Aozora Bank: long-term performance & turning pointsSales (revenue) and profit-margin ratio
Sales (¥ bn)Net margin (%)
1957The property lender among the long-term credit banks
1957Nippon Fudosan Bank established under the Long-Term Credit Bank Law
1964Lists on the Tokyo Stock Exchange; starts foreign-exchange business
1977Renamed the Nippon Credit Bank
1997Three affiliated non-banks declared bankrupt; ¥290bn capital raise
1998Placed under special public management; shares delisted
In April 1957 the Nippon Fudosan Bank was established under the Long-Term Credit Bank Law with capital of $2.8M (¥1bn). It was the third of Japan’s three long-term credit banks, but it stood apart from the other two: where the Industrial Bank of Japan and the Long-Term Credit Bank of Japan practised pure industrial finance, Nippon Fudosan funded itself by issuing bank debentures and lent long against real-estate collateral. That design carried its own weakness from the first day — earnings tied to the market value of the security rather than to the borrower’s industry. Decades later a president, Kubota Hiroshi, described the founding impulse plainly: the bank had stretched itself trying to make up for arriving late and catch the Industrial Bank of Japan.
It listed on the Tokyo Stock Exchange in September 1964, began foreign-exchange business the same year, and in 1977 renamed itself the Nippon Credit Bank in an attempt to grow from a property specialist into a full long-term credit bank; a trust subsidiary followed in 1994. Internally it remained closed. For years the man called the bank’s founding father ran it personally — its directors were described as a set of yes-men — with decisions settled between the chairman and a single responsible executive. The first internally promoted president, appointed in 1987, tried to break that habit, launching a bank-wide sales-information system in 1990 to stop field intelligence disappearing into what staff called the bank’s “Bermuda Triangle of information.” Its people were the youngest of any city or long-term credit bank, average age thirty-two, and its operating profit was 43% of IBJ’s.
The bubble’s collapse exposed the design. By March 1996 the bank alone carried ¥1.4 trillion of bad loans, and three affiliated non-banks were foundering under ¥2 trillion of debt. Kubota refused a merger — he wanted “a bank combining real-estate-backed finance and market business” — and in 1997 pushed through the bankruptcy of the three non-banks, a retreat from overseas, and a ¥290 billion capital increase. Togo Shigeoki, from the Bank of Japan, took over as president that August and told the institutions that had subscribed he would “return it to you with the share price ten times higher.” Seventeen months later, in December 1998, the bank was placed under special public management and delisted — the second long-term credit bank to be nationalised in three months, and the beginning of the end of the long-term credit bank as an institution in Japan.
2000Reprivatised to a SoftBank / Orix / Tokio Marine consortium
2001Renamed Aozora Bank
2003Cerberus becomes lead shareholder
2006Converts to an ordinary bank; relists on the TSE first section
In September 2000 special public management ended and the bank was sold to a consortium led by SoftBank, Orix and Tokio Marine — the first time a failed Japanese bank had been handed to a privately led rescue rather than kept on the state’s books. In January 2001 it dropped the name that had failed and became Aozora Bank. The transition was not smooth: the man designated as its first president died suddenly in November 1999, strong candidates declined the job, and it was December 2000 before Maruyama Hiroshi of Orix took the post. His programme was deliberately unglamorous — nine-tenths of income from ordinary banking, capital from the many regional banks that had invested channelled into small-business and venture lending, caution about the internet ventures SoftBank favoured.
That lasted two years. In 2003 SoftBank sold its stake to the American fund Cerberus, producing something rare in Japanese banking: a bank run by a US private-equity firm. Cerberus treated Aozora as a platform for consolidating Japanese finance, installed international bankers, and replaced the long-term credit bank’s instincts with a governance that put return on capital first. It also rebuilt the asset-liability and credit-investment machinery on a different design — the foundation on which the later expansion into overseas credit would be built.
In April 2006 Aozora surrendered its long-term credit bank licence and converted to an ordinary bank, set up a securities arm the same month, and in November returned to the first section of the Tokyo Stock Exchange, eight years after being delisted. Forty-nine years of statutory constraint fell away at once. What replaced it was freedom — and exposure — in equal measure: the seeds of both the internet retail deposit franchise and the overseas credit portfolio were sown in these years under foreign ownership.
2009Net loss of $2.6B (¥243bn); internet deposit business launched
2012Capital restructuring plan sets the repayment framework
2015Public funds repaid in full, fifteen years after the failure
2017Head office moves to Kojimachi, Chiyoda-ku, Tokyo
2018GMO Aozora Net Bank launched for corporate customers
Eighteen months after relisting, the second cliff arrived. Impairments on the overseas credit portfolio built up under Cerberus produced an ordinary loss in the year to March 2008, and a net loss of $2.6B (¥243bn) the year after. Ordinary income shrank by more than a quarter over two years. The pattern of the bubble repeated in a new asset class: a bank that had to hold a distinctive position had concentrated its risk, and the concentration turned against it.
The response came from the same constraint that had caused the damage. In April 2009, having no branch network at all, Aozora began taking retail deposits over the internet — turning zero counters from a handicap into a funding model with no branch cost. Built out from 2019 as the BANK brand, it held deposits in place even through negative interest rates. Meanwhile the balance sheet was repaired: a return to profit in the year to March 2011, a capital restructuring plan in August 2012 that set the terms for repaying the state, and a string of subsidiaries — regional research, asset management, real-estate investment advisory — that turned a bank into a group.
In June 2015, fifteen years after the failure, Aozora repaid the public funds in full. It was the first moment in its history at which it could draw its own blueprint: free of the long-term credit bank statute, free of the state, and free of the American fund that had exited two years earlier. What it chose was to stay branchless — corporate structured finance on one side, internet retail deposits on the other — and in October 2018 it went further, launching GMO Aozora Net Bank with GMO Internet as Japan’s first internet bank built specifically for corporate customers, gathering settlement balances that a rate rise would turn into interest income.
2019The niche investment bank, and the end of independence
Revenue (¥ bn, bars) · net margin (%, line)
Source: securities reports
FY2019 · consolidated
Revenue$1.5B
Net income$331M
Net margin22.5%
→
FY2026 · consolidated
Revenue$1.5B
Net income$162M
Net margin10.6%
2019Tanigawa Akira becomes president; BANK deposit brand launched
202015% stake in Vietnam’s Orient Commercial Joint Stock Bank
2024Net loss of $329.4M (¥50bn) on US offices and foreign bonds; Omi Hideto becomes president
2024Daiwa Securities Group takes 24% via a $342.6M (¥52bn) placement
Under Tanigawa Akira, president from June 2019, Aozora sharpened the logic to a point: dig deeper into domestic niches it had entered before the megabanks arrived — LBO finance, venture debt, environmental finance — and let the time advantage of an early entrant do the work that scale could not. LBO loan balances passed ¥300 billion from fiscal 2022; a 15% stake in Vietnam’s Orient Commercial Joint Stock Bank in 2020 added a Southeast Asian foothold. The BANK deposits funded it all without a single branch.
In February 2024 the third cliff appeared. Aozora announced a net loss of $329.4M (¥50bn) for the year, driven by provisions on non-recourse loans against US offices — where remote work had cut valuations in New York, Chicago and San Francisco until borrowers sat above 100% loan-to-value — and by mark-downs on a foreign bond portfolio caught by the rate rises of 2022. It was the first loss since Lehman, and the dividend was cut with it. Omi Hideto replaced Tanigawa that month, promising to “push forward an Aozora-style investment banking business matched to social change.” The logic of lending against property, which had defined the bank in 1957, had failed again — this time in American offices.
Capital was now too thin to rebuild alone. On 1 July 2024 Aozora issued $342.6M (¥52bn) of new shares to Daiwa Securities Group, handing it 24% of the voting rights and a board seat. The alliance is meant to graft Daiwa’s listed-company network and distribution onto Aozora’s LBO, real-estate finance, M&A advisory and wealth businesses, with a target of ¥10 billion of additional operating profit by fiscal 2027. For a bank that had belonged to no group for more than twenty years, it was the plainest possible statement that rebuilding the earnings base mattered more than the pride of independence — and the fourth wholesale reconstruction of the institution, after the long-term credit bank’s dissolution, the Cerberus years, and the repayment of the state.
A design that could not survive the fall in collateral values
Togo Shigeoki’s line — “all I can do is return it to you with the share price ten times higher” — reads less as an ambition than as a confession of debt to the institutions that had subscribed to the capital increase. By the time ¥290 billion had been raised, the only road left to the Nippon Credit Bank was to write off its bad loans within whatever time the market allowed, and that time ran out in seventeen months. There is nothing in the contemporary press that lets the responsibility be narrowed to one person; what appears again and again is the explanation that the design chosen in 1957 — to make up for a late start by lending against real-estate collateral — could not withstand a fall in the value of that collateral.
It is too early to say that reprivatisation saved this bank. Even after the private consortium took it on, the obligation to repay the ¥320 billion that had been injected remained, the loan book stayed shrunken, and what borrowers were left were heavily in property. When Maruyama Hiroshi said that “recovering the lost ground comes first,” that was, it appears, a candid expression of the fact that the content of the rescue was still blank. Even so, the framework itself — the state declining to carry the failure alone and preparing a private vehicle to receive it — survived as the template for the resolutions that followed.
