jiefeng jiang
Investment thesis
Marvell Technology (NASDAQ:MRVL) has been sold off like a cyclical semiconductor stock, but in my view, the company has idiosyncratic factors that will drive positive alpha in the long-term when compared to other more cyclical semiconductor peers. As a result, I take the view that Marvell will benefit from the multi-year data center growth story. Although supply remains to be a key constraint in the data center segment’s growth, this appears to be able to ease by the end of the year. At the same time, demand for the data center has been encouraging as it continues to outpace and exceed the supply in the market. At the same time, while automotive is a small percentage of the company’s total revenue mix today, I believe that this will grow to become more significant over time as the content in automotives continues to increase and the segment benefits from increasing electric vehicle, hybrids adoption as well as improving technologies from autonomous vehicles. My 1-year target price for Marvell is $75, implying an upside potential of 90% from current levels.
More of a supply story
In the near-term at least, we will likely see that Marvell’s investment thesis will rely on the supply side of the equation as demand continues to outstrip supply. The main tightness in supply comes from the products that are higher in complexity that have relatively longer manufacturing cycle times like the data center segment. The fact remains that because these more complex products in the data center segment utilize the top and leading technology as well as advanced packaging, there is relatively limited supply for these components in the market given how difficult it is to produce these components. While the Marvell operations team are actively securing capacity for the company’s long-term growth, I think that Marvell has limited ability to control the supply side of the data center segment. As a result, the supply situation for the data center segment, in my view, will remain tight in the near-term. On the other hand, Marvell is starting to see supply improvement for less complex products and certain nodes and supply chain that require simpler package technology.
Management expects that in the third quarter, supply will continue to be a constraint in meeting demand. For the third quarter, as a result of the challenging supply situation, it expects the overall revenues from the data center segment will decline sequentially in the single digit range. On the brighter side, there is an expectation for the supply situation to improve in the fourth quarter and we will likely see the data center segment revenues increase sequentially in the fourth quarter, up from the third quarter, as the effects of the ramp up in new products in cloud materializes.
As I have alluded to earlier, Marvell’s management will continue to invest in and secure the key supply for their longer-term growth. As a result, int he second quarter, we saw that the company increased the long-term purchase commitments by almost 15% to $3.4 billion, which in my view is consistent with the commentary that they are continuing to secure supply. I think that the priority of management is to find the necessary supply needed for their new design wins in the near-term and that the current strategy is aligned to the top priority for the company.
Demand remains solid relative to peers
While the supply situation is the main constraint, demand for the data center remains solid. I was encouraged to not hear in the second quarter earnings call any signs of weakness in demand for the data center segment.
I think that rather than show any weakness in demand for Marvell’s data center segment, the management painted a rather strong demand picture for the company’s data center segment. Management continues to expect that this single segment will be the largest contributor to Marvell’s growth in the long-term, and that the demand that they are seeing for the segment continues to be healthy in the second quarter. With regards to the highest growth segments within the data center segment, management expects customized solutions, cloud switching as well as data center storage to be interesting segments within its data center portfolio.
While other peers like NVIDIA (NVDA) have commented about weakness in Chinese hyperscale data centers, this problem seems not affect Marvell. In the second quarter earnings call, Marvell management reiterated that its cloud business comprises of mainly US hyperscale data centers, while having very limited exposure to Chinese hyperscale data centers in terms of revenues.
Long-term automotive upside
While the automotive segment for Marvell only makes up 6% of total revenues as of the second quarter of FY2022, this is one of the growth focuses for Marvell as it leverages on increasing design wins for both the traditional internal combustion engines as well as in the higher content hybrids and electric vehicles.
This segment is one of the fast-growing segments of Marvell as it saw 46% growth year on year in the second quarter results. The main driver for Marvell’s strong auto segment is due to increasing adoption of its Ethernet technology. As management shared in the last earnings call, their Ethernet design wins have an admirable 8 of the 10 largest OEMs globally and in total, it has 36 OEM design wins.
As management highlighted in the second quarter earnings call, these new design wins are rather substantial given that Marvell won over the highest volume internal combustion engine models as well as the electric vehicles and hybrids with higher content levels for the large OEMs. I think that the automotive segment looks to me to be very competitive given the large number and quality of design wins despite the smaller revenue mix it represents today. As a result of continued innovation and new products like the Ethernet technology that is seeing increasing adoption today, both large and smaller OEMs see the value in being a customer of Marvell as the content in automotives continues to grow.
As a result of continued advancement in the autonomous vehicle front, as well as increasing content and need for bandwidth in hybrids and electric vehicles, the automotive segment, in my view, will have a long-term secular tailwind supporting the demand for Marvell’s automotive products.
Valuation
Marvell currently trades at just 14.6x 1 year forward earnings, while earnings CAGR is expected to be in the range of high double-digit range. My 1-year target price for Marvel is based on an equal weight of DCF method and P/E multiple method. I assume a P/E multiple of 26x for Marvell, which is a slight discount to the company’s 3-year average P/E multiple. Also, I apply a discount rate of 10% and terminal P/E multiple of 15x as key assumptions for the DCF model.
My 1-year target price for Marvell is $75, implying an upside potential of 90% from current levels.
Risks
Data center demand
While it currently is performing strongly, there is a risk that data center demand may take a hit in the weakening global macroeconomic environment. This will be a big hit to Marvell as its growth story hinges on the secular growth of the data center segment, that has been seen to be more resilient than other segments in more challenging operating conditions.
Macroeconomic environment
If the macroeconomic environment weakens more than expected, this may have a greater-than-expected impact on the various segments of Marvell. As such, we will likely have to revise the near-term forecasts in the financial model downwards.
Loss in market share
There is a risk that Marvell may end up losing market share in either the storage or networking space, or even both. This might come as competitive pressures intensify for Marvell as smaller companies might innovate and produce a networking or storage product that can compete meaningfully with Marvell. In addition, Marvell operates in a competitive environment that has large established players, like Broadcom (AVGO).
Conclusion
Marvell continues to be a long-term data center growth story as the company is able to generate significant alpha for investors given the outpacing of demand over supply for the company’s data center segment. Demand remains strong even as the macroeconomic environment has weakened, and other semiconductor companies are cutting back on spend and growth estimates. On the other hand, Marvell’s near-term challenge is unlike that of peers. It has more than enough demand and the main constraint remains supply, which it is proactively addressing by securing the necessary supply beforehand. In addition, the supply situation is expected to improve closer to year end as Marvell’s more complex products like the data center segment sees a less tight supply situation. Lastly, Marvell’s automotive segment looks set to benefit from the strong tailwinds from the increasing adoption and technology improvement of autonomous vehicles and electric vehicles, and as such, will also be a focus for long-term growth for the company. My 1-year target price for Marvell is $75, implying an upside potential of 90% from current levels.
Be the first to comment