 |
Fabrinet CFO Mark Schwartz |
Don’t be surprised if you walk into a Fabrinet manufacturing plant
and recognize someone from JDS Uniphase. The Thailand-based EMS company
has purchased no less than six plants from JDS, one of its biggest
customers. It’s one big reason why, starting with its January 2000
acquisition of Seagate’s 230,000 sq. ft. plant in Bangkok, Fabrinet has
more than tripled its global manufacturing capacity and leveraged its
skills building disk drives into optical components and medical
electronics.
In February, chief financial officer Mark Schwartz shared some of the lessons Fabrinet has learned over the past six years.
CA:
During the past year you have made a number of acquisitions including
three JDS Uniphase plants. How did that transition go, and what have
you learned about taking over plants in Asia versus in North America?
MS:
JDSU has actively sought to restructure its operations over the past
several years. As their manufacturing partner, we have assisted where
our expertise in product transfer could accelerate the wind-down of
manufacturing operations. In December 2004 we acquired two JDSU
factories in Asia [in Singapore and Bintan, Indonesia]. In May 2005, we
acquired plants in Fuzhou, China, and in Mountain Lakes and Ewing, New
Jersey. We have since closed all the factories but CASIX in Fuzhou,
China, and VitroCom in Mountain Lakes. And we announced [on Feb. 15]
the acquisition of the manufacturing operations of its Ottawa plant and
we will wind this location down too.
In Singapore and
Indonesia, JDSU had great factories. We wanted to find a way to
maintain at least the design teams. The cost of operating the two
factories together was within 10 or 15% of our costs in Thailand. But
the products were similar to the products we were building for JDSU in
Thailand and it made more sense to drive to scale, and aggregate
production in Thailand. The personnel in Singapore and Indonesia were
very capable; they could have done the job too.
The CASIX
acquisition in China was interesting on a lot of levels. It gave us our
first footprint in China. CASIX fabricates base optical components,
namely fabrication and coatings for waveplates, laser crystals, optical
prisms, windows, lenses, mirrors and beamsplitters. We were their
seventh largest customer before we acquired them. Our customers today
tend to want to move up the food chain; for example, those at the
component level trying to get into higher level modules and even some
system builds. For us, it was an opportunity to get into a product area
that didn’t compete against our customers.
Similarly,
VitroCom is a producer of custom and standard borosilicate, clear fused
quartz, and synthetic fused silica glass tubing and rods. Both
acquisitions provide Fabrinet with control over critical sources of
supply by vertically integrating our supply chain. From a management
perspective, the GMs of both companies report directly into our CEO and
both GMs have weekly calls with Fabrinet’s management team to review
quarterly and annual objectives.
CA: How do you go about convincing former OEM plants to adapt to your corporate culture?
MS:
In every case where we have done this, JDSU has communicated openly
with its employees about the transition or the wind-down of operations.
At every opportunity we have to communicate with the employees prior to
closing, we try to demonstrate that we are mindful of the
circumstances, sympathetic to the impact it has on each employee and
their families, and, perhaps most importantly, that what we say is
honest and the best information we have at that time. We believe it is
important to earn the trust of each employee and make them comfortable
with their decision to become an employee of Fabrinet. So, for example,
we work hard to ensure that, where possible, all benefits and company
policies remain the same, such as salaries, pay scales, pay periods,
health and insurance benefits: we try to duplicate as best as we can
what currently exists. When we tell employees these things – that
salaries won’t change, that if you expect a pay raise you’ll still get
it – and follow through, that’s how we start to build the right
relationship. In most situations, I believe the employees really felt
that we cared, which we did and do. We strive to take the stress out of
the situation for each employee as best as we can.
So
every week, no matter where the site was, we’d travel to the site and
have an all-hands meeting. By the third or fourth meeting most people
felt what we were telling them was true, and by that time they’d seen
or heard from the payroll and benefits companies, which is critical for
their families. We always try to keep continuity in management, even in
China. The intent is not to revamp management but to use the best that
exists. In Singapore we borrowed one of the senior JDSU managers to
lead our team, because they recognized him and they knew him, rather
than bringing in someone totally new, which could have been even more
disruptive. In North America, we’ve brought in some outside folks that
have either joined the team or have special areas of expertise. In
Ottawa, complex manufacturing processes will be transferred to Asia and
Europe. So on our team we have a dedicated transfer director with over
20 years experience in process transfers. Another member of our team
has similar experience in the transfer of equipment and inventory.
We
approach each acquisition on two different levels. There’s the
technical level, the product and the transfer. And there’s the HR
aspect, which is certainly more critical upfront. We have to be sure
everyone is willing to be frank with us on potential technical concerns
and cost drivers such as product yield or manufacturing cycle time, as
well as any concerns we need to be aware of in the product transfer. So
to get the comfort level from the employees is critical.
On
the HR side, as factories wind-down, we often plan job fairs where we
bring in companies from the local area that might be hiring, as well as
provide career training and placement services.
CA: Characterize the differences between manufacturing in China versus Thailand.
MS:
Since we acquired our first factory in China in May of 2005, I don’t
know if we have enough experience in China to provide a comparison.
Communication in both geographies is the key. We’re still learning some
of the political and financial processes in China, like exporting and
importing goods. We had always heard to beware that the rules change
every day, but in our experience I don’t think that is actually true.
The rules are not very different from what we see in Thailand and other
places. You have to pay your taxes, keep accurate records, be a good
citizen.
One difference between China and Thailand is the
ease of obtaining visas in Thailand for travel to foreign countries. In
many cases, we ask our engineers and operators to travel to our
customer sites for training or support. Typically, our Thai folks are
able to obtain visas in shorter time and with fewer difficulties.
