Guest contribution: Barbarians at the gate

Comparing the U.S.-led takeover of Toshiba to the Nippon Steel bid for U.S. Steel: who is damaging national economic security?

As U.S. politics and national economic security interests become the increasingly dominant force in the cross-border market for corporate control and investment, my friend and fellow investor/activist/strategist Andrew McDermott (see below for a detailed bio) offers great analysis and insights into the U.S.-Japan dynamics. He contrasts Japan’s reaction to the foreign takeover of Toshiba to the U.S. response to the Nippon Steel bid for U.S. Steel. Which system is “Capitalism that works” ?

For the future of the U.S.-Japan alliance, there may still be some hope that the new U.S. administration will act rationally and allow Nippon Steel to help engineer an American industrial renaissance. By all accounts, Nippon Steel has got the capital, the technology, and the management focus to deliver….

Personally, I am more interested in the Japan-specific dynamics substantiated by Andrew’s analysis: the negative impact of U.S.-led-finance-first capitalism (of which the Toshiba case is Exhibit A….) is, in my view, starting some serious re-thinking of economic- and financial policy priorities amongst the Japanese ruling elite. Whether we’ll get all the way to a “Japan that says ‘No’ “ (to America) remains to be seen; but the growing confidence amongst the elite that, yes, Japan is “Capitalism that works”, is, in my view, a key mega-trend to watch out for. And yes, the U.S. reaction to Nippon Steel’s bid made me more confident in this view.

Here are Andrews’s thoughts and analysis. As always, comments welcome — many cheers ;-j

Turnabout (un)fair play: comparing Nippon Steel/US and Toshiba/Japan

"Reciprocity" defines US trade policy on both sides of the political aisle. We expect it from our adversaries and demand it from our allies. As Joe Biden seems intent on making his legacy a trade dispute with America's most important geopolitical and economic ally, it seems worth comparing Biden's actions in regard to US Steel with the result of then-Japanese Prime Minister Shinzo Abe's actions in regard to Toshiba in 2017. In preview, the US financial engineers Japan trusted with Toshiba have run it into the ground, leaving it in much the same position as US Steel: financially levered, loss-making absent subsidies, lagging competitors in R&D and innovation, and led by over-compensated financial engineers with little knowledge of the business at hand. The difference between the two transactions is that the Americans took capital and skills out of a world-class business while the Japanese offered capital, expertise and skilled leadership to a broken US icon.

Chips and Ships

If, as many contend, chips are to modern warfare what ships were to the 1940s, then the national security implications of a foreign takeover of Toshiba's memory chip operation would dwarf the risks posed by taking over US Steel. In 2017, Toshiba was a market-leader in specialized NAND memory, which Toshiba invented in 1987. This technology would form the foundation for high bandwidth memory, an essential component of all AI systems. By contrast, US Steel has underinvested in technology and innovation to such an extent that it is not even a top 20 global steel maker. US Steel supplies exactly 0% of its output to the US military and has so underinvested in presses that the most recent US nuclear power plant parts had to be pressed in Japan.

But let's give President Biden the benefit of the doubt and assume strategic equivalence between the two assets. What would a neutral observer infer from how they have evolved?

First, a little background on how US investors came to control Toshiba Memory in the first place. In 2017, Japan, as part of Mr. Abe’s attempt to improve relations with the US on all fronts, invited US-backed financial players to restructure Toshiba. It is important to note that Toshiba was in financial trouble in the first place because of poor decisions by US management at Westinghouse’s nuclear business, which Toshiba had bought in 2006 (Daniel Roderick, the executive directly responsible for the debacle, walked away with a cool $19 million. He most recently appeared as state's evidence in the ongoing Federal fraud case against his own employees).

Toshiba's Japanese managers had made a series of accounting errors, but their products and services were rock-solid. In the normal course of events, Toshiba would have been “rescued” by a consortium of Japanese customers, suppliers and banks who would have nursed it back to health over a decade. In 2017, this slow-and steady approach seemed too patient to many investors (including yours truly). In the fullness of time, the recovery of almost every Japanese conglomerate other than Toshiba speaks to the efficiency and efficacy of this "public market" approach.

