Description

  Overview

Hopefully, the second time is a charm.  I initially wrote up Magnachip in 2013 and I was surprised by accounting irregularities from the former management team (tied to channel stuffing).  These irregularities resulted in customer defections and a significant erosion of value.  In the interim period, a new management team has taken over and a number of activists have attempted to have the company sold.  In August 2015, MX formed a strategic review committee and hired Barclays to explore strategic alternatives.  An Electronic Times articles in February 2016 reported that the Company was in the midst of selling the business in pieces to a South Korean fabless business, Synaptics and SMIC.  In May 2016, Engaged Capital settled with the Company and got two seats on the board.  Given Engaged’s strong track record this year of getting company’s sold, the market apparently had high expectations for a sale.

Last week, MX indicated in their earnings that although the company will continue to review strategic alternatives, there was no current transaction to sell the entire company.  The lack of a sale of the entire company overshadowed the solid earnings report.  As a result, the stock declined precipitously.  I was surprised by the decline as my view has been that there is more value to the business in the long run as a standalone business, in particular, one as asset rich as this one, with a compelling AMOLED display driver asset (grew over 100% this year to $160mm+ in revenues) which might make MX the best security in the market to invest in to benefit from the growth of AMOLED over the next 5-10 years.  I am aggressively advocating taking advantage of this dislocation as I believe this has created a compelling opportunity to invest in a business that has shown a material inflection in operations (growing very nicely). I believe the inability to sell the company may have been a disconnect between the prospective upside of the current business plan and the price buyers were willing to pay.  I firmly believe the assets, on a liquidation basis, are worth $10-$16 (60% - 160% upside) and on an ongoing basis (at an 8x-10x ebitda multiple), worth $11-$24 (70% - 285% upside), with potential for real multiple expansion. A lot of the heavy lifting has been done and at the current run-rate, the consolidated business should generate greater than double digit revenue growth in 2017, leading to material EBITDA and FCF growth.  In a blue sky scenario, where MX can generate peer like gross margins (28-30%) and EBITDA returns to historical (pre 2013) levels, the stock could be worth $25-$30 (300% - 400% upside).     

Business

Magnachip operates a hybrid business model, serving as both a foundry provider for fabless semiconductor manufacturers while also designing and manufacturing its own power management and display drivers.  The company is the only independent provider of AMOLED drivers and owns fabs with production capacity of 121,000 eight-inch equivalent semiconductor wafers per months (located in Korea).

Management and Activists

The current CEO, YJ Kim, was hired from CAVM to run the display business and was promoted to CEO after the accounting irregularities were unsurfaced.  Last year, Pleasant Lake's John Lennon attempted to take the company private (not a credible offer) and within the last few months, the company agreed to add two Engaged Capital nominees to the Board of Directors.

AMOLED Growth 

OLED growth over the next five to 10 years should be very strong with IHS indicating a CAGR of 23%.  MX is the only independent provider of OLED display drivers and announced in early 2016 that they had already shipped drivers for 170mm phones (and that was prior to the business doubling this year).  MX explains that this growth mostly comes from the "wave" of smartphone makers in China that started to adopt AMOLED displays in their mobile phones. MagnaChip's ICs were designed into 29 smartphone models, and Magnachip believes it is the world's second largest supplier of AMOLED ICs.  In addition, the company indicated that they had shipped 7mm OLED TV drivers through May.  OLED TV’s are forecast by IHS to continue to grow rapidly. 

OLED TV market forecast (million units) 

 * Source: IHS Inc. OLED TV shipment forecast

According to Goldman Sachs research, only 10% of total smartphones are currently OLED but could represent 50% of the 1.7B in total smartphones over time, an increase of 300%. 

The SYNA conference call on 10/27 discusses their expectations for this phone display inflection:

"Let's turn our discussion to OLED, as many of us have been monitoring its market adoption closely. We believe the supply of OLED displays will start to expand meaningfully in late 2016 and continue to be featured primarily on flagship smartphones in 2018 due to its high cost, versus LCD technology. Wider adoption of OLED technology in mainstream smartphones will likely take place in 2019."
 

I am attaching below a presentation from Universal Display (OLED) on the OLED market.  OLED has a different business model – they sell materials with high gross margins (but with significant patent risk) and will generate roughly flat revenues year over year in 2016 (at roughly 200mm) but trades at 12x revenues and 36x ebitda.  My monkey math would suggest that the OLED business for MX will generate $200mm of revenues and $25mm of EBITDA in 2017 and should grow to $40mm of EBITDA in 2018.  A 12x multiple (one third of OLED mult) of EBITDA would value this business at $480mm, or $14 a share. 

