Breakfast Bytes Blogs

Jay Vleeschhouwer did his annual...well, he did it last year, too...View from Wall Street in the DAC Pavilion. I've known Jay for years since I spent a couple of years being the person in Cadence who was house-trained on what I could and could not say, and I knew the whole product line in enough detail to cover anything a financial analyst needed to know. (I used to think I was one of the people that knew Jay well enough I could even spell his name, but then I messed it up in my post a couple of weeks ago.) Anyway, in those meetings almost twenty years ago with Jay and other analysts, one thing I knew not to talk about was anything to do with finance—that was other people's jobs.

It's the same today in this post, so let me emphasize that everything in this post is me reporting Jay's opinions on the EDA marketplace, and in no sense should be taken as official guidance. See our 10-K blah blah blah.

Jay started off by saying that there are two arms races in technology: software development and silicon development. Semiconductor companies still provide the bulk of the EDA industry's revenue, but system companies are increasingly doing design to support their strategies. They are a fast-growing part of the business that is helping support the EDA industry's momentum. Lots of tool categories are doing well, not just one or two. This is a contrast to much of the past where growth was often more narrowly focused.

This is the big picture. TTM (trailing twelve months) revenue. There was a big dip a decade ago in the financial downturn, and then steady growth. Cadence and the other major player together are now 2/3 of the industry, up from 52% in 2008 a decade ago. Now, Cadence and the three other players have nearly 90% of industry revenue. For 2019, Jay estimates growth of 5.3% to 5.9% to $7.95-8.0B.

Since one of the major players is now part of Siemens, Jay doesn't have accurate data on their financials. He has models, but his focus is on "the big two" including Cadence. He is, after all, a financial analyst and there is little point in spending too much time analyzing shares you can't actually purchase, except to the extent that they impact the ones that you can.

Combined big-two operating income in 2019 was $1.34B or 25.3% of revenues. Combined R&D was $1.977B or 27.2% of combined revenue (up from $1.765B in 2017). For 2019, he is estimating combined operating income of $1.524B or 26.9% of revenue, and combined R&D of $2.1B or 37.2% of estimated revenue.

Looking at geographies, the big story is that Asia (basically China) is now bigger than Japan and Europe combined (yes, I know Japan is in Asia but traditionally it has been treated as its own region).

Semiconductor R&D is a driver for EDA. A composite of 29 semiconductor companies showed R&D of $43.1B, up more than 6%. Total R&D spending (excluding a player that accounted for almost 1/3 of the $43.1B) was up 8% in 2018, with some much bigger increases that were up to 21%.

The stock market has been liking what it has seen. The market cap for the big two has tripled in just five years. The combined market cap is $36.1B or 6.3X estimated combined revenues for 2019, as compared with $26.3B at DAC 2018 or 5X the then-estimated combined revenues for 2018. Five years ago, the combined market cap was $10.3B.

This chart is Jay's estimate of emulation and FPGA prototyping revenue. The numbers are not fully disclosed by the companies. Note that one of the key EDA players also has a hardware business but is not included in this chart.

One reason that people like Jay focus on hardware is that it has such a different license model than for software, and you can't model an EDA company without estimating this. You have to know how much ratable business is done and what the average term is. Hardware has no term (except for maintenance) and is not recognized ratably.

IP has had an incredibly stable growth rate over the last five years. Note that this is just the big two, not any smaller IP companies. The combined big two's IP revenue exceeded $1B for the first time (trailing twelve months).

This shows the big two's operating margin. Cadence dipped down years ago, leading to a change of CEO. Since then it has climbed back and has exceeded the other major player’s operating margin since 2013.

To summarize, Jay said:

EDA has been doing well. It is provided by an increasingly consolidated set of vendors. Their role is becoming even more essential to semiconductor and now system companies. Microsoft, for example, has 1000 chip engineers, and is doing custom processors, Xbox, etc. System vendors are increasingly important to EDA.

We have recommended Cadence...for eight years.

Jay was asked in the Q&A whether he expected to see indigenous Chinese EDA companies. He said that it's taken three decades of R&D for the EDA industry to get to the point it is now, which is a big barrier to entry. Even if money is not an issue, time is. "I can't preclude it, but it takes a lot more than just policy. EDA companies can't assume there will never be competition, but it is a hard thing to replicate."

Sign up to get the weekly Breakfast Bytes email: