Description

Summary

ON Semiconductor (ONNN) fits the mold of a general investment thesis that has worked quite well for us over time: a decent company with one of its segments significantly underperforming and masking the earnings power of the company, longer-term.  In short we think, we think ONNN can earn close to $1.00 per share in the next 1-2 years and has a fair value of $10-12. 

ON is cheap because its SANYO division is depressing earnings and folks are skeptical that the company will effectively turn the segment around.  The stock trades at 12x 2014 consensus EPS of 70c, but we think the company could earn closer to 75 or 80c with a run-rate approaching $1.00 towards the back-half of 2014 – which we think would lead to a higher multiple for the stock.  Any industry cyclical improvement could provide upside to these forecasts (semi industry revenues are still 5-10% below the 15-year trend-line in 2013). 

Company Description

ON sells a combination of analog chips and standard products.  In general, producing analog chips is a decent business with about 15 main global competitors.  Analog chips are defined as semiconductors that integrate and regulate “real world” functions such as temperature, speed, sound and electrical current. Digital semiconductors process binary information, such as that used by computers.

The market for analog semis differs from the market for digital semiconductors. The analog industry is typically characterized by longer product life cycles than the digital industry. In addition, analog semi manufacturers tend to have lower capital investment requirements for manufacturing because their facilities tend to be less dependent than digital producers on state-of-the-art production equipment to manufacture leading edge process technologies. ON and other analog companies will typically buy fab equipment after it’s been depreciated by higher-end manufacturers like Intel or Taiwan Semi.  The end- markets are more varied and more specialized than the relatively standardized digital semiconductor product markets. In general, analog has a higher barrier-to-entry – not a great business, but a decent one. 

The company has a portfolio of 40k+ products and sells ~40b units per year at average price of ~$0.07. Manufactures semis, so has operating leverage, but not R&D intensive. End markets (2Q13): Computing = 16%, Consumer = 20%, Auto = 28%, Industrial = 19%, Communications = 17%. Customers: OEM's = ~55%, Distributors = ~40%, EMS = ~10%. Acquired SANYO in Jan 2011. Prior to 1999, was sub of Motorola.

SANYO Acquisition – in 1Q11, ON paid $480m for SANYO.  On top of the amount paid, the company has posted $323m of operating losses.  SANYO was primarily a manufacturer of simpler, commoditized, consumer chips in Japan and other parts of Asia.  Following the acquisition, ON got hit with two issues: 1) a step-down in the competitiveness of Japanese consumer electronics companies and 2) floods at their Thailand manufacturing facility that caused a shift in demand to other regions of Asia.  The company blames most of its problems on these issues; other industry insiders say the SANYO business has been terrible for years and that ON misjudged its ability to streamline its operations.  Either way, the acquisition was bad; but, the analysis must look forward. 

Earnings Walk (Street May Be Miscalculating Potential Improvement)

It seems that the street is calculating gross margins for the consolidated company, when we think you should calculate them for SANYO and Ex-SANYO separately – this leads to a different result, as our gross margin estimate is roughly 100bps+ higher than FC for 3Q/4Q14, with EPS of 45c v. FC at 39c (for the two quarters combined). 

• Revenue – We model SANYO and Ex-SANYO separately – for SANYO, it seems that revenues have stabilized after dropping 40% from the 2011 level.  Still we are modeling a slight decline in 2014 as problems continue. Giving sales of 608m v. 619m estimate for 2013.  For Ex-SANYO, revenues have been relatively stable as a percentage of analog industry sales (measured by SIA data) at ~5.5% market share since 2010.  For all semis, we have an industry forecast, which effectively is long-term trend (of about 6% per year, which has held true for past 15+ years).  We think industry growth slightly faster than nominal world GDP makes sense in the medium-term.  We give ON revenues (ex-SANYO) growth in-line with our industry forecast which includes a slight catch-up to trend. For 2014, we forecast sales (ex-SANYO) of $2.3b (~8% y/y, but still below 2011 levels), so we think this is reasonable. 
• In sum, we’re a bit ahead of the street for revenues, but not significantly ($2.92b v. 2.85b FC). 
• Gross Margin – Since acquiring SANYO, the division’s gross margins have been horrible – averaging about 13% as they suffered from overcapacity, an inflated cost structure, and high levels of inventory. But, as the company reaches break-even in this segment, which should occur in 1H14, gross margins should increase to ~25% (from 22% in 3Q).  We assume 25% for 1/2Q14 and 30% for 3Q/4Q14.  This is still lower than the ~40% gross margins that the ex-SANYO division has posted, though this makes sense as SANYO products are a bit simpler with lower gross margin, but lower R&D as well.  For ex-SANYO, we are using 55% contribution margins in the near-term, which gets that division to ~40%, which is a level ON has consistently achieved in the past. 
• In sum, we’re about 100bps+ ahead of the street for gross margin in 2H14 – I believe mostly because we are modeling the business as the company is managing the business.
• Opex – Again, for opex – it’s important to distinguish between SANYO and ex-SANYO.  For SANYO, we use 24% of sales to start (near the recent results) and then fold in the additional cost cutting (letting go 600-700 people and closing a factory in Japan in 1Q14), which should further right-size the operations.  For the back half of 2014, we’re giving 7.5m per quarter in savings for SANYO (in-line with management estimates).  Overall, for opex, we calculate about 22-23% of sales, which is consistent with recent results and near the high-end of management’s targeted range.
• EPS - In sum, this results in EPS of 22 and 23c for 3Q and 4Q14, respectively and if the company can start to show signs of its ability to hit $1.00 run-rate earnings (even if it happens more in 2015), I think the shares could trade towards $10-12      over the next 1-2 years. 

