Description

I believe Xperi has a highly favorable risk/reward ratio at current prices. While the stock is up 72% from its lows of 2018, it is also still more than 50% below its highs of early 2017 while the primary causes of the peak to trough decline have been addressed and I believe the company has been de-risked in the medium term. Arguably, the stock has under-reacted to the positive developments of the last 14 months. Indeed, the stock is only about 10% above where insiders made substantial purchased in late 2017. Xperi was also the subject of a September 2017 write-up by Shooter McGavin and despite the positive developments he anticipated is still about 25% below the write-up price. Furthermore, I think that the large stock price decline in 2018 was at least partially due to a misreading of their results by quantitative investors in 2018, a situation that may reverse itself when they report Q4 2018 results.

Historically Xperi (fka Tessera Technologies) was primarily known as a licensor of semiconductor packaging technology to the DRAM industry. While the company had its ups and downs as various patent deals expired or were renewed, it is important to understand that Xperi is not a patent troll that acquired some obscure patents to hold up an industry, but an active developer of semiconductor packaging IP that invests significant amounts into R&D and collaborates with its licensees to bring its inventions to market. In the 2013-2015 timeframe activist fund Starboard Value acquired a large stake in XPER (slightly below current prices), successfully implemented changes in the board and management team that led to reductions in money-losing new R&D projects and a return of capital and exited the stock. In late 2016 Xperi purchased DTS Inc. for $975M. DTSI developed and licensed audio technologies (but generally did not rely on patent assertion) including the technology behind the HD radio standard which is being adopted for North American AM/FM terrestrial radio. Today Xperi reports its business is two segments: Semiconductor IP licensing and Products (which is largely DTSI/HD Radio and the Fotonation imaging product that was part of Tessera).

XPER stock peaked at around $45 in early 2017, but the company dropped a bomb on the market when it revealed that its largest patent licensee, Samsung, had declined to renew its license at the end of 2016. This led investors to focus on the upcoming expirations of its other major patent licensing deals with DRAM producers Micron (believed to be year-end 2019) and Hynix (2020) and two outsourced assembly and test vendors (Amkor and PowerTech) which the company has indicated it is not planning to renew when they expire at the end of 2018. IP consultants I have spoken with have indicated that some of the key patents which Xperi used to force long-term licensing deals in the past have expired, although the company has filed numerous continuation patents, new patent applications and acquired significant other IP over the past few years. While DRAM is currently in a downturn, Xperi’s semi IP deals are fixed rather than per unit royalties so they are not dependent on DRAM price, volume or profitability.

In Q4 of 2017 a key patent licensing event occurred: the licensing of Broadcom Corporation to a multi-year deal that I believe to be worth about $120M over four or five years. The implication was about more than just the cash they would receive from Broadcom. XPER has traditionally generated the vast majority of its semiconductor IP licensing revenue from the DRAM industry, which is only about $100B of the approximately $400B global semiconductor market. The licensing of Broadcom represented the first major XPER licensee outside of DRAM, and theoretically opened up a greenfield licensing opportunity worth 3x the $190M per year or so of DRAM licensing revenue. News of the Broadcom license sent XPER shares up to as high as $28 in December 2017. 

XPER currently sits at around $21. What happened in 2018? First, in the company’s Q4 2017 call, the company gave a surprisingly downbeat long-term outlook for its semiconductor IP licensing business. While some analysts believed that licensing Broadcom opened up a substantial growth opportunity given the greenfield opportunity outside of DRAM, the company’s investment presentation called for 2022 semiconductor IP licensing revenue of $160M to $222M vs. $205M in 2017. I believe this is due to management’s conservatism after being burned by the unexpected non-renewal of Samsung and if they had Samsung under license (spoiler alert: they licensed Samsung in December 2018) they may have guided for modest growth.

Additionally, a change in accounting standards may have led to confusion among both human and algorithmic investors. ASC 606 superseded ASC 605 relating to revenue recognition for revenue from contracts with customers. The past standard allowed a company such as XPER to recognize the revenue from a multi-year patent license agreement with a customer ratably over the course of the agreement, which typically matched well with cash flows when a license was paid quarterly or annually. ASC 606 requires a licensing company to recognize most or all of a multi-year patent license agreement upfront and applies retroactively even to deals signed before the advent of ASC 606, so at the end of 2017 Xperi was forced to recognize several years of earnings onto its balance sheet from its ongoing fixed license agreements despite the fact that it has not collected the cash yet. This lead to 2018 revenue and GAAP earnings estimates being substantially revised downward despite no change in non-GAAP earnings (calculated under what the company defines as “billings”, more or less corresponding the old accounting standard) and the underlying cash flows. I do not believe that the I/B/E/S GAAP revenue and earnings estimates appropriately reflected the change in accounting standards at the end of 2017 and the stock action since Q1 2018 is at least partially due to this apparent degradation of its GAAP results as the company’s GAAP FCF and non-GAAP earnings guidance did not change. While 2017 results under the old accounting regime show a strong revenue gain due to the acquisition of DTSI, starting in Q1 2018 the headline results showed declines in revenue and earnings, and I believe this may have contributed to the stock’s poor performance for most of 2018 as shares were sold by quantitative investors.

