On Aug. 3 the Portuguese government announced a €4.9 billion ($A7.04bn) bailout for Banco Espírito Santo -- another reminder that the 'too big to fail' doctrine still prevails six years after the financial crisis. At least in this case, junior bondholders (those who invested less than a year ago) and shareholders were forced to take a haircut. That's progress for those who argue that economic recovery is impeded when monetary and fiscal authorities rescue private institutions from the consequences of their decisions.
Too big to fail remains unresolved in the US. Last week the Federal Reserve and the Federal Deposit Insurance Corp said that not one of the nation's 11 largest banks could fail without threatening the broader financial system. The news came after regulators reviewed the banks' 'living wills', the emergency plans required under the 2010 Dodd-Frank law.
Living wills, according to resolution plans from the FDIC and the Fed, should "not rely on the provision of extraordinary support by the United States" and should "prevent or mitigate any adverse effects of such failure or discontinuation on the financial stability of the United States."
Most of the living wills, including those for the four largest banks -- JPMorgan Chase, Bank of America, Citigroup and Wells Fargo -- are based on a combination of Chapter 11 bankruptcy of the parent firm and sales of subsidiaries.
But selling business units in a liquidity or solvency crisis is hardly a viable plan. When Lehman Brothers filed for Chapter 11 in September 2008, the US financial system spiraled into disarray, which led to public intervention. And Lehman was small compared with Bank of America and Citigroup, both of which have roughly three times the assets that Lehman had when it went under.
Instead of living wills or government bailouts, we propose that banks issue a class of bonds to privately secure the financial system against a cascade of failures. Let's call them Class R or 'reorganisation' bonds. Class R bonds would function like any other corporate-debt instrument in normal times, meaning that bondholders would have no control over the corporation.
In the event of the firm's imminent failure, Class R bondholders would form a committee to develop contingency plans to appoint a new board of directors and reorganise senior management.
In bankruptcy, the existing board of directors would be dismissed, the equity of the firm would be eliminated and the Class R bonds would immediately be converted to equity. The Class R bondholders might take a haircut, but they would also become the owners of the bank, free of claims from prior management or shareholders. This bondholder capital avoids the involvement of the federal government, allows the firm to shed a substantial portion of its liabilities, but compensates Class R bondholders with control of a firm with a healthier balance sheet.
Bear Stearns is an example of how Class R bonds would work. Bear had about $350 billion in assets before its collapse, so if the firm had issued Class R bonds for 8 per cent of its assets, the face value of the bond issue would have been roughly $28bn. The firm was sold to JPMorgan Chase for $1.2bn packaged with $29bn in asset guarantees by the Federal Reserve. If the buyer needed guarantees of 24 times the sale price, the firm was worthless at the point of sale. Hence, Class R bondholders would have taken a substantial haircut when they came into possession of the firm.
Reorganisation through bondholder committees has a long tradition in corporate bankruptcy, but it has not been used with banks. In finance, the Fed and other regulatory agencies have been charged with guarding against systemic risk. But that approach did not prevent the 2008 crisis.
For depository institutions, our proposal would work well with policies in place at the FDIC. Since 2008 the FDIC has handled 489 small-to medium-size bank failures using procedures that transfer ownership of most assets, and responsibilities for most liabilities, to new owners.
This approach has been successful in avoiding serious bank runs during reorganization, but the Class R approach could improve the process. The FDIC could continue to insure deposits, but it would no longer need to assume loss-share agreements with a bank that is taking over a failed institution. The Class R bondholders would assume the losses.
For a well-managed institution, the risk in these bonds should be minimal. If the viability of the institution comes into doubt, it should be signaled in lower Class R bond prices, but their owners would have the potential upside of coming into control of the firm.
We estimate that banks would need to issue Class R bonds for up to 8 per cent of their institution's balance-sheet liability, which includes demand and time deposits, loans from other banks and corporate bonds.
In the event of a failure, the firm's liabilities will be reduced by 8 per cent. In every case except AIG, this was well in excess of the amount of support that the government provided to maintain firms in the 2008 crash; and in most cases the firm's loss approximated only about half of their outstanding corporate bonds. Crucially, to maximise a bank's ability to finance new growth after a failure, its losses need to be absorbed by incumbent investors to enable a fresh start.
The living wills discussion is part of the process of finding a better institution for dealing with bank failures. Our proposal seeks to promote systemic stability by securing each of its private elements.
Mr. Gjerstad is a presidential fellow at Chapman University in Orange, Calif. Mr. Smith is a professor of economics and finance at Chapman and a 2002 Nobel laureate in economics. They are the authors of "Rethinking Housing Bubbles" (Cambridge University Press, 2014).
This article was originally published at the Wall Street Journal. Reproduced with permission.