Description

Investing in the 8.5% 2023 bonds will provide for solid risk-adjusted returns in the current environment.  10%+ yield, highly unlikely you lose money.  We will earn a 16% total return if the company is acquired in a year (101 put plus interest).  The company is being shopped and is an attractive acquisition target for a competitor looking to gain scale, but my thesis does not require M&A.  

Note about statistics in box above

For price I am using bond price.  I am of course using stock price for P/FCF, P/E.  For P/E, I am using stock price and FCF per share.  For P/FCF, I am backing most of the cash (all the cash less an interest payment) which is why the multiple is so low.  EBIT is EBITDA less maintenance capex.  EPS rises in 2020 substantially but this due to the rise in revenues coupled with operating leverage.  I am recommending the credit versus equity as the credit does not depend on improving metrics.

Company overview

UTAC Holdings (UTAC) is a leading independent provider of outsourced semiconductor assembly and test (OSAT) services for a broad range of integrated circuits with diversified uses, including in communications devices (such as mobile, Bluetooth, and WiFi), consumer devices, computing devices, automotive applications, industrial applications, and medical applications.  The company provides assembly and test services primarily for four key semiconductor product categories: analog, mixed-signal, logic, memory and others.  The company's customers are primarily fabless companies, integrated device manufactures, and wafer foundries.  

UTAC is headquartered in Singapore with production facilities in Singapore, Thailand, Taiwan, China, Indonesia, and Malaysia.  

Revenues by service type for Q1 2019 broke out as follows: 68.3% assembly, 29.3% test, 2.4% other.  This mix has been relatively stable over time.  Revenues by product category: 51.7% analog, 30.9% mixed-signal and logic, 7.9% memory, 9.5% other.  Revenues by customer geography: 53.4% US, 25.2% Japan, 13.7% Europe, 7.2% Asia ex-Japan, 0.5% other.  

Test services typically have higher unit gross margins than assembly since test services require minimal material and component costs.  Test operations however are more capital intensive and have higher fixed costs.  The majority of test services are in the mixed-signal and logic category.  A meaningful amount of assembly revenues comes from analog.  Post-restructuring, the company has a renewed focus on test services as it helps bring in customers for full turnkey solutions whereby it provides test and assembly services.         

The company's top 10 customers in 2018 were Analog Devices International, Broadcom, Formosa Advanced Technologies, Maxim Integrated, Microchip Technology, ON Semiconductor, Panasonic, STMicroelectronics, TSMC, and Texas Instruments.  

UTAC does not focus on early stage products, so its ASPs tend to decline more gradually compared to some peers when providing services across the semiconductor cycle.

Equity vs. credit with industry comments

UTAC has unlisted reorg equity that can be purchased via a broker.  While the equity is not expensive, I personally prefer the 8.5% 1st lien bonds due Jan 2023 (USD).  The credit is attractive as it has downside protection, matures in a little over 3.5 years, has no duration risk, and trades at over a 10% yield.  The bonds are registered and the minimum size is $200k, so this works for PAs and smaller funds.  Global A&T Electronics (GATSP) is the issuer of the bonds, and UTAC is GATSP's parent company.  UTAC has hired bankers to sell the company.  If there is a change of control, we can put the bonds at 101.     

UTAC will continue to exist over the long term.  It is the 7th largest global OSAT player in a space which is relatively consolidated with the top 10 contractors having over 80% of the market.  Amkor is the only large U.S. based player, with the majority of players being in Asia.  Amkor shifted its HQ from South Korea to Arizona in 1999.  Amkor's long term stock price gives a fair glimpse of the industry: it's an industry that provides value to customers, but not so much for shareholders.  Amkor's stock price has been essentially flat for two decades.

The semiconductor industry has contributed to global growth and has had a large impact on how we live.  But it's a difficult industry which has faced constant challenges.  Semiconductors were invented in the U.S. and the U.S. dominated the industry for the first 30 years.  But in the late 1980s the Japanese became the dominant player.  In the 1990s, the U.S. reestablished its dominance partially thanks to Intel, U.S. government policies, and a Japanese industry which was slow to pivot from a Japanese mainframe market to an international PC/consumer market.  The industry then had another major challenge to tackle: the rising cost of fabrication as each generation of technology was becoming significantly more expensive.  This led to the proliferation of pure-play foundries and OSATs and a ramp in production in Asia.  Later, prices plunged as the business mix shifted even more towards consumers and away from corporates.  More recently, design costs have been increasing, particularly since the 40 nm mode.  The increase in costs at smaller modes is particularly concerning as new generation mobile, wireless, and AI requires high complexity silicon and will use these advanced nodes. 

