Sundry Photography
In the not-always-rational world of Wall Street, I’m actually a little surprised that the Street hasn’t adopted more of an expectation of ongoing beat-and-raise quarters from Broadcom (NASDAQ:AVGO). Then again, given that outlooks are withering at many peer semiconductor companies, I suppose that there’s little inclination to look a gift horse in the mouth.
Broadcom shares have rebounded about 20% since my last update on the company (only about a month and a half ago), and while that is better than the performance of the SOX index (up about 16%) and certain select peers (like Marvell (MRVL)), a few like Microchip (MCHP) and NVIDIA (NVDA) have done even better. Pull the comparisons out a year, though, and Broadcom is handily outperforming all of those except Microchip (which it has outperformed, but not by a wide margin).
I still view Broadcom as a core holding. I love the company’s leverage to high-end networking/data center demand for connectivity and AI acceleration, as well as leverage to home broadband, and I believe businesses like wireless, storage, and software are still more than worthwhile over the longer term. Roughly 20% below my fair value estimate and priced for a high single-digit long-term annualized return, this remains an attractive idea in my view.
A Beat Is A Beat
Broadcom didn’t blow away sell-side expectations for the fiscal fourth quarter, but given reports and guidance from peer companies, it was still a better-than-expected performance and the pace of business seems to be holding up better than for most (something I’ve expected and previewed in prior pieces).
Revenue rose 21% year over year and 5.5% quarter over quarter, coming in just ahead of expectations. Semiconductor revenue rose 26% yoy and 7% qoq, beating by closer to 1%, while Software revenue rose 4% yoy and fell slightly from the prior quarter, missing by about 1%.
Non-GAAP adjusted gross margin improved 10bp yoy but declined 120bp qoq and missed by 80bp. Operating income (also non-GAAP adjusted) rose almost 26 yoy and 6% qoq, slightly ahead of expectations, with operating margin up 240bp yoy and 20bp qoq to 61.6%, beating by 20bp.
Broadcom doesn’t provide detailed segment breakdowns but does provide enough information to make some extrapolations/estimates.
Within Semiconductors, Networking grew about 30% yoy and 7% qoq, with the company enjoying strong networking demand (Arista (ANET), Ciena (CIEN), and Cisco (CSCO) all posted healthy growth), as well as strong demand for its custom ASICs (up 33% for the full year. Storage revenue rose around 50% yoy and 8% qoq, while Wireless rose 13% yoy and 26% qoq. Broadband grew 20% yoy and strong PON, DOCSIS, and WiFi6 demand, while Industrial (a small business at Broadcom) rose 19% yoy and declined 4% qoq on some weakness in China.
Markets And Customers Matter, And Broadcom Has The Right Exposures Now
One of the points I’ve tried to hammer home in recent articles on Broadcom is that not only does the company have some truly superior products (the Tomahawk, Jericho, and custom ASIC businesses, for instance), but they also have attractively differentiated exposure that should keep the company outperforming its peer group (at least in terms of financials).
While overall cloud capex spending has been growing at a mid-teens clip this year, that drops into the high single-digits if you exclude Meta (META), but Broadcom benefits here from its leverage to Meta’s ongoing investments into AI (in part to power Reels as a rival to TikTok). Likewise, while I’m more negative on overall cloud spending next year, I see elevated spending on high-end connectivity and AI acceleration as hyperscalers and other major data center operators invest to stay on top of traffic growth – growth that I believe will be helped by even more adoption of cloud-based Infrastructure-as-a-Service as companies look to cut their capital intensity and transfer from fixed to variable costs.
Said differently, I expect strong spending on 400G+ connectivity and AI acceleration, and that feeds directly into what Broadcom does best. While I do see Marvell stepping up its game, and I do see some modest risk from Cisco turning more to in-house silicon, the overall trend is still very much in Broadcom’s favor. While management is looking for sequential weakness in fiscal Q1 (on “lumpy” trends), I believe the company will still see healthy growth in CY’23.
In Wireless, the outlook for Apple (AAPL) volumes isn’t great, but Broadcom continues to leverage strong design/performance capabilities to gain share in filtering and connectivity. What’s more, while Apple volumes may be looking weaker, they’re better than the virtual Superfund site that is the Chinese Android phone market right now.
Broadband has been surprisingly strong, and I’m not sure if Broadcom is taking some share back from MaxLinear (MXL), or if this is just a byproduct of different mixes (products and customers). In any case, both companies are seeing strong WiFi demand, and management at Broadcom is looking forward to WiFi 7 deployments.
Broadcom’s guidance for “flattish to up” growth in enterprise storage was more of a surprise, given the weaker guidance from Marvell. Broadcom benefits from less retail exposure and strong MegaRAID adoption, but I still find the outperformance interesting and suggestive of some share shifts. I’m not hugely bullish on the near-term outlook for storage, and I think the end market could be weak until later in CY’23.
Looking at guidance more comprehensively, management has about a year of semiconductor revenue already in the books, and the company has been diligent about scrubbing out over-ordering and enforcing its no-cancellation/no-rescheduling policies. Lead times have shrunk lately (down 15 days in October to 27.8 weeks versus an industry average of 6 days and 25.5 weeks, as per Christopher Rolland at Susquehanna), but not to a decree that concerns me, particularly given the aforementioned drivers in networking.
Uncertainties With VMware Are Not A Critical Issue
Here of late, it looks as though the proposed acquisition of VMware (VMW) is getting more scrutiny from regulators in the U.S. and Europe, with Britain’s CMA talking about investigating a “substantial lessening of competition” and the U.S. FTC investigating the “conglomerate effects”.
I personally find this all a little silly. While there will be cross-selling synergies between VMware and Broadcom’s legacy software businesses, there’s really no overlap, and proving any lessening of competition will be a stretch in my view. I actually do applaud this shift towards genuinely trying to enforce antitrust laws, but I think this is the wrong deal to use as an example and I think concerns like “conglomerate effects” are just silly compared to deals I’ve seen approved over the last year in the industrial sector (not to mention other sectors).
The Outlook
I’ve modestly raised my FY’23 revenue expectations, but also modestly trimmed my ’25-’27 numbers, so the net effect is pretty small (my FY’31 revenue estimate is about 2% higher, and I’m still looking for healthy mid-single-digit revenue growth (around 5%).
I’ve modestly trimmed back my margin assumptions, but that is more out of caution or conservatism than any real issue in the business. I still believe that the company can sustain adjusted operating margins in the low-60%s (with potential upside to the mid-60%s), and I’d note that upcoming contributors like Tomahawk 5 (with integrated photonics) could offer underappreciated margin uplift. Long term, I expect adjusted free cash flow margins in the high-40%s, with FCF growth of around 6%.
Between discounted cash flow and margin/return-driven EV/revenue, I believe Broadcom shares remain around 20% undervalued in the near term, with long-term total annualized return potential in the high single-digits. Were the market to sour even further on the chip sector, a return to prior trough valuation norms could create a downside to the $470s, but that assumes that Broadcom would get a “sector-trough” multiple, which I find unlikely.
The Bottom Line
Broadcom’s differentiated end-market leverage and strong portfolio of performance-driven products continue to serve the company well, and I don’t think today’s share price reflects that value. Broadcom is still a relatively popular recommendation, and I could see the shares lagging other chip names that have been beaten down more as sentiment recovers in 2023, but I still view this as a very solid core tech holding.
Be the first to comment