But at Toshiba, Japan followed a different playbook. Goldman Sachs led an "innovative" capital raise directly targeted at foreign (or foreign-backed) investors (including Harvard). This policy extended to Toshiba memory, whose eventual takeover by Bain cleared a rigorous national security review that excluded China-linked buyers. Notably, unlike Nippon Steel, Bain and other bidders were not forced to relinquish the PRC investments which even now dominate their portfolios. For example, in 2023, Nippon Steel exited its 50 year JV with China's Baosteel, in part to show its commitment to the US Steel deal. The same year, Bain Capital executed the biggest "mega-deal" in Greater China: a $3.2 billion purchase of Qinhuai Data. This deal was part of Bain's multi-year investment strategy to provide data centers in support of the PRC's push into AI and cloud computing. Bain's China office remains larger than its Tokyo office and seems as committed as ever to doing deals in China even as Japanese companies follow US guidance to move business elsewhere.

What did Japan get for its lenient open door policy? Goldman's investors supported a leveraged recapitalization of Toshiba from the get-go, immediately plowing most of the funds raised into buybacks and dividends to themselves. They recruited a former private equity executive to lead the firm. New management avoided playing any part in the recapitalization of the (now extremely valuable) US nuclear assets (which Canadian financiers bought for pennies on the dollar). Toshiba eventually vanished with a whimper into the arms of exactly the sort of consortium that rescued other Japanese industrial leaders, but seven years too late.

And what of the jewel in Toshiba’s crown, the memory business that alone among Japan’s semiconductor companies remained relevant thanks to its industry-leading innovation? Toshiba invented flash memory and was a leader in the memory field when Bain took over. There was not actually anything to “fix” at the memory business. It turns out there was much to break, and that’s what Bain did.

Today, high bandwidth memory (“HBM”) is essential to AI. The field is dominated by Toshiba’s erstwhile flash competitors: Samsung, SK Hynix and Micron have one hundred percent of the market. Not coincidentally, all are led by engineers who maintained the balance sheet strength necessary to invest in capacity and R&D through the downturn so that they could be poised to reap the benefits of AI’s rise. All three have reported profits recently even as Toshiba Memory (now called “Kioxia”) reported years of losses.

Financial or real engineering?

Bain, of course, considers this deal a great success, at least according to their former CEO’s last interview at Davos (before he retired to concentrate on managing his professional sports teams). Bain took much of its money out early during the “recapitalization” of Kioxia and has paid itself a ¥1 billion annual fee for “managing” the business, even as total SG&A in dollar terms plummeted (Bain's fee is multiples of what most Japanese CEOs receive in annual salary). For all this money, Bain presumably would have brought the best and the brightest to the management team, particularly after the untimely death of Kioxia’s leader Dr. Yasuo Naruke. Dr. Naruke not only led the company, he held multiple patents in the semiconductor field. Who did Bain invite to replace him? None other than Stacy Smith, a finance executive who spent most of his career at Intel leading the financialization that led to Intel’s current dire straits.

It is notable that Japan’s home-grown semiconductor leaders have fared far better than Kioxia over the past decade or so. Whether at Rapidus, the TSMC plant in Kyushu or at Renesas, the “home team”‘has shown that Japanese engineering talent coupled with sustained capital investment and foreign engineering (as opposed to financial) talent can succeed in the critical task of rebuilding our collective techno-industrial base. As Germany's failures have shown, this is not a simple task.

Japan can only hope that the incoming Trump administration takes time to read Secretary of State nominee Marco Rubio's latest book, Decades of Decadence. If Mr. Trump is serious about reinvigorating America's industrial base, strengthening its alliances with like-minded partners, and supporting middle-class jobs rather than those of financial elites whose vested interests are often aligned with our adversaries, then he will reverse the Biden decision at Nippon Steel and limit the damage that financial engineers can inflict on our common prosperity.

The opinions expressed are my own. They are not intended as investment advice. I could own or may in the future own positions in securities mentioned in this article.

Andrew McDermott - Nashville January 7, 2024

For Andrew’s bio and more of his thoughts, please see his previous guest contribution to Japan Optimist — Why not Hitachi?

and also :

Toshiba - the hidden cost of activism