|C. Dayakar Reddy, MD, MosChip:
Betting on a contrarian strategy
in Hyderabad's Tony Banjara Hills borough is one of India's hottest
semi-conductor companies. Another reading of that sentence isn't
warranted; you read it right the first time and I can produce a
doctor's certificate to prove that I wasn't under the influence
of any mind-altering substance when I wrote that.
If you'd still like a raison d' etre for this
story, try three for size: the company in question, MosChip, is,
arguably, the hottest listed little start-up in the country; it
is promoted by three Valley vets who, between them, have started
up and sold four semiconductor companies, one to National Semiconductor;
and it has got enough going for it to have had two electronics manufacturing
service biggies (you know the kind-they make products for other
companies), ESS Technologies and Flextronics, to have invested in
Chip design companies are as common as the
darshinis vending idlis in the capital of Nerd Nadu, Bangalore.
Karnataka's redoubtable it Secretary Vivek Kulkarni can give you,
if your interest runs to listings, a roster of 48 that have made
a home in Bangalore. Many of the 48, though, are into services;
MosChip is on holier, and more profitable ground, products-branded
products. In Indian technology companies' quest for that Holy Grail,
MosChip has made a start of sorts. Surely, that's enough to merit
#429 four pages of print, the same as #1 Bharti Tele-Ventures in
| THE UPSIDE
The three promoters are old Silicon Valley
hands with industry experience
Investors include prominent electronics companies, ESS Technology
Has a 16-product pipeline that's waiting to hit the market
The products are designed on existing standards and hence carry
If the products work out, MosChip could be a potential takeover
| THE DOWNSIDE
Is a relatively late entrant in the business,
and does not have a strong parent to back it up
The business is capital intensive and needs superlative technological
skills to launch products
Despite the impressive product portfolio, ICs are mass-market
products with low margins
Also, the economic lifespan of a chip is short, averaging at
24 months. Sales volumes will be important
Stock price of Rs 21 underlines the market's concern over the
lack of positive cash flow
It's hard to miss MosChip's offices on the
Banjara Hills, Road 12, a three-storeyed light-green building with
the name MosChip emblazoned in red on one side. This is the Gen
HQ from which the company will set out to conquer the world early
in 2003 when it launches the first three of the 16 products in its
pipeline now; every passing quarter will see the launch of two more,
an arithmetic progression that Moschip hopes will wipe away all
traces of red from its balance sheet. That, there's aplenty.
MosChip was founded in July 1999, by three
men who returned from the Valley, Ramachandra Reddy, C. Dayakar
Reddy, and Vinay Kumar. In January 2001, the company raised Rs 10.50
crore through an initial public offering. Soon after, in July 2001,
it acquired, for Rs 39 crore, a US-based company that had been founded
by its founding trio, NetMos, and renamed it MosChip US. The acquisition
gave the company six ready products, but it is the sixteen in the
offing that could propel it to technology greatness.
Anticipation of that, it is, which makes MosChip
a company to watch. There's little else: its revenues dipped from
close to Rs 2 crore in 2000-01 to a mere Rs 34 lakh in 2001-02;
net profits, from Rs 32 lakh to a negative Rs 4.38 crore. The numbers
don't look pretty but that could well be, as Managing Director Dayakar
Reddy reasons, because they do not incorporate those of NetMos.
On a consolidated basis, MosChip boasted sales of Rs 7.7 crore and
a net loss of Rs 3.8 crore in 2001-02.
This isn't one of those instances of unscrupulous
promoters launching a company, taking it public, finding some investors,
and then, acquiring a company they had founded earlier at a sizeable
premium. What it is, instead, in MosChip's case, is a high-risk
product play. Before it acquired NetMos, a significant portion of
MosChip's revenues came from design services it offered to the company.
Post acquisition, the company exited the segment-although it remains
lucrative-and moved all its engineering resources into product development.
"We have taken a temporary hit to focus on products,"
says Reddy. Just wait till next year, "when our products come
out." Only, the road to creation is long and expensive; for
the three months between April and June 2002, MosChip returned a
net loss of Rs 2.78 crore on sales of Rs 1.23 crore.
The Profit Zones
Just what are these products MosChip is betting
so heavily on? For the technically inclined, the company designs
products that facilitate high-frequency data communication, interconnectivity
and wireless applications. And oh yes, it also does high-end systems
on chip. For people like us, MosChip's products target companies
that manufacture peripherals, PCs, and consumer durables. There's
no denying the existence of a market for these products. "Based
on consumer feedback, we see the addressable market for a company
like MosChip at $200 million (Rs 980 crore)," says Dayakar
Reddy. Only, the company will need to keep its product pipeline
flowing-the 16 in the offing should take care of it for some time-and
build inventories. High-end as its products are, there's more money
to be made in the volumes game than the value one.
