A Stronger Beat for Cadence

S&P sees this maker of chip-design software improving smartly in 2004 as the semiconductor industry picks up the pace again

By Richard Tortoriello

Shares of Cadence Design Systems (CDN ; recent price $15) have lagged this year as its archrival Synopsys (SNPS ; $31) appears to have stolen a march on the electronic design-automation (EDA) software outfit thanks to a strategic acquisition. But we at Standard & Poor's remain bullish on Cadence, as we believe that it's in store for a strong 2004.

We also believe the shares can be expected to outperform the overall market as the semiconductor industry enters the beginning stages of a cyclical upturn. Cadence carries S&P's highest investment recommendation of 5 STARS, or buy.

EDA software is used to create and test designs of semiconductors and, increasingly, entire electronic systems, comprising semiconductors and the printed-circuit boards into which the chips are embedded. Cadence is neck-and-neck with its primary competitor Synopsys (which is ranked 3 STARS, or hold) for the leading position in the EDA software market. Mentor Graphics (MENT ; 4 STARS, or accumulate; $15) is a distant third, followed by a number of much smaller competitors.


  Cadence has historically held a strong position in the so-called place-and-route (or physical) segment of digital integrated-circuit (IC) design, in which a previously defined logical IC design is translated into a physical layout that maps each transistor and the conductive lines ("wiring") that connects transistors. It's also a leader in custom, or analog, IC design.

Synopsys has occupied the top position in logic synthesis, or the logical design of a semiconductor. But Synopsys moved to capture some of Cadence's territory with its June, 2002, acquisition of Avant!, Cadence's main competitor in the place-and-route market (see BW Online, 8/18/03, "A Promising Story Line for Synopsys").

The acquisition increased both Synopsys' size and product breadth. With the integration of the two companies going well -- and Synopsys' results improving at a faster pace than Cadence's -- industry observers began to speculate that Synopsys was gaining market share. For the 12 months ended June, 2003, Synopsys' sales were $1.148 billion, while Cadence's were $1.135 billion. For the 12 months ended June, 2002, just prior to the Avant! acquisition, combined sales for Synopsys and Avant! were $1.106 billion vs. $1.430 billion for Cadence.


  We attribute the shift in revenues, from a decisive lead by Cadence to relatively equal results, to three factors. First, in mid-2002, Cadence began to see a sharp increase in subscription sales, which allow customers to pay for software over time (usually three years). They also delay revenue recognition. This means about two-thirds of subscription sales are recognized in future years, reducing current revenue. We think this factor accounts for the lion's share of the revenue shift.

Second, we believe Cadence lost some market share to Synopsys as a result of Synopsys' increased competitive position (a newly integrated logical and physical digital-design product). Finally, we think some of the change, as well as differences in 2002 bookings between Synopsys and Cadence, is accounted for by differences in renewal cycles between Cadence and Synopsys/Avant!.

However, we still maintain that Cadence should do well in 2004. Here are our reasons:


  As it converts its revenue model, Cadence has gone from 48% subscription sales in the second quarter of 2002 to 82% in the third quarter of 2003, resulting in a strong revenue backlog. Cadence expects that 70% of product sales in 2004 will come from the backlog. We believe this will add a significant amount of predictability to Cadence's sales.

We think it has been making significant enhancements to its product lines, making them more competitive with Synopsys' offerings. In particular, Cadence has been releasing a new version of its Encounter digital design platform about every four months. Based on customer tests, Cadence says it has the technological lead with Encounter. It claims that tests have shown a three-fold design-time advantage over competitors (e.g., a design that took three days to complete on a competitor's platform took 24 hours on Encounter).

Just as important, however, Cadence has been improving the ease of use and integration (ability to mesh with other products) of Encounter and expects to be on par with the competition -- Synopsys and Magma Design Automation (LAVA ) -- within six to nine months. With customers already evaluating advance versions of this program, we expect Cadence's competitive position in digital design will begin to improve.


  Cadence is winning new customers with its Incisive design-verification platform, and we believe it's very competitive in this large and growing market. Cadence also released the Virtuoso analog design platform, which we expect to boost its already-strong analog design sales. Virtuoso can be used together with Encounter to design the growing number of chips that have both analog and digital content. Early in 2004, Cadence plans to release a platform for printed-circuit-board and chip-packaging design.

Based on its product portfolio, we believe Cadence is well-positioned for the future. Semiconductor companies like to have two viable suppliers for any given product in order to keep the suppliers competitive, and we believe that the "easy" market-share gains Synopsys may have made as a result of the Avant! acquisition are over.

Finally, we note that Cadence has been substantially improving its operating margins. While sales for the first nine months of 2003 were about flat with the same period of 2002, selling, general, and administrative costs dropped 16%. We project that operating margins for 2003's fourth quarter will be 27%, vs. 5% in the fourth quarter of 2002, and we forecast operating margins of 24% in 2004, vs. 16% in 2003.


  We also note that recent reports from chipmakers, such as Samsung, Intel (INTC ), Texas Instruments (TXN ), and STMicroelectronics (STM ) demonstrate that worldwide chip sales continue to improve. We expect them to rise 12% to 14% in 2003 and by nearly 20% in 2004. While chipmakers continue to be very cautious in their spending, we expect EDA outlays to increase at least modestly as chipmakers' profits improve.

We believe that all of these positive factors will favor Cadence in 2004. We forecast earnings per share, excluding amortization, of 80 cents, up from our estimate of 48 cents for 2003 (this year's estimate also excludes restructuring and other charges). We're forecasting 13% sales growth for 2004, with most of the gains expected to come from the increase in the subscription backlog.

From a valuation standpoint, the shares recently traded at 19 times our 2004 EPS estimate, or a price-earnings-to-growth (PEG) ratio of 1.3, moderately above an estimated 2004 PEG of 1 for the S&P MidCap 400 index, of which Cadence is a member. We believe its premium to the index is justified because it's a cyclical stock and can be expected to outperform in the beginning stages of a semiconductor industry upturn.


  On a price-to-sales basis, Cadence recently sold for 3.8 times sales over the prior 12 months, in line with its 10-year historical average. Past cyclical peaks in price-to-sales have occurred at between 5 and 8 times sales. Our 12-month target price of $20 is based on a multiple of 5 times sales, the low end of this range. We note that our target price is based on trailing sales, and as such we view it as conservative, given our expectation of 13% sales growth in 2004.

Our discounted cash-flow model provides an intrinsic value range of $17 to $21. Our model assumes an average growth rate of 11% over the next 15 years and a perpetual growth rate of 3%. Our cost of capital estimate is 13.5%.

On a Standard & Poor's Core Earnings basis, we estimate a loss per share of 46 cents for Cadence in 2003 and forecast a loss of 4 cents in 2004. Like most software companies, Cadence issues a significant amount of stock options to compensate its employees. In addition, it has acquired a number of companies in recent years and as a result records substantial noncash amortization charges.

Like most technology stocks, Cadence is highly volatile and therefore more subject than other stocks to general market risk. Additional threats, in our view, include the potential for a loss of market share, a downturn in electronics and semiconductor markets, as well as the fast-paced nature of software development combined with lengthy development and sales cycles.

Analyst Tortoriello follows semiconductor capital equipment stocks for Standard & Poor's Equity Services

Before it's here, it's on the Bloomberg Terminal.