A previously successful and private approach to improving banks with large amounts of troubled assets could promise real help for today’s subprime mortgagestrained financial institutions. The subprime mortgage crisis and its spread to other lending sectors has meant extreme difficulty not just for the U.S. credit markets, but for those overseas as well. It also has caused concern about the well-being of Freddie Mac and Fannie Mae. At the same time, leading banks and brokerage houses are taking significant writedowns, which have impacted their viability and forced the sale of some of them, such as Bear Stearns, albeit primarily because of liquidity concerns. Meanwhile, average borrowers are locked out from traditional sources of mortgage funding and working business capital.

Against the backdrop of the obvious political problems and economic challenges presented by a government-financed solution or bailout, it is time to revisit a successful bank restructuring model that apparently has been disfavored by regulatory policymakers. This “Good Bank-Bad Bank” solution relies on traditional, private market forces to save distressed and potentially failing banks, rather than relying on government spending. The Good Bank-Bad Bank structure presents a viable alternative approach for banks to divest low quality assets and, thereby, allows management to focus its efforts and thinking on the future and on growth.

Background and Description of Mellon Transaction

Twenty years ago, U.S. banks faced another crisis brought on by a significant decline in real estate values and the collapse in the price of oil. In late October 1988, the OCC chartered Grant Street National Bank (in Liquidation), Pittsburgh, Pennsylvania (“Grant Street”), as a de novo uninsured national bank. Grant Street was structured to liquidate, pursuant to a Plan of Voluntary Complete Liquidation (“Liquidation Plan”), a portfolio of low quality assets, and then go out of business.1 Mellon Bank Corporation (“Mellon”), then a Pittsburgh-based bank-holding company, created and organized Grant Street as a way to remove low quality assets from Mellon and its subsidiaries’ balance sheets.

Mellon sold its low quality assets to Grant Street, and Grant Street’s shares were dividended to Mellon’s shareholders as a stand-alone liquidating national bank. Grant Street’s assets were serviced and divested pursuant to a Liquidation Plan approved by the OCC. The liquidation was implemented by a wholly owned Mellon nonbank subsidiary, Collection Services Corporation (“CSC”) (approved by the Federal Reserve Board as a section 4(c)(8) nonbank subsidiary), which had an arm’s-length service contract with Grant Street. In this way, CSC undertook certain responsibilities to administer the assets of Grant Street and to ensure they were liquidated, while it maximized the returns from the liquidation.

The central benefit of the “Good Bank-Bad Bank” solution is both direct and indirect: It allowed Mellon to clean up its balance sheet, while enabling existing management to focus on growing Mellon’s businesses, rather than being distracted with troubled asset workouts. The outcome is a well-known success story, as the “Good Bank” morphed into the global BNY Mellon asset manager and service provider powerhouse.

The specifics in the restructuring warrant examination. The assets Grant Street purchased consisted of 191 loans made up of commercial and industrial loans and leases (including participation interests); real estate interests (including 1,140 residential mortgages that were deemed one asset); and energy properties. No foreign or LDC debt (loans to Lesser Developed Countries) was sold to Grant Street. The assets sold to Grant Street had an origination value of almost $1.4 billion. By the time the assets were sold to Grant Street, their book value was written down to about $640 million, or 47 percent of their original valuation.

The purchase was funded by the public sale of two classes of short-term, extendible pay-through notes ($225 million of Class A-1 Notes, due Nov. 1, 1992, and $288 million of Class A-2 Notes, due Nov. 1, 1994), aggregating $513 million, at interest rates of 10.25 and 14.25 percent, respectively. The balance of the purchase price was funded by Mellon contributing $35 million for Grant Street’s common stock, $90 million for Grant Street’s senior preferred stock (which had less than a 5 percent voting power), and $2.6 million of Grant Street’s junior preferred stock for Grant Street’s directors. Grant Street’s common stock was declared a special dividend and was spun off to Mellon shareholders, and Grant Street was established as a discrete unaffiliated entity.

As indicated above, Grant Street’s assets were serviced by CSC, the Mellon workout subsidiary hired by Grant Street to liquidate the assets. All the notes were paid off before maturity, and Grant Street’s existence was terminated July 31, 1995.

Grant Street was the only operating limited-life liquidating national bank established with federal banking agency approval. Since Grant Street, the OCC has not chartered a bank whose sole purpose was to liquidate assets and then go out of existence. The OCC likely believed that the Grant Street transaction raised a policy issue that a newly chartered bank should service its communities and markets on an ongoing basis, and not be established with the goal of being liquidated.

For today’s troubled banks, the basic premise and benefits of such a structure has renewed promise, as the OCC may now be willing to reconsider the Grant Street example as a private sector approach to dealing with the many problem assets that are currently on national bank and thrift balance sheets.2

Good Bank-Bad Bank Approaches Within and Outside the United States

Frank Cahouet, Mellon’s CEO from 1987 to 1998, originally used a variation on the Good Bank-Bad Bank structure in 1986 when he was CEO at Crocker National Bank and was brought in to clean up Crocker’s asset problems. In the case of Crocker, Britain’s Midland Bank PLC, Crocker’s parent, downstreamed capital in exchange for $3.5 billion of Crocker’s nonperforming loans that were transferred to a workout subsidiary of Midland, and were removed from Crocker’s balance sheet before Crocker was sold to Wells Fargo. The Mellon restructuring through Grant Street borrowed from and expanded considerably upon the Crocker scenario.