A bank held as an investment redraws itself with each owner
Sumitomo Mitsui Banking Corporation put its hand up, and the seller decided on terms anyway. SoftBank’s position was consistent: the Aozora shares were something to be “sold at the time of listing for a profit,” and with its own bond redemptions running at over ¥40 billion a year there was no reason not to pick the higher bidder. Who owns a bank was settled here, it seems fair to say, less by the design of financial policy than by the seller’s cash position.
Nor were the buyer’s ten years spent growing a bank. Cerberus did force through the conversion to an ordinary bank and the relisting, but the shares closed their first day at ¥502 against an offer price of ¥570, the loan book remained 26.9% weighted to the property sector, and in 2013 the fund exited its stake for around ¥150 billion. That outside capital can push a change of business model through quickly is one thing; that the change takes root is another. A bank held as a pure investment, it could be said, had its outline redrawn each time the shareholder changed.
The deadline of October 2012 determined the shape of this decision. If ¥155.3 billion could not be repaid, the government’s preferred shares would convert into common stock and the state would join the shareholder register as an ordinary owner. What Aozora chose was to cut capital from ¥420 billion to ¥100 billion to create the resources, and to repay in cash — a ¥22.7 billion buy-back plus a special preferred dividend of ¥20.5 billion a year. A capital reduction is not a policy that sounds good, but at the time no other route appears to have been visible that avoided dilution and state involvement at once.
That said, finishing in three years was not achieved by its own efforts alone. The share price had recovered, and Cerberus, the largest shareholder, stepped off its holding for about ¥150 billion — which made it possible to complete the repayment and tidy the capital structure in the same movement. Set that beside Shinsei Bank, whose preferred shares had already been converted to common stock and which therefore could not negotiate a staged repayment, and the difference in outcome looks less like a difference in management effort than a difference in how the public funds were injected in the first place: as preferred stock, or as common.
The order of events — deciding on a ¥51.9 billion capital increase three and a half months after suspending the dividend — tells you the character of this decision. Since the year to March 2016, after repaying the public funds, Aozora had flown a payout ratio of about 50% and gathered individual shareholders with quarterly dividends; it took that sign down and still did not have enough capital. That the counterparty was Daiwa Securities Group owes much to a commercial reason — a client base in which Daiwa acts as lead manager for half of Japan’s roughly 4,000 listed companies — but the closeness of the relationship also helped terms come together quickly: Daiwa had been lead manager of the relisting, and the two presidents had known each other for a decade.
What the alliance filled, however, was capital and a referral channel, not the earnings structure itself. The ¥10 billion of operating profit expected by fiscal 2027 is an estimate built bottom-up from more than two hundred working-level meetings, and remains to be confirmed in actual figures. Meanwhile, at the time of writing, the retail deposits gathered at high rates have run off by ¥376.3 billion in a year, and a deposit yield of 0.43% is above the assumption in the medium-term plan. What a bank that belonged to no group for more than twenty years obtained by giving up its independence was time and optionality — not a guarantee of earning power.
Each heading links to the full Japanese analysis — background, decision and outcome, with sources.
This is a condensed English edition. The full, source-by-source history — with the detailed narrative, financial tables, shareholders and executives — is maintained in Japanese: 日本語版（詳細）— Aozora Bank full history in Japanese →
Aozora Bank, Ltd. — 有価証券報告書 (annual securities reports), 統合報告書 (integrated reports) and earnings briefings (決算説明会).
Nikkei Business — 日経ビジネス (Nikkei BP), 16 July 1990.
Nikkei Business — 日経ビジネス (Nikkei BP), 5 August 1996 (interview with Kubota Hiroshi, president of the Nippon Credit Bank).
Shukan Toyo Keizai — 週刊東洋経済 (Toyo Keizai Inc.), 13 September 1997 (interview with Togo Shigeoki).
Shukan Toyo Keizai — 週刊東洋経済 (Toyo Keizai Inc.), 20 January 2001 (profile of Maruyama Hiroshi, president of Aozora Bank).
Nihon Keizai Shimbun — 日本経済新聞 (Nikkei Inc.), 26 January 2024.
This page is provided for general information only and is not investment advice, nor a recommendation to buy or sell any security.
Figures are compiled independently and include our own estimates, approximations and machine-processed data; we make no warranty as to their accuracy or completeness.
Sources are primarily each company’s securities reports and other public filings, but errors and omissions may remain.
Any use of this information is at the reader’s own risk. Past performance does not indicate future results.
Company names, logos and other marks belong to their respective owners.
Data API
Aozora Bank’s history, financials, executives and
shareholders are published as static JSON — no key, plain GET.