Merging the Sumitomo and Mitsubishi wafer businesses into a 300mm specialist (2002)

Bundling a business, and making it earn

Read this as a merger of the Sumitomo and Mitsubishi camps and you miss the point. What the fifty-fifty joint company formed in 1999 took on was the job of mass-producing the next-generation 300mm wafer as a single firm. Its plants at Imari and Yonezawa were still in ramp-up — president Mori Reijiro spoke of taking output from a little over 40,000 wafers a month to 100,000 — and neither Sumitomo nor Mitsubishi could carry the investment beyond that alone. Gathering Japan’s three separate silicon lines into one was not a matter of tidying up the industry’s appearance; it was a design for loading an ever-growing investment burden onto a single balance sheet.

Bundling alone, however, produced no profit. In the year to January 2004, two years after the merger, Mitsubishi Sumitomo Silicon posted a net loss of $314M (¥36bn) on sales of $1.4B (¥166bn). 300mm did not become a profit pillar until demand took hold around 2006, after cumulative investment of roughly ¥130bn — and the confluence of the three lines was itself only completed that year, with the acquisition of Komatsu Electronic Metals. Between gathering a business into one company and having that business earn there lay the distance of one full semiconductor demand cycle. Whether the decision to bundle was right showed up not in the results that followed it, but in the years over which the investment could be sustained.