Buying Toshiba's memory unit is a complex issue, especially since the company makes for an attractive target as positioning for the transition to solid-state solutions.
To save Toshiba from the brink of a financial black hole, the company has decided to sell off one of its prized businesses—the memory unit—which, in turn, created a storm of acquisition activity.
Buying the memory unit is a complex issue. There are technical considerations, since Toshiba is a leading developer of NAND technology, making for an attractive target as positioning for the coming transition from spinning hard drives to all solid-state solutions.
Figure 1: Toshiba's quadruple-level cell (QLC) 4-bit flash memory in a 3D flash device. (Source: Toshiba)
There are financial considerations, of course. Toshiba places third globally in flash sales, and with current margins buoyed by a under-capacity problem across the industry the flash business is a profitable entity. This could drive up the asking price for the business, reputedly around $27 billion right now against an evaluation for the business at around $18 billion, but Toshiba is in a distress sale situation.
Finally, there are political considerations. Toshiba is Japanese and that nation is reluctant to lose a key element of its technology portfolio. On the other hand, China is extremely keen and publicly committed to chip self-sufficiency within a decade or so, which one could easily translate into “having a dominant position” in key components such as CPUs, DRAM and flash.
Complicating things is that the world’s largest hard disk drive (HDD) vendor, Western Digital (WD), is already in a joint venture with Toshiba. This is a key strategic element for WD, who is investing very heavily in flash/SSD to replace dwindling HDD demand. The partnership pools the Toshiba technical expertise with WD’s deep pockets and storage market savvy and it’s key to keeping WD at the leading edge of SSD art, while also protecting WD from flash die shortfalls and market price fluctuations.
So who are the bidders? Japan made it clear early on that they view Toshiba as a national resource. That precludes some of the more overt takeover activity we’ve seen from China in the last year as they used a fund reputed to be more than $60 billion to buy up smaller, niche flash players in other countries.
Nonetheless, a major company with strong ties to China is leading one bidding group. This is Hon Hai (Foxconn), a Taiwanese contract manufacturer with a huge footprint in the PRC. They are partnered with Apple and Amazon, and the group members each stand to gain control over a key component of the products they make or use—iPhones, Kindles and the Amazon cloud servers, among many other products). Right now they are reputedly the $27 billion bidder.
Japan has a couple of government-backed investors who are involved directly with Toshiba. These are the Innovation Network Corp of Japan (INCJ) and the Development Bank of Japan (DBJ). They are very tight-lipped on their involvement, but their status viz-a-viz the Japanese government suggest they are there to assist bid groups that will keep both IP and the business unit within Japan.
Broadcom, the U.S. chip giant, has joined with Silver Lake, a U.S. technology investment fund and three Japanese bank groups, Mizuho, Sumitomo and Mitsubishi UFC, in a significant bid. Details are sketchy, but the bid seems to be around $20 billion. This would expand the scope of Broadcom’s product line-card and an acquisition fits the mould of investing into growth markets that Broadcom has followed for the last several years.
Equity firm Kohlberg Kravis Roberts is nosing around the opportunity, too. It isn’t clear what their aim is, other than a purchase followed by a relatively quick sale of the component elements of Toshiba’s flash business. KKR’s bid is around $16.36 billion, but they are rumoured to have taken advantage of the Japanese government backing mentioned above, with INCJ and DBJ bringing their bid price up significantly.
It appears that INCJ is not fully committed to KKR, however, and is considering adding to a bid by Bain Capital and SK Hynix who together are offering $9 billion for 51% of the business.
That leaves Western Digital, who attempted to protect their joint venture investment (and strategic die source) by demanding an arbitration process before any sale proceeded. Many lawyers later, this is in the dust of history and now WD is a lead bidder for the memory unit, though the joint venture complicates the whole bidding process, since WD is reasonably demanding protection.
You might throw your hands up and say this is a wide open race, but the leaks around the bidding process point to political influence as being a major factor. The Japanese government wants to keep one of the few chip makers it has in Japan. This all but nixes the Hon Hai bid, given the expected fierce opposition of the Japanese public to any China-related deal.
The KKR deal looks like opportunistic investment of the “buy low, sell high” type, which also misses the goal of keeping a viable Japanese chip maker. SK Hynix is a South Korean company, and even though they claim they will operate the flash business at arm’s length, there is a strong sense that this just concentrates even more of the flash industry in Korea, where Samsung is the world leader.
That leaves Broadcom and Western Digital. There are rumours circulating that, in fact, these are the two finalists in the bidding. One can see why this would be so. Broadcom is a pure chip play, but they will add process expertise, coupled with a broad technical reach, to the Toshiba business. This will strengthen that business and will also have the credibility to raise necessary capital as flash evolves.
WD already has a large, successful Japanese-based acquisition under their belt. Their purchase of HGST, a Hitachi unit, has been well received. Moreover, they are in the same business as Toshiba, in selling storage, so the synergies of an end-to-end business should strengthen the (Toshiba) memory business and thus Japan’s technical position.
With the usual assurances that the Toshiba unit won’t be dismembered or moved offshore, either Broadcom or WD appear to be a good fit. With the Japanese banks apparently there to supplement bids as necessary, we should see an announcement before the end of June.
Looking at this from a different direction, what would a win do for Broadcom or WD? The flash-based storage units of WD, HGST and SanDisk have integrated remarkably well into the WD operational flow. Since WD would be essential extending an existing supply relationship with Toshiba, one would expect the acquisition of the memory unit to also go smoothly and add to WD’s bottom line quickly. Clearly, the added technical IP and expertise should accelerate and broaden WD’s flash-based product lines. Overall, this would be a good match.
Broadcom is somewhat lacking in the storage synergy, but makes up with chip-making and selling expertise. They don’t have two-plus years of working together, but they have great experience in successfully integrating acquisitions. Their payback is a major foothold in a fast-expanding market.
First published by EBN.
New products & solutions, whitepaper downloads, reference designs, videos
Register, join the conference, and visit the booths for a chance to win great prizes.