Skip to main content
Japan's Lost Decades

The Keiretsu System and Corporate Japan

Pomegra Learn

How Did Japan's Corporate Structure Both Create and Prolong Its Crisis?

Japan's keiretsu system — the dense web of cross-shareholdings, bank relationships, and inter-corporate obligations that defined Japanese business organization for most of the postwar period — was widely admired in the 1980s as a model of patient capital and long-term coordination superior to Anglo-American short-termism. The same features that were celebrated as strengths — stability of ownership, close bank-corporate relationships, protection from hostile takeovers — turned out to be amplifiers of the bubble and impediments to the recovery. Understanding the keiretsu is essential to understanding both why Japan's bubble was so extreme and why the resolution of the subsequent crisis took so long.

Keiretsu: A group of Japanese companies bound together by cross-shareholdings, centered on a main bank (horizontal keiretsu) or a large manufacturer (vertical keiretsu), characterized by stable long-term relationships, coordinated strategic decisions, and mutual obligation — a corporate structure that amplified asset price movements in both directions.

Key Takeaways​

  • Horizontal keiretsu (Mitsubishi, Mitsui, Sumitomo groups) organized around main banks; vertical keiretsu (Toyota, Sony groups) organized around manufacturers.
  • Cross-shareholding typically ranged from 2–5 percent of each member company's shares — small individually, but collectively locking up 30–50 percent of major companies' outstanding shares in stable, non-traded positions.
  • Rising equity prices during the bubble increased the value of cross-held positions, strengthening apparent bank capital and enabling more lending — the bubble amplification mechanism.
  • Falling equity prices during the crisis reduced cross-held positions' value, impairing bank capital and restricting lending — the depression amplification mechanism.
  • The main bank system created implicit guarantees: member companies expected their main bank to provide rescue financing during stress, and banks expected to support members. These implicit commitments became explicit obligations during the crisis.
  • The keiretsu structure made hostile takeovers essentially impossible — removing the market discipline mechanism that would have forced restructuring of underperforming companies.
  • Unwinding cross-shareholdings became a major corporate reform project in the 1990s and 2000s, with completion rates varying by company and sector.

The Structure of Keiretsu​

The term keiretsu evolved from the zaibatsu — the family-controlled industrial conglomerates that dominated Japanese industry before World War II. US occupation authorities dissolved the zaibatsu in 1945–47 as part of Japan's democratization, breaking up the family holding companies and separating the affiliated industrial firms. Over the following decades, however, the separated firms gradually rebuilt coordination networks through reciprocal shareholding and regular management meetings — the keiretsu emerged as the postwar evolution of the zaibatsu.

Horizontal keiretsu (also called financial keiretsu or kigyo shudan) are organized around a main bank and typically a general trading company (sogo shosha). The major horizontal keiretsu include the Mitsubishi, Mitsui, Sumitomo, Fuji (Fuyo), Sanwa, and Daiichi Kangyo groups. Member companies in these groups span multiple industries — manufacturing, trading, finance, insurance, real estate — coordinated by the main bank and the Presidents' Council (shacho-kai), a regular meeting of member company CEOs.

Vertical keiretsu (seisan keiretsu) are organized around a large manufacturer and its supplier and distributor networks. Toyota's supplier network — with Toyota owning partial stakes in key tier-1 suppliers and exercising significant influence over their management — is the canonical example. These vertical keiretsu created efficient supply chains but also created dependent supplier relationships that made it difficult for suppliers to develop alternative customers.

Cross-Shareholding: The Structural Core​

The cross-shareholding arrangements within keiretsu were the key mechanism linking asset prices to banking system health. The typical pattern involved each member company holding 2–5 percent of other members' shares. These positions were:

Stable and long-term. Cross-held shares were not purchased for investment return or traded actively. They were held as expressions of the relationship between companies — more analogous to a business partnership commitment than a financial investment.

Protective. With 30–50 percent of outstanding shares held in stable cross-held positions, hostile takeover attempts were essentially impossible. An acquirer would need to purchase the remaining 50–70 percent of freely floating shares at a premium to gain control — a prohibitive cost given the available floating supply.

Capital-relevant. Japanese accounting rules allowed banks to count a portion of unrealized gains on cross-held equity positions toward their capital requirements. As equity prices rose during the bubble, bank capital appeared to grow — even though no new capital had been raised — because the market values of cross-held positions had increased.

This capital accounting treatment was the primary mechanism by which keiretsu cross-shareholding amplified the bubble. Rising equity prices → higher unrealized gains → higher reported bank capital → more lending capacity → more lending → higher asset prices. The feedback loop worked as long as prices were rising.

The Implicit Guarantee System​

The main bank system embedded implicit guarantees that became explicit liabilities during the crisis.

In the keiretsu relationship, the main bank held a special role. It provided the primary corporate banking services, held a stable equity position, and typically had a representative on the borrower's board of directors. In exchange for this privileged relationship, the main bank was expected to provide rescue financing — additional credit, debt restructuring, even providing a bank officer to serve as a de facto turnaround manager — when a member company faced financial distress.

This rescue function had worked well during Japan's high-growth period. When a manufacturing company faced temporary difficulty, the main bank provided bridge financing that allowed restructuring and recovery. The implicit system substituted for the formal bankruptcy procedures that would have been used in a US context.

During the post-bubble crisis, the implicit guarantee system became a trap. Banks that had provided rescue financing to member companies were now providing continued lending to companies that were genuinely insolvent — not temporarily illiquid. The obligation to support members prevented banks from recognizing the loans as bad and restructuring or closing the member companies. The zombie dynamic was embedded in the keiretsu structure itself.

The Downside Amplification: 1990–2003​

The same mechanism that had amplified the bubble — cross-shareholding gains strengthening bank capital — worked powerfully in reverse during the crisis.

