The technological transition from hard disk drives (HDD) to solid state drives (SSD) is well known. This transition is especially critical for Marvell Technology Group (MRVL). The question presented in this article is whether profitability will be the casualty of this transition. However, lack of transparency in financial reporting is masking these profitability related risks. This article will outline two key reasons for concern.
Marvell is a fables semiconductor company. HDD storage controller is the main product. As a result of industry consolidation in recent years, the HDD controller market has become a duopoly with Marvell and LSI Corp (LSI) being the two companies sharing the market.
In addition to HDD controllers, Marvell also designs SSD controllers. The storage segment revenue includes both HDD and SSD controllers. In 2012 the storage controller segment represented 47% of Marvell's total revenue. Therefore the company is highly dependent on the storage segment.
According to the Q1 2014 conference call transcript, HDDs are still the dominant source of revenue by a wide margin. Recently the HDD revenue has been declining. The decline in HDDs has been mitigated by a rise in SSD revenue. In other words, technologically more advanced SSDs are taking share as the transition gains momentum.
In addition to the storage segment, Marvell has two smaller segments each accounting for roughly a quarter of total revenue: The Networking segment consists of switching solutions, which enable high-speed data transmission in communications systems. The Mobile & Wireless segment consists of complete system-on-a-chip (SoC) smartphone platform solutions, including communications and applications processors, wireless, power management and other platform components.
Increased Competition in Storage Controllers
The technological transition from HDDs to SSDs has left the door open for competition. The following factors outline the change in competitive environment:
Firstly, SSD manufacturers are faced with an option to design the controller chips in-house instead of purchasing the chips from Marvell or LSI. For example, Seagate (STX), the major HDD manufacturer, has shown interest in acquiring SSD manufacturers like OCZ Technology (OCZ) that have in-house controller design capabilities. Similarly, major SSD manufacturer Micron (MU) has been internally developing controllers. Here's an excerpt of Micron's Q4/2012 conference call transcript:
"We've also made significant strides in the quarter and in the fiscal year with our internal controller development. Our controller strategy is really 2-pronged, using both internal and externally-designed controllers. We're focused at the high-end of our system solution for our internally design controllers. However, our approach is modularized and can be scaled down to the lower-end client systems over time."
Secondly, new competitors have entered the controller market with the transition: Samsung (OTC:SSNLF) is manufacturing SSDs in-house including the controller chip. SK Hynix (OTC:HXSCL) is the world's second largest DRAM and flash memory chip manufacturer after Samsung. With its recent Link A Media acquisition, SK Hynix expanded its capabilities from flash memory to SSDs and controllers. Other competitors include companies like Toshiba, OCZ Technology and Silicon Motion (SIMO). Even Apple acquired an SSD controller company called Anobit. All these companies are offering a variety of different types of controllers aimed for different end markets. Even though the SSD market is becoming crowded, Marvell and LSI (SandForce) are faring quite well. However, as the SSD market grows, it will attract more intense competition.
Thirdly, SSDs are competing against HDDs. SSDs will be taking increasing share especially in the high-end storage market. This in turn will negatively affect the HDD average selling prices. Therefore, while Marvell and LSI will maintain their strong position in the HDD market, the value of that position will diminish.
Lastly, the market share balance between Marvell and LSI might unexpectedly shift one way or the other as a result of the technological transition.
Overall, the storage controller market is facing a lot of uncertainty. Increased competition makes historical duopolistic profit margins enjoyed by Marvell and LSI unsustainable in the future.
It's customary that companies disclose both revenue and operating profit for each operating segment. However, Marvell discloses only revenue for each segment. It's not in the interest of a company to disclose profitability for each of its operating segments if profitability is highly concentrated on a single segment like storage controllers. Even more so if the segment in question would be facing significant uncertainty going forward. Especially if this uncertainty would relate to profitability. It would be a significant concentration of risk.
In 2012 storage controller segment represented 47% of Marvell's revenue. The other operating segments don't have nearly the same scale or competitive position. Therefore it's likely that the storage controller segment's share of total operating income is much higher than its share of total revenue. This view is supported by management comments, which indicate that the smartphone SoC segment has a negative impact on gross margin.
