Tariffs are Taxing the Supply Chain

The breaking tariff situation in the electronics industry is equal parts fascinating and chilling because of its lack of near-term precedence and unpredictability. We’ve spoken with several EMS companies (read the article here) to gauge the extent of the disarray and get a sense of how they are (attempting to) resolve the issue.

Our reporting is ongoing, so be sure to check back occasionally for updates.

 

 

Just What Gets Counted in the CIRCUITS ASSEMBLY Top 50?

I received a note today from an EMS company that had revenues exceeding $400 million in 2017. Why, they asked, did they make the MMI Top 50 EMS list but not the recently released CIRCUITS ASSEMBLY Top 50?

It’s a fair question, and one that comes up each year. In short, MMI calculates its Top 50 differently than we do. While it states it uses calendar 2017 EMS sales, that is demonstrably incorrect.

Key Tronic, for example, which is No. 35 on the MMI list, had sales by quarter of

  • Q1 113.6 million
  • Q2 118.5 million
  • Q3 109.2 million
  • Q4 111.7 million

for a total of $453.1 million. MMI lists Key Tronic’s revenue at $467 million, which was actually the amount for its fiscal year ended July 1, 2017.

Another example: Ducommun. It is similar to many EMS/ODM companies in that it is part of a larger corporation that has other divisions unrelated to contract assembly. MMI has Ducommun listed as 32 on its ranking, yet its EMS sales were, by quarter

  • Q1 $78.7 million
  • Q2 $81.8 million
  • Q3 $79 million
  • Q4 $77.2 million

for a total of $316.7 million. The rest of its sales come from unrelated products such as motors, switches and other non-PCB components. The no. 50 company on the CIRCUITS ASSEMBLY Top 50 was Global Brands Manufacture, with EMS sales of $430.2 million in 2017. As such, in EMS only, Ducommun does not belong. (And again, MMI used the fiscal year, not calendar year, for the revenue total, even though its chart explicitly states otherwise.)

Another reason why the lists don’t match up is in the definition of “EMS.” CIRCUITS ASSEMBLY counts ODMs such as Foxconn, Quanta, Compal, Wistron, Pegatron, etc., because they design and assemble products for other companies. (We also take care to remove the revenue unrelated to direct design and manufacturing for third-party customers.) MMI includes some of these companies but not all. But MMI does include Alpha Networks, which is an OEM, selling almost all its products under its own name.

MMI also counts revenue from bare board fabrication (Flex, Sanmina, etc.), which we remove. Yet MMI does not include companies such as ZDT, Mektron and MFlex which perform vast amounts of contract assembly although their primary business is flex circuit fabrication. We do so because their respective assembly operations are huge – in the billions. If you buy components and assemble them onto circuit boards for external customers, you are an EMS. To not include such firms is a huge omission.

I will say, it’s not easy to align companies’ EMS/ODM revenues, especially when they use different fiscal years, accounting standards, and reporting methodologies. Throw in the fact that most of these firms have businesses unrelated to pure EMS, some are private, and several report in languages other than English, and the task becomes very exacting. This certainly should not read as a criticism of MMI (or anyone else’s) methods, because 1. I know how hard it is to get everything right and 2. I truly respect the data that MMI (and others) add to the industry domain.

 

Filipino Fiasco

The situation in the Philippines is starting to feel a lot like that of Turkey from July of this year: A paranoid leader turned strongman seeks to exert his supreme dominance over a democratic nation.

Will the country stand back while Philippines President Rodrigo Duterte effectively declares martial law? Or will the army — no fan of China — assert itself and ironically return law and order to the nation by overthrowing a budding dictator?

There are more than 30 EMS companies in the Philippines, the largest of which include IMI, ranked 28th worldwide in EMS revenue at $800 million last year, EMS Components Assembly ($110 million), and Ionics ($63 million), which has seven plants there.

Other major players with smaller operations include Siix, Celestica, Cal-Comp and Wistron.

Although US-centric for decades, the Philippines under Duterte are pivoting toward China. US companies have invested nearly $5 billion in the country; even if they no longer feel welcome, extracting that won’t be easy.