As equity prices fell from 1990 onward, the unrealized gains on cross-held positions shrank and eventually became unrealized losses. Bank capital, which had appeared robust based on the inflated cross-shareholding values, contracted. Reduced capital constrained lending. Reduced lending contributed to slower economic growth, which reduced corporate earnings, which reduced stock prices further.

The feedback loop on the downside was potentially worse than on the upside because the consequences of reduced bank capital were more immediate than the benefits of increased bank capital. When capital falls below regulatory minimums, a bank must either raise new capital or reduce its loan book — both actions that directly reduce economic activity.

Japanese banks in the 1990s were caught in a doom loop: declining equity prices reduced capital, reduced capital required asset sales or reduced lending, reduced lending weakened the economy, economic weakness reduced equity prices further. The only exit from this loop was either rapid capital injection (which the government was initially unwilling to provide) or a rapid equity market recovery (which did not occur).

Corporate Governance Failures​

The keiretsu structure produced corporate governance failures that contributed to the economic damage of the lost decades.

No hostile takeover discipline. In US markets, underperforming management faces the threat of hostile acquisition — a buyer who believes management is underperforming can offer a premium to shareholders, circumventing management opposition. With 30–50 percent of shares in stable cross-held positions, this mechanism was unavailable in Japan. Management was effectively insulated from market discipline.

Board dominated by insiders. Japanese corporate boards in the 1980s and 1990s were composed almost entirely of insiders — current and former executives of the company — rather than independent directors. Boards were incapable of providing the oversight and correction that would have required identifying strategic mistakes and replacing management.

Prioritization of employees and banks over shareholders. Japanese corporate governance traditionally prioritized the interests of employees (lifetime employment commitment) and main banks (loan repayment certainty) over shareholders (profit maximization). While these priorities had social benefits, they reduced the incentive to maximize capital efficiency and allocate resources to highest-return uses.

No market for corporate control. Without takeover discipline, poorly managed assets could not be transferred to better management without the incumbent management's cooperation — which was generally not forthcoming. This reduced the economy's ability to reallocate capital from low-productivity uses to high-productivity ones.

The Unwinding of Cross-Shareholdings​

Corporate Japan began unwinding keiretsu cross-shareholdings in the 1990s and accelerated this process in the 2000s, driven by several forces:

Accounting changes. Revision of Japanese accounting standards to require mark-to-market valuation of equity positions (rather than historical cost) removed the benefit of hiding unrealized losses and created incentive to sell depressed cross-held positions rather than continue carrying them.

Capital requirements. Under Basel international bank capital standards, cross-held equity positions received unfavorable capital treatment. Japanese banks facing capital pressure had incentive to sell cross-held equity positions to reduce risk-weighted assets.

Shareholder activism. Foreign investors — particularly US institutional investors, who became significant shareholders in Japanese equities through the 1990s and 2000s — applied pressure for governance reforms including reduced cross-shareholdings, more independent boards, and greater dividend returns.

The unwinding has been substantial but incomplete. Studies suggest that Japanese cross-shareholding ratios (the fraction of a company's shares held by other Japanese companies in stable positions) fell from roughly 45–50 percent in the early 1990s to approximately 20–25 percent by the mid-2010s. The unwinding process itself sometimes created selling pressure on declining markets, as companies that had been prevented from selling cross-held positions began doing so.

Common Mistakes in Analyzing the Keiretsu​

Treating keiretsu as purely negative. The keiretsu system had genuine economic benefits: reduced transaction costs within the network, patient capital that enabled long-term investment, efficient supply chain coordination. The problem was not the structure itself but its interaction with the specific conditions of a post-bubble crisis. A structure that provides stability during normal times can provide excessive stability — resistance to necessary restructuring — during crises.

Assuming Japan's governance model is uniquely problematic. Other East Asian economies had similar corporate governance structures (family-controlled conglomerates with bank relationships in Korea, Singapore, Indonesia). Their experiences with post-crisis restructuring varied significantly based on how aggressively governance reforms were implemented. Japan was not uniquely resistant to reform; it was slow to reform relative to what the severity of its crisis required.

Frequently Asked Questions​

Are keiretsu still important in Japanese business today? Cross-shareholdings have declined substantially from their peak. However, the concept of stable, long-term relationships between banks and corporate customers — and among corporate partners — remains more prominent in Japan than in the US or UK. The governance reforms of the 2010s-20s (particularly under the Abe government's Corporate Governance Code) have produced further changes, but Japan's corporate governance still differs materially from Anglo-American norms.

Did the keiretsu system prevent any foreign acquisition of Japanese companies? The combination of cross-shareholdings and cultural resistance to foreign ownership made Japan one of the most difficult major economies for foreign acquirers to enter. The reforms of the 1990s-2000s made it somewhat easier, and there have been notable foreign acquisitions. But Japan remains below-average among OECD countries in the level of inbound foreign direct investment as a percentage of GDP.

Summary​

The keiretsu system — Japan's distinctive corporate structure of cross-shareholdings, main bank relationships, and inter-corporate obligations — was the mechanism through which asset price changes were amplified across the economy. Rising equity prices during the bubble increased cross-shareholding values, strengthened apparent bank capital, and enabled further lending. Falling prices during the crisis reversed this process, creating the doom loop of declining capital, reduced lending, economic weakness, and further equity decline. The system's governance features — insulation from hostile takeover, insider boards, employee and bank priority over shareholders — prevented the creative destruction that would have facilitated recovery. The gradual unwinding of cross-shareholdings since the 1990s has improved Japanese corporate governance, but the system's legacy shaped the character of Japan's lost decades as fundamentally as any purely macroeconomic factor.

Next​

→ Zombie Banks and Forbearance