Looked at another way - ON Earned 68c annual run-rate in 3Q13 and lost 1c at SANYO. If they can get 10% margin on SANYO at 200m quarterly revs, that would add 20c. 10% increase in revenue at 20% incremental op margins could add 10c+.

More Detail on SANYO

Admittedly, the thesis assumes ON will be able to get its SANYO division back to a reasonable operating margin.  But, the company is making progress and we don’t think the hurdle is that high before earnings estimates get revised upwards.  By mid-2015, ON thinks the break-even for SANYO is ~$140m of revenue (vs. $157m in 3Q13).  Today, ~65% of revenues are coming from outside Japan and they expect that to be closer to 75% by the end of 2014.  In other words, the drag from the competitive position in Japan is diminishing. 

Other

• Capital Management - Given the SANYO acquisition and the amount of value and time it destroyed, ON says they will be more hesitant about future acquisitions, weighing them more strictly against the benefits of value created from debt reduction, buybacks, or a dividend.  The company bought back ~1% of shares out in 3Q13.  It’s also at about 0.75x Net Debt / EBITDA as of 3Q13.  The company says it wants to get back to net debt neutral (which it could probably do by      4Q14), but we don’t see why they cant run comfortably at ~1x Net Debt / EBITDA. 
• Revenue Recognition Conservative – ON recognizes revenue on a sell-through basis.  In other words, when it ships products to a distributor (which accounts for ~40% of sales), it recognizes deferred income (on balance sheet), but doesn’t book actual revenue and income until chip is shipped to final customer.  This is more conservative than semi companies that recognize revenue on a sell-in basis and have ability to stuff channel to inflate earnings. 
• Inventory – One red flag is that since the SANYO acquisition, DSO’s have been consistently high.  We would like to see these come down and management has said they would like to get them below 100 days (from 113 in 3Q13).  For the      industry, more inventory has been sitting at the manufacturers (compared with distributors, OEM’s, EMC’s) than in the past as the customer base has consolidated while semi industry has remained fragmented.  We have seen this trend at FCS, TXN, LLTC to name a few companies. 
• NOL’s – ON carries significant NOL’s and will have tax rate of 5-10% for the next five years at least. 
• Yen – weak Yen is overall helpful because a lot of costs are in Yen.  But, the weak Yen has also impacted SANYO’s top-line because it is mainly driven by the consumer segment where the Korean and Taiwan companies have been taking share.  Industrial, Auto, and Chemical industries in Japan have rebounded, but ONNN isn’t as exposed to those segments in SANYO.  This headwind should abate going forward. 
• Proxy – Near-term awards based on adjusted EBITDA and EPS.  Also, goals are separated by SANYO performance and ex-SANYO performance.  Longer-term incentives are a combination of restricted stock awards and options based on target      EBITDA and EPS. 
• GAAP v. Non-GAAP – Semi-Conservative.  ON includes stock-based compensation in NON-GAAP results, but excludes amortization of intangibles and restructuring.  We’d argue that some portion of amortization of intangibles is a real cost as the company is able to delay R&D by effectively buying it through acquisitions.  But, amortization is likely inflated currently from the SANYO acquisition.  We hit them for about half of the amortization expense in our assumption of normal earnings power.      

Conclusion

ON is taking its medicine and buying the stock now, when the windshield is still dirty, should be rewarding. 

Disclosure

We and our affiliates are long ON Semi (“ONNN”) and may buy additional shares or sell some or all of our securities, at any time.  We have no obligation to inform anybody of any changes in our views of ONNN.  Our research should not be taken for certainty.  Please conduct your own research and reach your own conclusions.   

Catalyst