Japan's banks are a disaster. So why do Americans keep buying them?

Oh no, another burst of Japan optimism. Every year or so, news comes out of Japan that makes enthusiasts believe that—at last!—the Japanese establishment is finally restructuring the economy as it should have done 15 years ago. The current boomlet: Sony just announced it will cut 10 percent of its global workforce. Meanwhile, a bunch of U.S. financial institutions have recently struck deals to acquire or manage the bad loans of Japanese banks. Earlier this month, Merrill Lynch agreed to invest $91 million in a corporate turnaround unit of Nishi Nippon Bank—after having invested $1 billion to form a joint venture with UFJ to manage distressed Japanese loans. On Oct. 8, Sumitomo Mitsui Financial Group—a bank in which Goldman has already invested $1.27 billion—said it would transfer $9.1 billion in bad loans to a new joint venture with Goldman. And in August, Cerberus Capital successfully acquired a majority stake in Japanese bank Aozora.

American bankers—and, presumably their Japanese counterparts—hope that the gaijin can achieve what Japan's public and private sector have been unable and unwilling to do for more than a decade: confront the nation's gruesome bad-debt problem.

But along with their PDAs, laptops, and spreadsheets, U.S. bankers winging their way across the Pacific should also pack a copy of Saving the Sun,the new book by Financial Times correspondent Gillian Tett. It tells the ongoing story of the brash U.S. investment fund, Ripplewood, which in 2000 acquired Japan's Long-Term Credit Bank. LTCB had amassed $50 billion in bad loans and hadn't the slightest clue how to deal with them. Saving the Sun should leave us skeptical that U.S. capital will ever really fix the perverse Japanese banking system.

In the '80s and '90s, Japanese banks poured loans into unproductive businesses and, even worse, real estate deals. When the real estate bubble popped, the result was a mountain of bad debt—as in the United States' own S & L crisis. To any outsider, the answer to the Japanese banking morass has long been clear: Foreclose on bad loans, recognize losses, sell off assets, force companies to restructure, and put new capital to work. That's precisely what the United States did in the early '90s after the S & L debacle.

But Long-Term Credit Bank, like other Japanese banks, didn't confront its problems directly. And Japanese regulators never forced it to. In Japan, as Tett explains, banks existed not to make money for their shareholders but to support long-term clients—even when they didn't make any efforts to pay back their loans.

LTCB lent recklessly in the 1980s—particularly to real estate developers. When things went sour, it wrote off a few bad loans, shifted others to subsidiaries, then foolishly continued to extend credit to long-standing clients that were essentially bankrupt. Meanwhile, it intentionally mischaracterized the scope of its problems. The government approved of all of this. It was in keeping with a Japanese business culture that had worked well for decades. "Japanese culture has traditionally assumed that a distinction exists between 'public' or 'official' reality (known as tatemae) and 'private' truth (known as honne)," Tett writes. Unlike other most writers of business narratives, Tett, who trained as an anthropologist, shrewdly believes cultural differences—not personality conflicts or the quest for cash—account for the conflicts in the story.

In the spring of 2000, the desperate Japanese government allowed Ripplewood, an upstart U.S. private equity firm that had no experience in banking or in Japan, to take over LTCB and rename it Shinsei.

Although Ripplewood installed a Japanese CEO, the story since the takeover has been the clash of civilizations. Ripplewood's lead U.S. executive "believed that the world's banks were driven by a set of rules as universal as the Newtonian laws of physics." But the Japanese bureaucrats overseeing Shinsei clung to the belief that banks existed not to return cash to their shareholders but to serve a societal need.

The LTCB deal allowed Ripplewood to shift any really bad loans to the Japanese government's books. Ripplewood did this, causing a public relations and political disaster. Dumping the loans meant abandoning long-standing relationships and exposing well-known companies as deadbeats—a no-no in Japan. "Wall Street was founded on the presumption that if there was a showdown between a legal transaction and a relationship, it was the transaction and not the relationship that should be honored," Tett writes.

To its credit, the Japanese government honored its agreement, accepting $10 billion in bad debt. In the past three years, Shinsei's management has installed better credit controls, securitized loans, built up its retail presence, and drastically slashed the number of bad loans. But we're left with the impression that much of its success stems from its ability to shift the burden of bad debt onto Japan's taxpayers. And Ripplewood has not yet been able to cash out: It has been trying and failing to do an IPO since July 2002.