A man is reflected in a nand flash mass
There have been rumors in the press recently about a possible merger between Kioxia and Western Digital’s NAND flash and SSD division. Western Digital may have looked at a possible acquisition of what had been Toshiba Memory (since renamed Kioxia) when it was spun off from Toshiba to raise cash to fund the troubled parent company’s operations in 2018. More than likely Western Digital was using its flash memory joint venture with Toshiba to get the best terms for its ongoing NAND flash and SSD business.
Recent speculation has been fueled by a settlement in June 2022 with activist investor Elliot Management saying WDC would consider spinning off its hard drive and flash memory businesses. Since then, the company has considered whether creating separate businesses, selling the flash business or other options make sense. Recent rumors indicate that a decision on the future of WDC’s storage businesses will be made in the near future.
Let’s look at what the benefits and costs might be for WDC to sell its flash memory business to Kioxia. First, let’s look at the market share of the major NAND suppliers. According to Trendforce in the third quarter of 2022 Samsung had 31.4% of the market followed by Kioxia with 20.6%, SK hynix and Solidigm (formerly Intel’s NAND business) with 18.5%, WDC with 12.6%, Micron with 12.3% and other companies with 4.6%. At first glance, the combination of Kioxia and WDC’s NAND business would lead to the largest NAND producer at 39.1%, but things never work that way.
If Kioxia and WDC are combined, it is likely that some customers will shift some of their business from the new joint venture to other vendors in order to reduce their dependence on that company and support its competitors. This was certainly the case when there were big mergers of HDD companies in the 2010s. So a combined Kioxia and WDC may not be the market leader, at least not for long.
Also, let’s look at WDC’s flash and HDD revenues and profits from their quarterly reports to see how WDC’s two businesses compare to each other. The chart below shows the company’s NAND flash and HDD quarterly revenue from CQ1 2020 to CQ3 2022. From a revenue perspective, the two businesses somewhat track each other, with the HDD business typically generating more revenue each quarter.
History of WDC NAND and HDD Business
In the chart below we show the gross margin of WDC’s NAND flash and HDD business. WDC’s NAND gross margins generally exceed those of its HDD business. Gross margins should generally be related to the profitability of businesses. Note that the slight decline in HDD business margins in CQ2 2022 was due to a large correction in the company’s HDD business with continued declines in the company’s legacy business following a spike during the pandemic and excess inventory in approximately HDD business. We’ll talk about WDC’s NAND profitability again after we talk about Kioxia’s financials over the last few years.
Comparison of WDC NAND flash and HDD gross margin history
The chart below shows Kioxia’s sales and profit in yen from their quarterly reports. Sales have generally increased from CQ1 2020 to CQ3 2022, but the company’s profit is much more dynamic. Kioxia and WDC get their NAND flash from the same factories in Japan, why is there such a difference if the profitability of the two businesses?
Kioxia’s Sales and Profit History
Analyst Jim Handy of Objective Analysis wrote an article analyzing the profitability of WDC’s NAND business and Kioxia’s business in 2019. Jim points out that the business agreement between WDC and Kioxia (then Toshiba Memory) allows WDC to take up to 50 % of production of the joint venture fab. However, they do not have to get their full 50% stake and only have to pay Kioxia for the product they get from the joint venture plant. NAND flash factories are very expensive to build, and once built, it makes sense to produce as much product as the factory is capable of making to get a return on the capital cost of the factory.
When demand for flash memory is low and prices are falling, this provides an advantage for WDC. The company can control potential losses by taking only as much output from the joint plant as it can sell, while Kioxia must try to sell its share plus what WDC does not take. This causes additional price pressure for Kioxia. As a result of this agreement, then the profitability of the WDC NAND business tends to be more stable than that of Kioxia.
With this information, would it make sense for WDC to sell its NAND flash business to Kioxia? Kioxia would definitely win, but WDC would lose a good source of revenue and profit. So it would be surprising for the WDC to do this. They could separate the two businesses as separate companies, as long as the new NAND business maintains the product agreement with Kioxia, but it doesn’t seem to make sense for WDC to sell its NAND business to Kioxia.
Looking at the fab deal between Kioxia and WDC and the profitable history of the NAND business for the two companies, a merger would be much more beneficial to Kioxia than WDC. This makes it hard to believe that WDC would divest this business to Kioxia, but time will tell.