"We are not getting into radically new
markets but are instead focusing on established standards,"
says Reddy. Keeping up with relevant standards will pose a challenge:
product lifecycles in the semi conductor industry rarely exceed
30 months; the norm is 24. And each new product could cost anything
between $200,000 to half-a-million dollars (Rs 1-1.25 crore).
| THE CHIP CASTE SYSTEM
No, we aren't caste-conscious, but MosChip's business
model does make it a Brahmin among India's chip designing
There are chip companies, and
then there are chip companies. MosChip is a fabless chip company.
That means it manufactures its own products but outsources
the silicon wafer processing to large foundries like Taiwan
Semiconductor Manufacturing Co. (TSMC). That gives it an edge
over chipless chip companies and chip design services firms.
The former, also termed chip IP companies, do not build complete
chips but design and licence out parts of them (or modules)
to other firms. Intel, TI, and other integrated device manufacturers
increasingly outsource designs and IP from chipless firms.
The emergence of fabless and chipless firms has pushed plain
vanilla chip design companies-they dealt not in IP but in
chip design services-to the background. Still, most analysts
reckon fabless is the way to go (See Fabless Rules).
» Foundries can make huge capital
investments and leverage economies of scale. Ergo, fabless
firms can source chips at volume prices.
It costs upwards of $3 billion to build a fabrication facility,
and there's no telling whether this can be run to capacity.
Fabless companies won't suffer from excess capacity or inventories
when the next downturn in the semi-conductor industry comes
The upside? MosChip will operate in the profit
zone of designing and marketing chips to customers such as Sunrich,
Hong Kong-based company engaged in the business of cards, cables
and other devices and Megarich, a Taiwan-based company that deals
in peripherals, cards, and connectors, and if all goes well, a clutch
of consumer electronics majors some of whose products adorn your
homes. The commodity business of manufacturing the chips (or the
foundry work, as it is called), the company will outsource to Flextronics-that's
one reason for its investment in MosChip.
Its peers believe MosChip has what it takes
to make it to the big league. "What is admirable about MosChip
is that it has entered the system-level design (market), right from
the beginning," says Bishwadip Mitra, the Managing Director
of TI India. His reference is to the traditional route adopted by
Indian chip-design wannabes: a little outsourcing work for larger
chip design companies, then a little more, until, gradually, the
company becomes a full-fledged chip design one itself. Even then,
few are willing to risk branding and marketing their chips like
MosChip will do. "It is not a cost play any more; it is a very
strong IP (Intellectual Property) play," adds Mitra. Purely
on the kind of competition it faces, MosChip is on to something
big. These include, in the peripherals space (let's just call it
chips for printers), Taiwan's ITE and UK's Oxford Semiconductors;
and in data communication, TI, Cypress, and Philips. A combination
of price-play and value add, such as free system design (buy our
chips and we'll help build you a device around it, is the promise),
explains a Mumbai-based analyst who has been tracking the company,
should help MosChip hold its own against the competition.
Marketing can make or break the company. Unlike
chip-design hotshops that merely have to go out there and seek business
from a chipmaker that wants to outsource part of the design process,
MosChip will have to find takers for its chips. Founder Ramachandra
Reddy has considerable experience in international marketing in
the semiconductor industry, and that should help.
India's National Association of Software and
Service Companies (NASSCOM) believes the Indian chip design industry
today is placed exactly where the it services one was a decade ago.
The transnational heavyweights are all here (TI, Motorola, ST Micro,
Cadence, Synopsys), as is a host of Indian and MNC startups. The
first reports about the scarcity of microelectronics pros are just
starting to make the rounds but hey, things are just about getting
|Purely on the kind of competitors it faces-these
include the likes of ITE, Oxford Semiconductors, Texas Instruments,
and Philips-MosChip could be on to something big
The Next Infosys?
If MosChip is really as hot as we think it
is, what can explain its stock price (Rs 21 on October 19)? "We
are doing well even from the stock market point of view," argues
cfo Vivek Bhargava. "We have some reputed investors, but the
company has yet to deliver."
He can say that again. Still, MosChip hopes
to clock revenues between $35 million (Rs 171.5 crore) and $50 million
(Rs 245 crore) by 2005. That's not bad for a $2 million (Rs 9.8
crore) firm. And the first profits, claims Bhargava, should come
by the first quarter of 2003-04. Premature as that may seem, it
is a possibility: investments in the semi conductor business are
invariably front-ended; and a smart volumes player can recoup them,
and turn a tidy profit before the 30-month deadline.
The market doesn't have much to go by. When
it looks at MosChip it sees a company making huge losses; one that
burned some money acquiring a company founded by its promoters;
and one that operates in the esoteric world of fabless chip-making.
When we look at it, we see the possibility-just that, nothing more-of
it being the next Infosys.