Certain players have had incredibly success (Intel, Texas Instruments, Linear Technology Corp which was acquired by ADI, Maxim Integrated, etc.), but most firms simply earn their cost of capital at best.  Most companies in the space create value for consumers without being able to capture much value for themselves.  While UTAC's equity will do well if they can scale, I am recommending the bonds since I have higher conviction in the bonds.  The equity is cheap if the company can grow and execute and take its fair share in a large market; however, this is not assured and it's a competitive space where there is a cap on return on investment.  As mentioned before, the best U.S. comp for UTAC is Amkor.  Amkor is a larger company which benefits from its increased size, but UTAC has the potential to grow its revenues at a higher pace and improve its profitability more given its capabilities and smaller scale.    

Despite my comments on how it is difficult to extract significant value in the semiconductor industry, I do believe in the OSAT model.  The semiconductor industry has historically grown a few hundred bps higher than GDP per annum and OSAT share of assembly and test revenue has grown over time.  For instance, OSAT share was 50% in 2013 and is now around 53%.  In 2008 it was around 45%.  Given the increasingly expensive design and manufacturing costs, integrated device manufacturers prefer to focus on their core competencies.    

History

The company started operations in 1997 in Singapore and in 1999 acquired the memory semiconductor test assets and operations of Fujitsu which were also in Singapore.  The company offered memory test services and then started offering turnkey assembly and testing of DRAM.  It established operations in China in 2003 and went public in Singapore in 2004.  In 2005, it acquired UltraTera Corporation, a Taiwanese testing company.  In 2006, it acquired NS Electronics Bangkok, a Thai analog assembly and service busines.  TPG and Affinity Equity Partners (a well-known private equity firm based in Asia) won the auction for UTAC for $1.8B USD in Q4 2007.  In 2010, it acquired ASAT Limited, a semiconductor assembly and test business with operations in China.  In 2014, it acquired OSAT assets from Panasonic.  The Pansonic acquisition has been a nice one as UTAC acquired good assets and now has Panasonic as a customer.  

Due to the LBO and subsequent M&A, the company was overlevered with well over $1B in debt.  The company was overlevered by the sponsors, but the sponsors were also incredibly aggressive and, in my opinion, illegally made the original 2nd lien pari passu with the 1st lien.  This of course made the 1st lien drop substantially in price, and the creditors sued through various stages in NY.  The court documents (SDNY 17-23991) are interesting and informative.  The creditors won, and GATSP then filed a pre-pack in December 2017 and emerged from bankrupcy in January 2018.  Creditors received these first lien notes I am recommending and some cash and 31% of the equity.  69% of the equity went to the original private equity sponsors.  The financials I mention include the results of the GATSP segment and the legacy UMS segment -- UMS was contributed to the GATSP debtor level which benefits these notes.  The UMS contribution is meaningful.      

There is always concern with jurisdiction when it comes to investor rights.  These bonds are NY law bonds, and Singapore is also a good jurisdiction.  As part of the original lawsuit, creditors were pursuing tortuous interference against the sponsors which would have allowed the creditors to go after the sponsors themselves as the typical corporate veil protections would not have applied.  This would have been a 1-week pre-pack but since JPM objected to the plan it took 1 month.  It was still a fast process.  

Summary financials and valuation

The company produces cash with its reduced leverage.  The company has about $1 million of finance leases/other loans, and $665 million of these 8.5% senior secured 1st lien bonds.  No other debt.  I estimate that 2019 EBITDA will be $170 million (off of $790 million of sales) and capex will be $90 million (some growth capex included, but I use $90 million for maintenance capex).  Cash interest is $56.6 million.  At year-end, the company will have about $215 million of cash (versus $232.5 million currently).  The market is trading the equity at 4.5x EBITDA, or 9.4x EBITDA less maintenance capex.  This is a reasonable valuation.  Net leverage (which is where we are creating our position) is 2.6x EBITDA or 5.6x EBITDA less maintenance capex.  Some forecasters use a 11x EBITDA less maintenance capex multiple on 2020E EBITDA of $195 million.  This of course yields very high upside for the equity, however I am not underwriting to this case.  UTAC has meaningful fixed costs and its facilities are operating slightly above 70% utilization on average, so there is ample capacity for growth and operating leverage.  

The bonds have tight covenants due to large, sophisticated creditors being involved in the restructuring negotiations.  In addition, by my estimates, the company conservatively has about $60 million of valuable Singaporean real estate that it owns.  I do not include this in any of my statistics.    

EBITDA less maintenance capex margins are slightly over 10% and return on tangible capital sligthly above my estimate of cost of capital.  Optically ROTC is higher than my number and some investors claim return on capial in the high teens, but my over-the-cycle estimate is lower than the equity bulls.  Returns can be volatile due to the cycle, contracts which are not long-term, currency, and restructuring/M&A.  Things can change quickly and the company does not have a lot of visibility.    