CA:
Five years ago optoelectronics was the next big thing. Then for most
companies it fell off the radar. Now that issues over automation have
been resolved, it appears to be making a comeback. Is that your
impression?
MS: I would suggest much less
emphasis has been put on automation. The demand, generally, for optical
communications components and modules, has not again reached the levels
of 1999 – 2000. Automation is volume driven. So companies performing
their cost-benefit analysis relative to automating process steps, have
in many cases opted not to invest in automation. We have not seen
significant deployment of automation in legacy products, although the
current generation of new products, showing signs of stronger demand,
is taking advantage of automation where cost effective. Certainly we
have witnessed improvements in fixturing, alignment, jigs, but overall,
the assembly line today doesn’t look much different than it did five
years ago.
CA: Fabrinet’s founder came from Seagate
and the disk-drive manufacturing business. What elements from that
business have been incorporated into Fabrinet?
MS:
Tom Mitchell was one of the founders of Seagate and its president from
the late 1970s through 1991. Fabrinet was his idea, spawned from a
visit in 1999 to an optical assembly plant in Taiwan. As he tells it,
the assembly lines reminded him of a disk drive manufacturing line from
the 1980s. Tom saw an opportunity to take the lessons learned from the
disk drive manufacturing world and apply them to optical communications
components and modules.
In terms of looking for the right
manufacturing environment, the assembly of optical components didn’t
require an SMT factory. What we needed were the engineering and
manufacturing skill sets from the disk drive industry that could be
leveraged into optical components and modules: cleanroom disciplines,
soldering and epoxy proficiency, mechanical assembly skills, process
engineering. When Tom acquired our first factory in January 2000 from
Seagate, along with approximately 1600 manufacturing engineers and
operators, we had almost no optical engineering knowledge. But our
customers were willing to train our employees because of the
manufacturing technologies we brought to the table. Of course, in the
past six-plus years we have acquired expertise in manipulating light
and assembling optical devices.
You do hear a lot of
comparisons between semiconductors and optics, though very few on our
team are from [the former]. By and large our manufacturing skill sets
came from assembling disk drives.
As an example, our
workforce had proficiency in soldering, from building disk drive
assemblies such as hydrodynamic motors. These motors required very
precise soldering and alignment tolerances that exceeded, in many
cases, the requirements for the assembly of optical devices. So when we
brought companies such as JDSU to Thailand for the first time to
understand our capabilities, they saw our proficiency in areas such as
soldering. It made sense that the transition for our skilled operators
from the manufacture of mechanical disk drive assemblies to optical
products would be easier than training folks from a standing start. Our
manufacturing processes really reduced the learning curve.
CA:
For certain leading-edge projects Fabrinet uses highly secure areas –
the so-called “factory within a factory” – that are separated from the
rest of the factory. Do these areas include any dedicated assembly
lines? Also, what restrictions, if any, does Fabrinet put on its
employees to ensure the IP of its customers?
MS:
IP protections are provided to every customer. Most of the protections
are driven by our customer’s demands, starting shortly after we began
operations, where Company A wasn’t so excited about having their
products manufactured in the same building as Company B. So we were
instructed that if we maintain proprietary protections A, B, C, etc.
then they would be comfortable with having their product manufactured
on the same manufacturing campus as their competitor. So in that
respect, our earliest customers drafted our IP protection policy. It
still varies from customer to customer. Documents may be protected
differently, engineering or manufacturing areas may have fogged
windows, keycard entry, dedicated employees, etc.
CA: You have a legal background. What’s your take on RoHS, and the liability of companies and employees?
MS:
I’m not in a position to comment on it. I know that our supply chain
spends a lot of time trying to understand the technical requirements
and the liabilities with our customers and supplier partners. Our
contracts with customers generally relieve us from liability for excess
materials, but that is a discussion we would rather not have with our
customers. Getting everyone on the same page with respect to when those
lead-free components cut in, for example, is more productive. You can’t
have any surprises with your customers.
CA: What lessons from all the acquisitions you’ve done can you share?
MS:
We’re constantly reflecting on and reviewing the lessons learned.
Particularly in a factory wind-down, you are organizing a company with
a short lifespan, and establishing and maintaining your corporate books
and records to reflect the nature of the organization. One critical
area is tax. Determining the allocation of earnings for the goods and
services created and delivered in each domicile, so all the revenue
collection organizations involved properly understand your tax
obligations and agree with your position. Another critical area is in
financial reporting. Our accounts are maintained under U.S. GAAP and we
review each transaction with our auditors. It can be a frustrating
experience, as GAAP rules do not always allow a company to portray a
transaction in a way that is most meaningful for shareholders and
others to absorb. It is something we struggle with and something we’re
learning to deal with each time. While the net profit and loss from any
particular transaction is accurately represented, the road to get there
has all sorts of twists and turns and footnotes and comments. I’m left
to wonder, on occasion, if GAAP reporting accurately portrays the
financial state of a company, particularly at the gross margin line.
I’m sure this is a concern many CFOs share.
Business
terms and conditions dictate accounting and tax treatment. But there
are often alternative business terms that are preferable from an
accounting and tax perspective, with little or no impact on the
business objective. At the front end of a transaction, we try to
structure the agreement documents to clearly tell the story of the
transaction for all downstream parties to understand. “Telling the
story,” often in the agreement’s preamble, assists in answering the
inevitable slew of questions from accounting and tax folks and can get
you, in a timely manner, to the proper accounting treatment of the
transaction. While the tail should not wag the dog, you must consider
the business objectives in parallel to the tax and accounting
treatment. In international transactions, every jurisdiction has
different laws in terms of notice obligations, severance obligations,
etc. Having both a legal and finance background is helpful to me in
anticipating issues and probable outcomes.