Reviving The IPO Route For IP Companies

K. Charles Janac, chairman and CEO of Arteris IP, sat down with Semiconductor Engineering to talk about the company’s recent decision to go public, including the benefits and risks of operating as a public IP company.

SE: The rule of thumb used to be $20 million in revenue was needed for an IP company to do an IPO at the turn of the Millennium, and then it increased to $40 million about a decade ago. What’s the number today?

Janac: If you want to be interesting to a low- to mid-cap fund, you need a minimum valuation of about $320 million, and preferably closer to $400 million. You need the revenue multiple, or the profit multiple, to get up to that valuation. An IP company typically is valued at 8X to 10X trailing revenue, and you need around $40 million or $50 million in revenue. If you have a lower multiple, then you have to be that much bigger. So if your industry multiple is 5X, then to reach a valuation of $400 million your revenue needs to be somewhere around $80 million.

SE: What’s the rationale for those numbers?

Janac: Below that, small- to mid-cap funds cannot have sufficient liquidity, so you fall into the micro-cap funds, which is a much thinner market.

SE: In the past, there were four companies specifically developing network on chip (NoC) technology — not including Arm, which has a much broader portfolio of IP. Three of them were bought by bigger companies, right?

Janac: Actually, there were four acquisitions, if you include our Qualcomm transaction. But recently it was Intel-NetSpeed, Facebook-Sonics, and Google-Provino. The system IP is getting so complex that it’s much cheaper to buy someone who has at least a partially proven technology than to build it yourself.

SE: So why did you decide to take the IPO route?

Janac: There are several reasons. One was that we really wanted to assure our customer base that we don’t really want to sell. We want to run this as an independent company and to grow it, and we want an independent infrastructure to which everyone can connect. Second, system IP is getting very complex, very R&D-intensive, and we really want to make some complementary acquisitions to augment our organic R&D investment. Third, all this regionalization of the semiconductor industry, the democratization of SoC design, and the fact that system companies are starting to build SoCs and dis-intermediate the supply chain has resulted in a tremendous shortage of semiconductor engineers — particularly interconnect engineers. We know there’s a shortage because people are trying to take ours. So it’s going to be helpful to have liquid RSUs (restricted stock units) for the purpose of retaining employees and attracting new ones.

SE: All of these shifts also position you well for the forthcoming chiplet market, as well, right?

Janac: Yes. Before Arteris, it took tens of days to build an interconnect. With Arteris, you can build an interconnect in days. Where we need to go next is providing the ability to build it in hours.

SE: That’s probably a pretty significant investment. How do you get there?

Janac: Part of the definition of system IP is there’s this NoC interface IP segment, and these are things that connect to your NoC. We just have to decide what is the most strategic thing to connect to the NoC. Security is obviously on the list. But the most important, as far as I’m concerned, is chiplets. In the past, people were asking us to make a multi-core processor subsystem look like a single program memory programming space. Now they’re asking us to make multiple dies look like a single programming space.

SE: Could you have done this without going public?

Janac: Possibly. By going public you get a bunch of capital. But you also get stuck with a bunch of profitability requirements. So just because you have cash doesn’t mean that you can run huge losses. You’ve got to manage [financial] quarters while you’re doing all these strategic initiatives.

SE: Does this leave Arteris IP more vulnerable to an acquisition now because it’s public?

Janac: No, and that’s why you build in a poison pill defense as part of the IPO structure. If you don’t want to be acquired, you basically trigger the poison pill, which triggers massive dilution. So you can defend yourself from the very beginning against against an acquisition. On the other hand, the minimum price you can get away with without being sued is whatever the market price is plus a 30% control premium. You’ve essentially established a floor.

SE: An IPO gives you the ability to borrow more, right, because you have a higher market cap?

Janac: Yes, you have access to the capital markets and you can do secondaries. We have no debt, so that gives us additional borrowing power. But having too much capital is as bad as not having enough.

SE: What other potential issues does going public create?

Janac: To go public you have to take on a much more higher cost structure. You need a much bigger finance team and a much bigger IT team. You need to put in HR. You wind up with significantly more infrastructure. The SOX (Sarbanes-Oxley) compliance costs are a couple million dollars a year. You’re scaling up part of the company that doesn’t really generate any value for the customer. From a product point of view, you have to be of sufficient size to carry that.

SE: What happens on the technology side? How does an IPO impact that?

Janac: Now you have some constraints about what you can fund. As a private company you can say, ‘We need to build this product so we’re going to run losses for a couple years.’ As a public company, you need to think about that much harder. But the engineering doesn’t change that much.

SE: It’s been years since we’ve seen an IP company IPO.

Janac: Yes, the last one I remember was in 2002.

SE: As a public company, are you more open to scrutiny for trade with countries like China?

Janac: Not really. Some of our customers were banned. We have Magillem software, which is of French origin, and we transferred FlexNoC to France because it’s a French-origin product. We’ve diversified an export risk. More than 50% of our revenue stream is coming out of France and subject to French export rules, and maybe 40% to 45% is subject to U.S. export rules.

SE: Still, Arteris is raising its visibility by going public. Does this put you above the radar for companies looking to get into lucrative markets? Right now there’s only one independent, pure-play NoC vendor.

Janac: In many IP markets, that’s the case. The outsized returns accrue to the winner, and system IP is one of those markets. On the other hand, the market is big enough for others, as well. It’s possible we will be facing some new competitors over the next three or four years. That said, the barriers to entry are extremely painful in system IP.

SE: But an IPO does make it easier to acquire other companies, right?

Janac: Totally. Acquisitions as a public company are much easier because instead of trying to solve an equation with two variables, you’re trying to solve an equation with one. Everybody knows what your stock is worth because the public markets expose that. So all you have to worry about is the value of the acquired companies. M&A becomes much easier with a public platform.