Having already built an impressive data center portfolio, in part through acquisitions, that included processors, storage, security, and Ethernet components, Marvell (MRVL) decided to go one large step further, announcing the acquisition of Inphi (IPHI) and its high-speed optical interconnect assets in a deal worth close to $8.7 billion. Marvell is paying around $8.7B to expand its served addressable market by about $3 billion a year, but adding Inphi should also create meaningful cost, development, and revenue synergies over time. Moreover, it takes a premier asset off the board, preventing another rival from acquiring it, and I wouldn’t dismiss the possibility of end-market growth exceeding current expectations.
I can’t fault Marvell’s ambition, but it wasn’t the company’s ambition or execution that concerned me coming out of the analyst day. Obviously, 5G infrastructure and data center offer very attractive multiyear growth opportunities, but there’s an increasingly high bar in terms of growth and margin performance to drive further re-rating.
A Major Deal In Data Center (And Wireless Infrastructure)
Between internal R&D and M&A (the Cavium deal in particular), Marvell has built an impressive suite of solutions for the fast-growing data center market. Adding Inphi’s top-tier optical interconnect portfolio fills one of the only major holes in that suite, giving Marvell the market leader in PAM4 DSP PHYs and 400G-ZR, as well as an underrated player in merchant DSPs and other components like drivers and TiAs.
Marvell is paying a high price to fill this important slot.
Marvell and Inphi agreed to a deal that will give Inphi shareholders $66/share in cash and 2.323 shares of Marvell for each Inphi share, giving them 17% ownership of the combined company. At pre-announcement prices, that’s an almost $158/share price for Inphi (the highest published sell-side fair value was $164), a 42% premium and 42.5x 2021 EPS.
Inphi would likely have been accretive to Marvell’s margins on a standalone basis, but Marvell management believes they can reap $125M/year in cost synergies. If management included this in its discussion of those synergies I missed it, but I would also note the scale synergies for product development – chip development at the leading nodes is getting increasingly expensive, and the added scale from this deal should be significant where that is concerned.
I also believe there are meaningful long-term revenue synergy opportunities. Only Broadcom (AVGO) has a similar portfolio of assets, and they have leveraged that successfully in the past to drive/hold greater market share and margins (not to mention optimizing product development/performance). This is admittedly difficult to quantify, but I believe it will be significant for Marvell as that company looks to gain share in data center and 5G wireless infrastructure.
I see minimal risk of the deal running into antitrust problems, and I think execution/integration risk here will be relatively low. Truly maximizing the potential of Inphi’s assets (particularly on the synergistic development side) could take some time, and Broadcom is likely to gain share in PAM4 next year, but I’m not worried about Marvell’s ability to integrate and leverage this deal.
I’ll be very interested to read the deal proxy to see the process that led to this deal. Given the scarcity value of Inphi, as well as the premium Marvell is paying, it wouldn’t surprise me if there was at least one other interested party. To that end, with Inphi off the board, there really isn’t a great list of independent names for investors to leverage the trends that were driving Inphi. Yes, there’s MaxLinear (MXL), but MaxLinear really isn’t in Inphi’s league (or Broadcom’s) in PAM4. Likewise, names like MACOM (MTSI) and Semtech (SMTC) offer some leverage to optical, but they’re not on the same level as Inphi.
Turning back to Marvell, I think deal accretion will likely be relatively modest given the premium paid, but Marvell will emerge from this deal as a more profitable, faster-growing company, with even more leverage to some of the most attractive markets in the semiconductor space. Given that the market has clearly shifted back to paying for growth (versus margins), that likely won’t hurt Marvell’s valuation so long as that growth-over-margin paradigm stays in place. I also don’t think that Marvell is necessarily done with M&A. While there aren’t any big gaps in the data center portfolio, I believe the company may still want to add some select optical capabilities/IP to fill out its portfolio even further.
Although I didn’t own Inphi, I have to admit I was watching the recent sell-off with interest, hoping I might get a chance to establish a “starter position”. This is where, if Seeking Alpha allowed GIFs, I’d insert that well-known “frustrated Skeletor” GIF, but I can’t say that I didn’t have my chances – anybody could have bought Inphi below $80 within the last year.
I won’t speak for how Inphi shareholders should feel about getting Marvell shares in this deal, but I imagine at least some investors will have mixed feelings. Not unlike how the Nvidia (NVDA) offer for Mellanox started looking less attractive as time wore on and the market for high-end interconnect continued to outgrow expectations, I can understand if Inphi shareholders feel as though, even with the high premium Marvell is offering, they’re getting sold out at least a little short of what Inphi could have become on its own. At least the Marvell shares will offer some opportunity to participate in that upside.
The Bottom Line
I already thought Marvell was highly valued (the shares are down about 13% from my last update earlier this month), and paying a high premium for another highly valued company doesn’t create some alchemical magic that makes Marvell cheap. By the same token, investors are clearly willing to pay up for growth stories, particularly in areas like 5G infrastructure and data center where underlying demand remains very strong, and Marvell continues to burnish its above-average, if not “exceptional”, growth credentials. I have real questions about the long-term annualized return potential from Marvell around $40, but so long as growth remains in favor, it won’t really matter in the short term.
Disclosure: I am/we are long AVGO. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.