Multiyear Enterprise Investment Cycle Continues For Broadcom Stock (NASDAQ:AVGO)

It hasn’t been that long since my last update on Broadcom (NASDAQ:AVGO), but the shares have continued to outperform (up 8% versus a slight decline in the SOX) on the company’s strong leverage/exposure to a red-hot enterprise/networking segment and ongoing excellence in execution on margins. Moreover, with lead-times still near record levels and not as much inventory-building in markets like networking and broadband, Broadcom is better-protected from some of the cyclical sentiment risk that has hit chip stocks.

I continue to like Broadcom today. Taking another look at the key data center market, I believe this market can continue to drive strong revenue growth for Broadcom for several more years, helping offset some of the risk in markets like wireless. With longer-term revenue growth potential above 6%, along with further margin leverage, these shares look undervalued below the $700’s.

Cycle And Sentiment Risks Are Still Real

While there are have definitely still been outperformers in the space, semiconductor stocks as a group have been weaker since the start of the year (SOX is down about 13% year-to-date). While macro issues like the war in Ukraine and higher rates are certainly playing some role, I believe investors are also concerned that the market is at or near a cyclical peak, and that orders will start to slow, leading to weaker revenue and margin performance in 2023 and perhaps into 2024.

Nothing about this is unusual for the space; it’s always been cyclical. Companies are talking about being more attentive to the risks of double-ordering and countering with tactics like non-cancellable orders, but as inventories rise with customers, orders and lead-times will eventually shrink (“when?” is still a very valid question).

You can see a microcosm of this in the wireless chip space, a space where customer inventories are quite a bit more robust than in areas like autos or networking chips. Companies like Qorvo (QRVO), Qualcomm (QCOM), and Skyworks (SWKS) have been lagging the SOX for some time now.

I also see some risk on the pricing side. February industry data from the sector (Semiconductor Industry Association) showed 25% year-over-year growth in chip sales ex-memory, with double-digit growth in pricing. Customers are paying premiums for supply assurance and companies are using the supply crunch to hike prices, and I do see some risk to margins as inventories and demand start to normalize.

Specific to Broadcom, I do see some slowdown risk in certain markets, but overall lead-times are still very strong and backlog more than doubled in the fiscal first quarter (and rose 14% qoq). Given demand, supply, and inventory dynamics in markets like enterprise and data center, not to mention broadband, I see relatively less risk here for Broadcom, though the extent to which the stock could still outperform the group is less clear to me, as sentiment can be a lot harder to predict.

Enterprise, Networking, Data Center … It’s All On Fire

All of the bits and pieces that fall under Broadcom’s “Networking” segment are red-hot right now, as seen in the 33% year-over-year growth in the last quarter and management’s guide for 30%-plus growth in this current quarter.

Businesses are undertaking campus switching upgrades, fueling demand for Broadcom ASICs at customers like Cisco (CSCO) (Broadcom supplies to the Arista 7800 platform, among others). Likewise, 5G backhaul spending remains strong.

In the data center space, strong traffic growth (Lumentum (LITE) recently talked about 30%-plus data traffic growth in 2022) continues to drive capex investment. Broadcom’s Tomahawk line is a key enabler for upgrades along the 200G/400G/800G food chain with Meta/Facebook (FB) upgrading to 200G and Microsoft (MSFT) underway with an upgrade to 400G, and the 800G cycle still in its early days.

On top of that, the hyperscale data center players (Amazon (AMZN), Google (GOOGL), Microsoft, et al) continue to invest in compute offload – driving strong demand for custom ASIC from Broadcom (up 50% in the fiscal first quarter), as well as opportunities for Marvell (MRVL).

Given the growth in data traffic and the opportunities in cloud services, I believe semi-based capex spending could grow 30% or more in 2022, with double-digit growth continuing through 2026. This isn’t a “trees growing to the sky” outlook, as I do expect a slowdown at some point, but between the 200G/400G/800G upgrade cycle and opportunities in compute offload, this remains a major growth opportunity for Broadcom and one where over-ordering/bloated lead-times seems like less of a risk.

I may have missed it, but one thing I haven’t heard from Broadcom is encouraging commentary on their PAM4 efforts. Broadcom has been making a renewed push into optical/photonics, but if anything, Marvell seems to be widening its lead over Broadcom and MaxLinear (MXL) in this market.

Broadband And Storage Healthy, Wireless Less So

As mentioned previously, the wireless space has already seen shrinking lead-times and more robust inventories with customers, a situation that hasn’t been helped by weaker demand from Chinese smartphone OEMs.

Broadcom’s Wireless segment results in FQ1’22 were softer (up 4% versus the 20% overall year-over-year growth in semiconductor sales) and the only segment to come in below expectation. Broadcom’s higher exposure to Apple (AAPL) does offer some protection, but this is the business with the weakest near-term outlook. It’s also a segment with more modest longer-term growth prospects, although the 5G upgrade cycle is still relevant, it’s not going to drive the same revenue growth as other segments are likely to see.

Broadband isn’t as strong as Networking, but 23% growth in FQ1 and an outlook for 20% growth in Q2 are still quite good. Component shortages are a risk for set-top box manufacturers, but Broadcom and MaxLinear continue to see strong demand driven by next-gen PON, DOCSIS, and Wi-Fi 6/6E upgrade cycles.

Storage, too, has been healthy, with management noting greater adoption of nearline HDDs, a business that’s basically a duopoly with Marvell. Strength in nearline HDDs should also be a net positive for the far less well-known Veeco (VECO), a capital equipment manufacturer that is the only provider of ion beam systems used to manufacture the thin-film heads for HDDs.

The Outlook

Given the robust upgrade cycles in enterprise and data center end-markets, I’m not as worried about Broadcom hitting a “speed bump” in 2023 as I am some other chip companies. That doesn’t mean that there’s no risk, mind you, but just less risk given the demand and supply dynamics of those end-markets (fewer relevant competitors, customers more concerned with performance than price, et al).

I’ve upgraded my overall data center spending assumptions, and that helps drive estimate changes with my Broadcom model that push my long-term revenue growth rate from the mid-5%’s to the mid-6%’s. Marvell is a threat in markets like custom ASIC, but I see very strong demand trends for not only custom ASIC, but also Tomahawk switch silicon on that 200G/400G/800G cycle.

Strong overall demand, and particularly for leading-edge products like Tomahawk 4 and custom data center ASICs should continue to generate a tailwind for margins, and I now expect operating margin to exceed 60% in FY’22. I’m modestly concerned that margin leverage could flatten in FY’23 and maybe into FY’24, but I think there’s less risk here than for other chip companies. With an improved margin outlook, my FY’26 adjusted FCF margin moves up about a point (to over 46%), while my long-term estimate moves up about 75bp (to almost 47%).

The Bottom Line

Discounted cash flow gives me a fair value close to $700 today and a long-term total annualized potential return in the high single-digits, and there could be more upside to revenue and margins over the next five years than I’m currently modeling. Likewise, margin-driven EV/revenue and EV/EBITDA suggest a near-term fair value above $710, and I could argue for a fair value closer to $800 without too much difficulty.

Looking at the semiconductor space and the risks of shrinking lead-times/margins, I’d rather own companies with stronger non-consumer leverage, and that includes names like Broadcom, and Marvell. Weaker sentiment is a sector-wide risk, but I believe Broadcom’s leverage to strong a-cyclical growth in markets like data center are an important offset, and I still think this is a chip stock worth owning.