Why are these bonds cheap?

This is a bit of an unusual bond with multi-jurisdictional assets.  During the restructuring process, for instance, there were NY distressed funds involved, Asian funds, etc.  Some of the credit trades out of Hong Kong, some trades out of NY.  It's an orphan name, and it is not widely followed.  I purchased the credit from a desk in Hong Kong.  Additionally, recent numbers have been weak, but this is largely due to general semiconductor trends.      

Near-term outlook

The company expected weakness (due to general semiconductor industry weakness) in 1H 2019 though it has met guidance.  Q1 2019 revenues were down 12% sequentially from Q4 2018.  This compares to Amkor (#2 player globally) being down 17%, ASE down 15% (#1 player), JCET (#3) down 22%.  In Q1, semiconductor players were hit by the inventory correction in both iOS and Android as well as weakness outside of mobile.  However, EBITDA is expected to be $170 million this year versus $180 million last year (my estimates, other investors are forecasting $190+ million for 2020).  EBITDA in 2017 was nearly $230 million.  The drop in 2018 was due to a weak smartphone market and the closing of USC, a facility in China.  Sales are down across segments and the company has partially countered this with cost containment.  Nevertheless, the company is investing in growth and catching up from losing out on opportunities during its period of financial distress.  Business did not disappear during that period, but the company did miss out on some new customer opportunities.  Management is seeing stabilization in the business this quarter.  The company and its customers are predicting a robust 2H outlook based on new wins and a solid project pipeline.  Requests for quotations are at high levels, and in the past this has been a decent indicator of future trends.  The company has successfully won a meaningful amount of business in power and 5G projects.  The industry in general is predicting rebounding mobile numbers from the launch of several flagship phones in 2H 2019, as well as recovery in non-mobile.      

I've done customer calls with 5 of the top 10 customers and the feedback was that UTAC has excellent customer service, execution, and engineering.  3 of the customers I spoke to said they held business back from UTAC during its period of financial distress and that they are now comforted by UTAC's financial profile.  TI, which is a powerhouse in analog with 30% market share, explained how UTAC has added value to its solutions.  UTAC's Thailand plant has been doing analog OSAT work for over 40 years.            

The company is well positioned from a technology perspective and has recent design wins in several areas including:

CuClip -- UTAC's cooper clip solution which is a favorable replacement for traditional wire bond interconnection used for high performance metal-oxide-semiconductor field-effect transistors.  Applications are in high-end computing servers/data centers, as well as automotive.  Cu Clip allows for higher current, higher frequency, and better efficiency.  

Image sensors -- UTAC's solution offers superior optical and reliability performance in camera applications. The company's ceramic packaging has lower moisture content and allows for higher pixel density.  

MEMs -- Has had success working with customers involved in production of oscillators, magnetometers, accelerometers, gyroscopes, and tuners.

SiP -- UTAC partnered with a leading supplier of substrates to create its System in Package solution.  UTAC's solution scores well on system miniaturization, electrical performance, thermal performance.    

Automotive -- UTAC plays in many areas, such as sensors and QFN (quad-flat-no-lead) packages.  QFN works well for applications where size, weight, thermal, and electrical properties are critical.

Plasma dicing -- Improved yields when doing wafer dicing.  Wafer dicing is generally done with conventional blade technology.  Blade cutting can lead to die chipping or cracking which can cause lower yields.  UTAC has a solution which can etch very narrow deep trenches into silicon to separate individual die.  

FOWLCSP -- Fan out wafer level packaging is used to reduce size and weight in electronics used in portable, wearable, and IoT settings.

Trade-war risks

The company has limited operations in China but lower sales of mobile handsets and electronics in the U.S. and other regions would of course negatively impact the company.

What happens in a downturn?

I examined prior downturns and think that $140 million is the EBITDA floor given the contributon of the UMS business.  The company has the ability to dial down capex.  I do not expect that EBITDA would stay at that level for long, especially with recent business wins.  Additionally, UTAC has a huge cash balance.  Even in this scenario, it is highly unlikely that the bonds are impaired.  And including the high cash coupon, it is even less likely we would lose money.    

Let's suppose I am wrong and the next downturn is worse than I expect and this gets a $500 million enterprise value valuation in bankrupty in 3-4 years.  TPG and Affinity will be wiped out and first lien holders will own the whole company and we will get some combination of debt and equity.  Let's assume the over $200 million of cash is gone due to poor M&A (unlikely as the last 2 acquisitions added valuable capabilities) and cash burn.  With the coupon payments received, we would be roughly breakeven.  You can vary your recovery assumptions, but it is hard to envision being impaired significantly. 

Catalyst

Ongoing sales process.  

Maturing of bond in January 2023.