Either way, politically the Philippines are a complete mess. Duterte has taken a stable nation and completely disrupted it, without any clear end-game. If his goal was to expand his nation’s markets and hedge its bets — understandable, given their neighborhood — he could have done so in a much simpler fashion. As it stands, he has alienated many of the Philippines primary trading partners, and for what? Business partnerships don’t have to be a zero-sum game. He could have ramped his dealings with Beijing without destroying his relationship with Washington.

Duterte likes to rail against the West for being what he considers corrupt and hypocritical would-be overlords. His own military very well appreciate the security the West historically has provided, however. His words might play well with the public, but he could very well pay the price with his life.

 

OEM/EMS Barrier Permanently Cut

For years we’ve been told that EMS companies are in the service business only and would never develop their own products. In one of the first interviews I did, back in late 1991, then IPC director Tony Hilvers — a leading proponent of the then-emerging CM industry (it wasn’t even called EMS then; that term was coined by Sue Mucha the following year) — insisted to me that contract assemblers wouldn’t go down the product development and branding path because it would put them in position of competing with their customers.

We can bury that old saw. With today’s news that Foxconn has, at long last, bought Sharp (for the low, low price of $3.4 billion), the loop between EMS and OEM has been drawn taut.

Not that this is ground-breaking in practice. Certainly, many, many EMS companies have, through acquisition or otherwise, developed and marketed their own products. Our 2009 EMS Company of the Year had a healthy, branded keyboard product line. And we estimated in this space in 2012 that 15 to 20% of the (then) 2,400 companies listed in our EMS directory did some degree of ODM/OEM work.

Going further, we wrote in 2015 we felt the line between EMS and ODM has been “permanently crossed.” But the Foxconn-Sharp marriage takes it to an entirely different scale.

Whether the Sharp name stays on its product lines, which range from Aquos televisions to smartphones to solar panels, and includes the OLED technology so prized by Apple that it compelled Foxconn to write the check in the first place, remains to be seen.

Either way, there’s no going back. EMS is now OEM. Going forward, who is the customer they will serve? And knowing the line keeping their suppliers from their end-customers has been permanently breached, will this spur OEMs  to reestablish their assembly operations?

Comparing Emerging Markets for EMS Production

How do various regions compare when it comes to locating EMS plants? Joe Fama, a veteran of 30 years with EMS companies around the world, wanted to explore this question.

He developed the initial table below, with some embellishments by CIRCUITS ASSEMBLY.

EMS Comparative Matrix

We thought it would be an interesting project to crowd source. As such, readers are encouraged to submit their ratings for any country with which they are familiar by clicking on this link and completing the short survey. There are 15 countries listed in all. We will continue to update this matrix as inputs are received. (Please note that all responses are anonymous.)

 

Will ‘OnCore’ Deal Spur Encores?

Fascinating how aggressive Natel Engineering has been with acquisitions over the past 18 months. First it gobbled up Epic, and this week it announced plans to nab OnCore. Epic was roughly 2.5 times the size of Natel at the time of that deal, and OnCore is almost the same size as Natel is now. Combined, they will form an EMS business with pro forma revenues of $770 million, 13 manufacturing sites and more than 3,700 employees.

And to think that as recently as September 2013, Natel had sales of $100 million spread across three factories, some of which were hybrid thick film, not SMT. That’s a stunning transition.

Can it hold? This latest deal is highly leveraged, and Moody’s gave Natel a B2 CFR rating, (obligations rated B are considered speculative and are subject to high credit risk; the 2 refers to mid-range) and a B1 LGD3 (loss given default) assessment (meaning ?30% and <50%, in Moody’s opinion). After the close, Natel will end up with $340 million in debt, between the new lender and a $60 million note issued by OnCore’s owner, Charlesbank Capital.

We’ve seen huge runups in the past, sponsored by equity capital, that have  burst into flames because the market couldn’t provide the necessary growth to sustain the acquirer’s debt payments. Viasystems is perhaps the most notorious example; that company ended up going through bankruptcy before finally stabilizing and operating in somewhat lower-key manner up until its announced acquisition by TTM Technologies last year. Flextronics went through one major flameout in 1990 before reappearing as a Singaporean company. Of the CIRCUITS ASSEMBLY Top 50 however, today most are few of undue private equity influence.