In connection with the 1987 Chemical Banking Corporation acquisition of Texas Commerce Bancshares, the lower quality loans of Texas Commerce were written down and transferred into a stripped-down existing subsidiary bank of Texas Commerce, i.e., a “bad bank,” which was then spun off to Texas Commerce’s shareholders. First Interstate Bancorp (of California) used a very similar approach when it acquired Allied Bancshares of Houston, in 1988. The same basic structure was used in the case of FDIC-assisted takeovers involving First City Bancorp and Nations Bank’s purchase of First Republic Bank of Dallas. All of these involved problem banks in Texas.3

The Good Bank-Bad Bank structure concept was also tried overseas in Japan and Sweden. Japan tried it during the 1990s with both Japanese government involvement and using an industry approach. In December 1992, the Bankers Association of Japan established the Credit Cooperative Purchasing Company as a centralized receptor of bad debts. Also, in 1994, Japanese banks were allowed to establish “bad banks” as Special Purpose Companies to “facilitate bad debt restructuring.” In 1996, the Housing Loan Administration Corporation and the Resolution and Collection Bank were established as receptors for bad real estate and credit co-op assets. These were subsidiaries of the Japanese counterpart to our FDIC.4 Sweden has also used a Good Bank-Bad Bank structure with the government’s asset management corporation, Securum, playing a key role in disposing of nonperforming assets.5

Germany also recently considered a variant on the Good Bank-Bad Bank structure in working out troubled assets at West LB, Germany’s third biggest landesbank. West LB had a significant amount of low quality collateralized debt obligations tied to U.S. subprime mortgages. It also had trading problems. West LB created a “Bad Bank,” an off-balance-sheet entity to hold about $33 billion of risky assets that were removed from West LB Bank’s balance sheet.

With the current subprime mortgage crisis and its spread to other types of balance sheet assets, top executives from major banking organizations have hinted at various forms of the “Good Bank-Bad Bank” structures using separate portfolios or subsidiaries to segregate troubled assets. For example, in reported discussions involving the monoline insurer Ambac Financial Group and the New York State Insurance Commissioner, consideration was given to splitting the business into a lower-risk municipal bond business on the one hand, and a higher-risk structured-finance business on the other. The contemplated proposal as reported included at one time a significant capital injection that would both strengthen Ambac’s capital, and enable it to absorb any writedowns it might need to take while, hopefully, preserving its triple-A rating. A key issue is whether these institutions will be allowed to take the troubled assets off their balance sheets, which was Mellon’s objective with Grant Street. 

Benefits of Such Restructuring

The Grant Street transaction had very positive results for Mellon’s shareholders, which retained their ownership of the Good Bank. First, investors place a much higher valuation, as they should, on an organization with a lower proportion of troubled assets, because of the enhanced earning power and stronger capital position.

Second, divestiture of lower quality assets immediately improves the Good Bank’s debt ratings and can significantly lower its funding costs. All of this translates into an immediate earnings impact.

Third, a most valued advantage of the restructuring was that it freed Mellon management from being preoccupied with having to deal with problem assets to focus upon the company’s future—which it did extremely well and to the Mellon shareholders’ significant benefit. This allowed Mellon to become an acquirer within less than two years.

Fourth, isolating the collection work into a separate nonbanking subsidiary whose client was an unaffiliated bank allowed the collection process to proceed unencumbered by the normal bank desire to protect borrower relationships. Also, the focus of the collections staff was to monetize the troubled assets within a short timeframe, so they could pay off the Class A-1 and A-2 Notes.

Fifth, the segregation of troubled assets at the Bad Bank allowed Grant Street’s management and CSC’s employees to focus on maximizing net present value. This approach deviates from the typical banker mentality “to pursue the last dollar of the problem loan no matter how long it may take.”6 This is a painful lesson to learn in the current environment, plagued by significant writedowns over time. Furthermore, CSC’s employees knew they would retain their jobs at Mellon even after the Grant Street assets were sold.

It is critical that raising capital precedes or accompanies such a restructuring at the Good Bank, so that after the writedowns on the transfer of assets, the Good Bank remains in a regulatorily acceptable capital position. In the case of Mellon, the holding company replenished its capital base even after the writedowns for Grant Street, by raising $525 million in equity capital, including $158 million from the Warburg Pincus investment funds.


Critical to the success of Grant Street was the mix of its assets, the liquidation of which provided the cash flow to pay off the $513 million in Grant Street Notes. The largest component of Grant Street’s assets were the commercial loans and leases, which represented almost 45 percent of Grant Street’s book value and were the most readily monetized. Of note is that some large borrowers decided to immediately pay off their delinquent loans rather than have them sold to Grant Street. The early cash flows and early monetization of Grant Street’s assets helped carry Grant Street through the longer workout timeframe required for the disposal of real estate assets.

The valuation process used by Mellon for each asset transferred to Grant Street was very important—that process used a liquidation scenario with a forecasted cash flow (using discount rates) to determine the fair market value of each asset. The valuation process was developed with the assistance of the Arthur Andersen & Co. accounting firm. The results were checked by Kenneth Leventhal & Company, which reviewed the reasonableness of the assumptions. The last step was a sensitivity analysis done by an independent third party to test the potential effect of variations from the underlying assumptions with respect to oil prices, economic growth, inflation and interest rates.7

In the current treacherous financial environment, all manner of rational alternatives for helping to resolve the weak balance sheets of imperiled banking organizations should be considered. A private sector approach that has worked in the past is the Good Bank-Bad Bank structure. It can work again.