After presumably engaging in discussions with Samsung subsequent to license expiry at the end of 2016, Xperi launched ten separate legal actions against Samsung at various United States and international venues. Chief among these was a suit at the United States International Trade Commission, which is a venue of choice among legitimate patent licensing companies due to the ITC’s relatively swift decisions, which can ban the import of any company’s products found to infringe the IP of a U.S. company. Indeed, the Xperi settlement with Broadcom was reached shortly before a final determination was due in an ITC suit brought by Xperi. Unfortunately for Xperi, while it was only one of the ten legal actions brought against Samsung, Xperi’s ITC suit against Samsung was thrown out on May 21st, 2018 due to the court finding that two of the asserted patents (acquired from a Japanese company) in the suit were subject to a previous arbitration agreement. This setback pushed back the timeline for Samsung resolution and apparently led to some investors throwing in the towel.

Xperi’s Q1-Q3 reports were largely non-events, although before the Samsung announcement it appeared that Xperi was on track to report 2018 results at the low end of its initial 2018 guidance. The Product licensing segment (DTSI/HD Radio/Fotonation) looks to have grown billings at the low end of expectations in 2018, although the company has mentioned its expectation that Product billings growth should re-accelerate in 2020. Interestingly enough, management voiced their confidence in signing a new IP license agreement during 2018, which most investors took to mean a greenfield opportunity (i.e. not Samsung). They presumably only said this if they had several productive discussions ongoing, so I would not be surprised to see a new greenfield license announced in the next few months.

It would be difficult to overstate the importance of successfully re licensing Samsung to Xperi’s prospects. In a December surprise, Xperi announced signing a new license with Samsung on December 10, 2018. Samsung recently has held out out on re-licensing patents unless forced to via litigation in numerous situations I follow. Based on the enormous breadth and depth of Xperi’s patents and its successful licensing of Broadcom—not to mention the fact that the DRAM industry was until recently awash in its highest profits ever and Xperi’s license ask is a rounding error to the DRAM industry—I believe that Xperi’s successful re-licensing of Samsung will lead to an improvement in the growth outlook for their IP licensing segment and makes the future Hynix and Micron renewals much more likely. Amazingly, Xperi’s current stock price currently sits below where it was immediately before they announced that the ITC had thrown out their Samsung case, when the renewal of Samsung looked uncertain and the company seemed to have significant litigation expenses in front of it not to mention the risk of renewing Micron and Hynix.

Xperi did not announce the per-year billings, term or total amount of the Samsung deal but they did raise the high end of their Q4 2018 billings guidance by $21M. Due to the fact that under ASC 606 almost all the revenue from a patent license deal is recognized up front, we should have a better idea of the scope of the Samsung deal when the company reports Q4 2018 GAAP results and files its 10-K. Note that this also means that they will show an enormous increase in revenue and earnings for Q4 2018, the reverse of the situation from Q1-Q3 2018 that I believe may have caused quantitative investors to sell. Given that Samsung previously paid them $50M per year in semi IP licenses, the previous guidance appeared to include $10M to $15M from generating a new semiconductor IP license (which didn't happen before the end of 2018) and there was probably some weakness around the margins in DTS/Fotonation from weaker smartphone sales/China/tariffs, I believe the new Samsung license is for around $40M per year. Based on this new guidance I expect Xperi generated around $3.00 per share in FCF during 2018, a current FCF yield of 14%. Granted, Xperi is facing the loss of its Amkor and Powertech licenses at the end of 2018, which will cause a loss of $63M in high margin cash collections. Looking to 2019, Wall Street forecasts already contemplate a decline in revenue and earnings. According to my numbers, even with the $63M revenue loss mentioned above, overall revenue should decline by only about $50M due to growth in DTSI/HD Radio as well as other parts of the business such as FotoNation, and litigation expenses should decline from about $30M in 2018 to $10M in 2019 due to terminating the 10 legal actions against Samsung. All of this leaves the company with $2.50 or so in FCF per share, still a 12% FCF yield. In fact, I estimate a current enterprise value of about $1.3B for XPER, which is less than the $975M XPER paid for DTSI plus the $385M or so it will collect from semi IP licenses under contract (another $105M from SK Hynix & Micron, $80M to be collected from Broadcom in 2019 and beyond and what I think will be a $160M of unbilled receivables from the Samsung deal) and assumes basically zero ongoing value for its thousands of patents in the semiconductor space built with hundreds of millions of cumulative R&D.

Thus, I believe the bear case of Xperi being unable to re-license expiring DRAM-related IP revenue or close new greenfield licenses is more or less debunked. What does Xperi look like now that they have successfully re-licensed Samsung? While revenue and earnings will likely decline in 2019 due to losing Powertech and Amkor at the end of 2018, I believe that revenue and earnings will return to growth in 2020 due to DTSI/Fotonation, even assuming that Micron’s renewal comes at a modest haircut. At 15x ongoing cash earnings per share of around $2.50, Xperi could be worth around $38 per share, upside of 80% to today’s prices and still below the stock’s previous peak in early 2017. The stock pays a substantial 3.8% dividend, the company is buying back stock and insiders have purchased not far below current prices. The company has also publicly hinted it is close to closing other new greenfield semiconductor IP licensing deals, which would be further virtually 100% margin upside to my estimates.

Disclosure: The fund that I work for owns shares of XPER. We may buy or sell shares of XPER at any time without notice.

Catalyst

• Large GAAP revenue and profitabilty growth in Q4 due to ASC 606 accounting will be apparent when they report
• New greenfield semiconductor IP licenses
• Continued share repurchases and dividend payments