Therefore, the increasing competition in the storage segment is likely to have a significant and negative long-term impact on Marvell's profitability. In other words, profitability might be the casualty of the transition from HDDs to SSDs. At the very least, it would be optimistic to extrapolate the historic profit margins far into the future. While this technological transition will take time and is not an immediate risk, the market can discount the risk in advance.
Companies that face uncertainty in their primary business often seek to diversify into other businesses. The riskiest way of diversifying is to make an entry to a new market with new products and new customers. A lot can go wrong.
In recent years, Marvell has been aggressive in growing its Mobile & Wireless segment that now accounts for around 25% of the total revenue. Marvell designs smartphone SoC platform solutions. It's utilizing the storage segment as a cash cow to finance investments into this Mobile segment. However, the problem is that Marvell lacks clear and sustainable competitive advantages in the Mobile segment.
In the Mobile business, Marvell has positioned to the low end of the market. It's unable to compete for a position in the high-end flagship smartphones. In other words, it's scraping the bottom of the barrel, taking what is left behind by larger and technologically more advanced competitors such as Qualcomm (QCOM), NVIDIA (NVDA), Samsung (OTC:SSNLF) and now Intel (INTC). A while ago Texas Instruments (TXN) withdrew from the market due to competition. Marvell has proven to be agile, finding niches in the market like it did in the TD-SCDMA (China 3G) market.
However, as the market matures, it's unlikely that Marvell can consistently replicate this agility. It's already losing its first-mover advantage in the TD-SCDMA market. The Mobile and Wireless segment revenue has declined over 26% in the last 2 years. Overall, margins are likely to be razor thin in the low-end smartphone SoCs. In addition, at less than $1 billion in annual revenue, the Mobile & Wireless segment is lacking adequate scale.
The risk is that Marvell is wasting a portion of the cash flows provided by the highly profitable HDD controller products in a futile attempt to capture share in a low-end commoditized smartphone SoC market where margins are thin and Marvell lacks competitive advantages. In other words, unsuccessful attempts to diversify the business can lead to destruction of shareholder value. Considering the circumstances, likelihood of failure is high.
On the Q2 2013 conference call, a Citigroup analyst was raising a question whether the company should withdraw from the smartphone SoC business as competition intensifies. Marvell's CEO responded that the company is committed to gaining share in the mobile market: "We are very, very serious in this business. We cannot blink. We will not step back for any reason." While the determination is admirable, absolutism in the face of adverse reality can be dangerous.
While the diversification related risks are more pronounced in Marvell's case, LSI could be facing similar diversification related challenges in the future. After all, nearly 80% of LSI's revenue came from its storage segment in 2012.
There are also positive factors to consider:
In the smartphone SoC business, the CEO correctly points out that it's the biggest market of semiconductors in the world. While the opportunity is huge, I'm not convinced Marvell is up for the task.
In the storage segment, transitional periods can present opportunities. During the transition from HDDs to SSDs, hybrid drives will emerge to combine the best of both worlds: the low cost of HDD storage capacity coupled with the high performance of SSDs. This is where Marvell and LSI have advantage over the competition since they are strong in both technologies. However, hybrid drives will cannibalize HDDs and SSDs, so they don't actually expand the market.
On the other hand, cloud storage in general provides opportunities for growth. Similarly, SSD controllers open new markets, especially in the mobile space. Despite the upside, the risk of losing the duopolistic HDD margins probably outweighs the opportunity to compete in a larger but more crowded and commoditized SSD controller market.
There are several short-term factors affecting the profitability. Instead, this article has taken a long-term view outlining two key reasons for concern: Firstly, increasing competition in the storage controller business is likely to decrease profitability.
Secondly, diversification into highly competitive and commoditized smartphone SoC markets without clear competitive advantages is likely to result in destruction of shareholder value. As for valuation, Marvell's shares appear relatively cheap -- but only if historic profitability is extrapolated.
The ongoing patent litigation between Carnegie Mellon University and Marvell is an additional risk factor to consider.