Presentation from OLED – growth of market on page 6-7

http://s2.q4cdn.com/831534118/files/doc_presentations/UDC-2016-IR-presentation.pdf

Valuation, Liquidation Value and a SMIC               

The broader analog semiconductor market appears to be growing 3-6% a year, resulting in a steady increase in demand for new capacity.  It is generally believed that capacity is much more expensive greenfield vs. brownfield (i.e. buying an underutilized existing fab).  SMIC’s recent purchase of Lfoundry for $3,750 per monthly eight-inch wafer start ($150M deal EV / 40,000 monthly wafer starts = $3,750), is a clear validation of that view. 

SMIC discussed the capacity situation on their Q2 call:

“This is the first time in our history when we build our capacity and all of them are fully utilized as soon as the capacity has been available. So at this moment, still that we have a lot of orders that has not been fulfilled, and so we need to still expedite our tool installations, as well as look. And that's the reason we have to look for external expansion through merger and acquisition. And we have been very, very fortunate to find a good partner such as LFoundry.”

At a similar price, the MX foundry alone would be worth $487mm, or $11 a share (MX has monthly capacity for 130k 8-inch wafer starts x $3,750).  It was illuminating when SMIC, discussing the acquisition, mentioned Lfoundry’s 18% gross margins and how SMIC feels it can take grow them to 30% over time (comparable to their core business).  In other words, SMIC would be able to flow 500mm of sales at 30% GM into the MX foundry, implying a 4.5x EBITDA purchase price at $3,750.  SMIC trades at 7.5-8.0x.

As I mentioned earlier, the AMOLED business is expected to grow 23% CAGR over the next 5 years and is currently generating 164mm in sales, growing to 200mm in 2017.  Under a liquidation scenario, I apply a 1.5x revenue multiple to the AMOLED business ($300mm of value) and value the rest of the company’s proprietary semi business, which is struggling with margin, at .5x revenue, or 130mm.  As a comp, SYNA, which appears to trade at the lowest multiple of sales in the industry, trades at 1.35x sales.  Adding up the Foundry and Non- Foundry businesses suggest a valuation of $917mm (less $150mm of net debt) or $22 a share.  Capitalizing a severance/ pension liability of roughly $130mm, would reduce the value by roughly $4 a share, or $18.  Since this number seems unreasonable high for a liquidation, I haircut it by 10-45% for my $10-$16 range.

Valuing the business on an ongoing basis, I assume free cash flow accrues to the balance sheet and I apply an 8x - 10x EBITDA multiple to $60-$70mm of ebitda in 2017 and $90-$100mm in 2018. I discount back the 2018 valuation by 1 year at a 10% discount rate.  This results in valuations ranging from $11 - $24 per share.  In addition, to the extent management was able to get 2018 margins to 28% (still 200 bps below peers), EBITDA would grow to $125mm (same as company’s 2012 levels) and the stock would be worth $25+ and would be generating $80mm+ of free cash flow (11x FCF at target price). As a note, the company has $280mm of NOLs so taxes should be manageable for a while.

From a downside prospective, current guidance for Q4, would suggest EBITDA of $12mm or $48mm, implying a valuation of 7.7x (with a seasonally weak AMOLED).  AMOLED, on average next year, should be 12-15mm higher per quarter, adding $3-$4mm per quarter, bridging to my ebitda expectations for the year of greater than $60mm.  On current year numbers, I think MX is trading at a little over 6x EBITDA.

Is the company still pursuing strategic alternatives?

We find it interesting that the company was very specific in its press release that it is not pursuing a complete sale of the company. Therefore, we would not be surprised if there were certain assets sold to maximize value.  The company said that they were no longer pursuing a sale of the whole company but that the strategic alternatives committee was still active.  As I highlighted above, there appears to be significant asset value in the company that can be optimized.   The best alternative, in my view, would be to sell the foundry business while keeping the semiconductor business.  The company would then have a massive net cash balance sheet and will have a remaining business that can consolidated grow in the mid-teens that would trade at a multiple significantly higher than slower growing peers (9.2x).    

