I dont claim to understand the EDA business, but as a one-time CDN holder (out at 29) with a continuing interest, I submit this analysis from todays NYT tech section...
April 22, 1999
Cadence Stock Crash Raises Hard Questions
Filed at 12:05 p.m. EDT
By Richard Goering for EE Times, CMPnet
After stunning Wall Street by saying it expects no revenue or
earnings growth on existing business in 1999, Cadence Design
Systems' stock plummeted by nearly 40 percent Wednesday.
While Cadence said it's simply repositioning itself for future gains,
analysts and competitors said they believe the EDA revenue leader
may have stumbled badly on the road to very deep submicron
design.
San Jose, Calif.-based Cadence made its surprise announcement in a
conference call following its first quarter 1999 report, which showed
revenue of $305 million, a 13 percent increase over the previous
year. That's a little less than expected, but it was the prediction of flat earnings for all of
1999 that caused Cadence's stock, which has been as high as 39 in the past year, to close at
12 5/8 Wednesday.
Jack Harding, president and CEO of Cadence, said the company decided to rebuild backlog
that was eaten up in 1998 because Cadence's services business was underperforming. He
also said the company has run into a "one- to two-quarter delay in absorption of 0.18-micron
design tools" among semiconductor makers.
While observers agree about Cadence's need for backlog, most strongly disagree that there's
any slowdown in demand for 0.18-micron tools.
"Everybody is moving to 0.18 microns," said Gary Smith, chief EDA analyst at Dataquest
Inc. (San Jose). "They [Cadence] are making comments that give most analysts the
impression they don't seem to know what's going on in the real world."
"We don't see any delay in the transition to 0.18 micron," said Paul Lippe, senior vice
president of Synopsys, in Mountain View, Calif., the EDA industry's second-largest
provider, after Cadence. "Cadence made some comments about the overall industry we don't
think are true."
The real question, observers said, is whether Cadence can effectively compete in a
0.18-micron-and-below world that will require a complete overhaul of existing tools.
Cadence's stronghold in physical design is under threat not only from traditional rival Avant,
but also from Synopsys, Mentor Graphics, and a host of start-ups with new technology
aimed at very deep submicron design.
There's also the question of whether design services is a sustainable long-term business for
Cadence. Harding acknowledged margins on services have been too low, and he said the
company has decided to take on only high-margin accounts, a move that will reduce revenue,
but ultimately increase profits.
So far, Cadence's woes don't seem to be extending to other EDA vendors, though Synopsys
stock slipped nine percent Wednesday, an apparent reaction to the problems at Cadence.
Avant had a strong first quarter report, as did Quickturn Design Systems, which Cadence is
set to acquire.
Harding emphasized Cadence's prediction of flat revenue applies only to the existing part of
the company, not to any acquisitions. With the addition of Quickturn and possibly other
companies, he said, Cadence will still show revenue and earnings growth for 1999.
"We're still going to do $100 million in profits every quarter this year," he said. "We're
trying to poise ourselves for the kind of expansion we can achieve into the billions. We gave
the street all the heads-up we could that we were doing to do some different things
strategically for the next three quarters for a long-term gain."
Harding characterized the 40 percent drop in Cadence's stock price as a gross overreaction.
"What happened today was not about math -- it was about emotion," he said.
If analysts were emotional, surprise was definitely a factor. "They committed one of the two
cardinal sins on Wall Street -- their guidance changed materially," said Erach Desai, analyst
at Credit Suisse First Boston. "When they restructured their services business in November,
they had given guidance of 20 percent product growth, and 50 percent services growth."
"My model went from 17 percent year-to-year growth to three percent," said Jennifer Smith,
analyst at BancBoston Robertson Stephens, one of several investment firms that cut
Cadence's ratings. Smith said she now expects a slight decline in product revenue, service
growth around 30 percent, and no growth in maintenance for 1999 -- and services, she said,
are only about 25 percent of Cadence's total business.
Dataquest's Smith said the analyst conference call was "a disaster," and that it's not yet clear
whether the stock crash was an overreaction. "Basically, Cadence is projecting the image that
they don't have a clue what they're doing or what's going on in the real world," he said.
On the product side, Harding said, Cadence's bookings are up 20 percent -- it's the decision
to build backlog that results in flat revenue. But he acknowledged the halcyon days of 80 to
100 percent annual growth in IC CAD are over for now, and not likely to return until
complete 0.18-micron design flows are available.
"What we saw is the Silicon Ensemble product cycle came to a crashing end," said John
Barr, analyst at Needham and Co. Desai also said star Cadence products such as IC
Craftsman and Silicon Ensemble have "petered out," and Cadence is losing market share in
physical verification to Mentor Graphics.
Can Cadence compete in the 0.18-micron arena? Most analysts said Cadence has some good
technology in the wings, and needs to get it into customer's hands as quickly as possible to
rebuild confidence. "They have some image issues," said BancBoston's Smith.
Harding said Cadence will ship a complete 0.18-micron design flow, "from authoring
designs right through physical design," by the fourth quarter of this year. "We have a design
flow that is integrated from a software perspective, we've just about frozen code, and we're
going into the validation phase," he said. "Checking five to 12 products, all tied together
with a single timing engine, is about as much effort as writing the code."
Dataquest's Smith said Cadence has strong technology and will make "one of the more
important" EDA announcements of the year in June. "The only technical question, in my
mind, is their IC layout tools -- can they put out a timing-based tool set in time?"
In addition to developing new tools, Cadence must rebuild depleted backlog. "We used a
tremendous amount of backlog in 1998 to make our bottom line, to account for the
under-performing services business," said Harding.
Related to the backlog question is Cadence's controversial three-year flexible access model
(FAM) licensing, which analysts said provides immediate revenue recognition at the expense
of the final two years.
"They've got to put in a firm plan to cut away from FAM," said Desai. Harding said Cadence
will supplement FAM licensing with an approach that takes in revenue every quarter over a
12-quarter period.
Meanwhile, Cadence will attempt to make its services business more profitable by selecting
accounts carefully. Last year, Harding said, "we were ramping the business quickly, there
were lots of innovations, a lot of new people had to be trained. Now we're only taking
transactions that fit us strategically."
Cadence's services model, which resulted in large-scale layoffs last year, has long been
controversial in the EDA industry. The company's recent foray into the industrial-services
market didn't help, said Dataquest's Smith. "I think everybody said, why did they get into
that stupid business?" he said.
But there's still no denying Cadence, with its $1.2 billion annual run rate, is far and away the
EDA industry revenue leader. According to the Electronic Design Automation Consortium,
Cadence held 37 percent of the overall EDA market in 1998, and enjoyed 33 percent revenue
growth. The upcoming battle for 0.18 micron and below design will determine whether
Cadence can hold onto that kind of lead. |
