireless ITC re-appeal decision is due any day now. This is really more headline/perception risk as there is no material financial impact. MOT and QCOM are 2 largest defendants. MOT already signed a contract when TSRA won the initial appeal and QCOM is not paying TSRA anyway (uses AMKR who refuses to pay TSRA).
  • DRAM ITC case is due late this year or early next year. Again this is more headline risk given TSRA already signed up 85%-90% of the DRAM industry.
  • AMKR arbitration. TSRA won last arbitration but will likely take 3 years to settle this and just began. Management believes they will win this as well.
• PTO Cases (Patent Trademark Office).
  • 3 defendants in TSRA suits (Siliconware, Powerchip, Amkor) are trying to invalidate 8 key patents in the PTO. Potential bad outcome is that patents become invalidated and current customers try to renegotiate terms on existing contracts.
    • Mangagement argues that PTO cases will be delayed/extended until patents expire so these cases are effectively pointless (6 of 8 patents expire this year). PTO cases are simply a common defensive reaction to a patent infringement case.
    • New customers have continued to sign up with TSRA, even as PTO suits are continuing, including MOT is mid 2009.
    • TSRA contracts have 100-200 patents listed.