Current Operations and Operational Bridges

The fallout of the departure of the former management team was severe as management had devoted little time to maintaining the backlog of the foundry business and it started to decline precipitously.  Foundries are very high fixed cost businesses and the decline in the loading (with utilization dropping to 60-70%) caused revenues and gross margins to decline from $800mm and 30% in 2012, to $633mm and 21% in 2015.  At the outset of this year, the street was projecting further declines in the business to roughly $600mm in sales while management was indicating higher aspirations as a result of strength in tape outs (new design wins) for the foundry business and growth in AMOLED.  The Company has proceeded to beat revenue expectations every quarter this year, and based on the guidance for next quarter, MX should do roughly 685mm in revenues (or 8% growth).   More importantly, the foundry and the non AMOLED power and display businesses are showing strength for the first time in many years - the company is guiding those businesses (which all grew double digits sequentially last quarter) for further growth.  We have now seen an inflection in those businesses which is the key to business stability and margin and cash flow growth.

Based on current trends, it is likely that revenue growth for 2017 will be 10%+.  The fourth quarter is seasonally a down quarter for MX, so using Q4 as an annual run-rate is probably conservative.  For the foundry business, Q3 revenues were $75mm and are guided to grow sequentially - assuming $2mm of sequential growth, a $77mm run-rate would be $308mm.  I believe the Non AMOLED power and display businesses will do $65mm in q4 (based on the imputed seasonal decline in AMOLED) or $260mm, assuming no growth.  The company guided the AMOLED business to grow at least at the industry rate in 2017 of 23% (base of 164mm), implying $203m in revenues - add it all up ($308 +$260 + $202) and you have $771mm, or 12%+ growth.  If the other businesses continue to grow sequentially, which seems very possible, revenues could approach $780-800mm in 2017.  Even if I am too aggressive and revenue is $750mm as the company prunes lower margin business, the inflection has happened and revenues are now reaching historical levels. 

With low to mid-teens revenue growth potential and with the associated fixed cost absorption, I think it reasonable to expect margin expansion in 2017 and beyond.  On the low end, peers such as SMIC generate 28-30% gross margins, which is the company's goal.  To stabilize the base and increase fixed cost absorption, management was willing to fill up the foundry with lower margin business, but now with the foundry loaded, management has the ability to optimize mix and benefit from higher capacity utilization.  This past quarter, margins decreased due to management’s willingness to accept certain low margin business but are now targeted to bounce back to 22-24% in Q4.  The company has said that the AMOLED gross margins are above the corporate average and the foundry business generated 23.5% gross margins, implying the non-oled power and display businesses produced a mid-teens margin.    In fact, I believe the non-oled power and display businesses generated roughly 18% gross margins for the full year.  This should be low hanging fruit in terms of optimizing the portfolio and eliminating unnecessary business (or raising prices to customers).   This business has intellectual property associated with it, so for it to trade at a margin lower than the foundry business does not make sense.  A margin improvement on this $260mm business to 21%, would add $7.5mm to GM and EBITDA – alternatively just stopping to produce the products (over time) and devote the fab capacity to foundry business (now that it is more fully utilized), would increase the margins by 600-700 bps or $12.5-$15mm.   For 2017, I am assuming AMOLED margins expand from 25% to 26% as they are moved to a third party foundry (maybe even better), assuming foundry grows margins by 1% (similar to 2016) and assuming a 3% increase in the remaining non-AMOLED display and power (to 21%), GM % would increase year over year to 23.5-24%, or gross margins dollars of $180-190mm.

With Opex of $36mm a quarter (including $6mm of depreciation), EBITDA in 2017 should grow to $60-$70mm and should further grow to approx $100mm in 2018 (assuming similar sales growth and 125 bps of gross margin expansion to 25%).  With ongoing capex of $15mm and $16mm of interest expense, and limited cash taxes, FCF should be meaningful (14%-17% FCF yield in 2017 and 30%+ in 2018). 

Conclusion

There are multiple paths for value creation and there is an opportunity to buy after a technical puke related to the “no sale” of the whole company.  The business has stabilized and is growing and has assets that can generate meaningful value to shareholders either as a standalone business or on a SOTP.  The turnaround has taken a long time, but after a 3-year journey MX is at that positive inflection point that we are constantly looking for as investors.

Catalyst

- OLED Growth

- Foundry Growth

- Margin Expansion

- Sale of Division

- Materially positive earnings revisions