*********
For those wondering what EMS or PCB companies might be veering toward financial distress, here’s an interesting tool. I’m guessing Jabil ranks relatively highly on this because of its high exposure to Apple. Companies also seem to be penalized for a high P-E ratio.

Optimal at Optimum

As corporate models go, Optimum Design Associates isn’t unique. Several firms have launched as design shops only to add EMS capability as they matured.

But one thing ODA does that I really like is provide a much greater level of detail about its financials on its website.  See below:

Financial Disclosure

Financials through 12/31/2013

Ownership Private
Annual Revenue $19 million
Credit Access $5 million
Loan Balance $133,000
Principal Payments $3,700
Access to additional financing Yes
Executive changes last 90 days No
Ownership changes last 90 days No
Return on Assets (last quarter) 3.6%
Current Ratio 2.2
Quick Ratio 1.21
Capitalization 23%
Cashflow 5.45
Debt to Equity Ratio .003

 

For a privately held company, that level of public disclosure is unusual. For a small business, it’s unheard of. Certainly any OEM worth its salt would ask for verification of financial stability prior to engaging, but having that data on hand upfront makes determining whether ODA is the right fit a little easier for potential customers (which could also be a time-saver for ODA), and moreover offers a high degree of confidence that ODA will be above board.

Perhaps that’s willingness to break the mold is why ODA is again on the Inc. list of the fastest-growing companies. (It also made the 2007 and 2008 rankings.)

Congratulations to Nick and the gang at ODA.

Investigation at Fabrinet

Weirdness abounds at Fabrinet. Consider the following:

On Aug. 1, CEO David Mitchell sells 40,000 shares of company stock in a transaction valued at $734,000.

On Aug. 12, JDS Uniphase lowers its outlook, saying its current quarter sales will be as much as 8% lower than the consensus analyst forecast. JDS is Fabrinet’s largest customer, and one of two 10%-plus customers of the EMS firm.

Today Fabrinet announces it will postpone its fourth-quarter earnings release in the wake of an internal investigation into “certain accounting issues” uncovered by company management during its most recent fiscal quarter. The firm says it is also looking into whether there may be any “deficiencies” with its disclosure controls and procedures.

There’s no obvious straight line here. I’m hoping the timing of Mitchell’s transaction was just good luck, and that the investigation isn’t related to any insider shenanigans. Based on similar announcements from other industry companies, the investigation has something to do with the company’s inventory management. Such tightly sequenced events bear further watching, however.

 

A ‘Worthington’ Idea

EMS firm Worthington Assembly last week announced a deal to market its EMS services via CircuitHub.

WAI is a small EMS company located in Western Massachusetts. Like many in the sub-$20 million space, WAI’s owners double as its salesmen, and the firm relies heavily on word of mouth (and engineers changing jobs) for prospecting.

CircuitHub developed a universal parts library and is offering that, along with BoM, bare board and assembly quoting. PCD&F did a piece on the company last year.

Chris Denney, WAI’s CTO (and a sometime CIRCUITS ASSEMBLY columnist) explains the partnership here.

Clearly, more opportunities to order boards from a variety of suppliers via a single website are popping up, with the site typically offering free software in order to gain visitors (FabStream, for example, offers use of a PCB CAD tool capable of up to 12 layer boards, and SnapEDA offers simulation).

I would not anticipate larger EMS firms would go this route. But for smaller ones, whose cost of sales would be proportionally high relative to its income if it employed direct outside sales, using app-based vendors could be a creative and low-cost way to find new customers.

M&A is Here to Stay

There’s been a flurry of EMS acquisition activity of late, with Natel’s acquisition of EPIC Technologies and Benchmark’s pickup of Suntron and CTS among the larger deals. Lincoln International, an M&A advisor, counts nine transactions in the fourth quarter alone, out of 24 total for the year. While Lincoln’s numbers shouldn’t be considered absolute – my guess is that worldwide they are off by well over 50% – they do provide a reasonable snapshot of the industry at a given time.

While I dare say Nam Tai will be the largest EMS company to close its doors in 2014, when all is said and done, I predict we will see a record number of shops close or be bought out in asset deals.