The economics of “Good bank-bad bank”
The spectre of nationalisation
From The Economist print edition
There are ways for governments to revitalise banks without taking them over
IT IS generally easier to remove a kidney from a dead donor than a live one. When regulators in Scandinavia and America in the early 1990s started extracting the bad assets from their crisis-hit banking systems, it helped that the banks they dealt with were bust or in the government’s hands. Today, policymakers are trying to excise toxic assets from banks that are still, at least officially, private and viable. That is a much trickier proposition.
Last autumn, governments around the world poured new capital into private banks and guaranteed their debts to protect them from further losses and help them raise private capital. But continued losses have overwhelmed those initial efforts. Some banks have needed more capital, and a few have been nationalised outright. Moreover, the haphazard implementation of rescues has kept private capital on the sidelines, fearful of being diluted or wiped out. What is needed, the experts say, is a more systematic approach through the creation of a “bad bank” to assume the bad assets, leaving “good banks” to resume lending. (In an Orwellian attempt to hide the nastiness, it may be known in America as an “aggregator bank”.)
The good bank/bad bank terminology dates at least back to 1988 when America’s Mellon Bank spun off its bad energy and property loans into Grant Street National Bank, which was financed with junk bonds and private equity. Such purely private solutions are not feasible during crises that encompass the entire banking system: there is not enough private capital around. In the early 1990s the governments of Sweden and Finland each nationalised some of their largest banks and set up “bad banks” to dispose of their assets. Around the same time, America created the Resolution Trust Corporation to sell off the loans and underlying collateral of hundreds of failed savings banks, or thrifts. In each case, the assets taken over by the bad bank were equal to about 8% of GDP, according to a study by Daniela Klingebiel of the World Bank.
In this crisis policymakers have adopted piecemeal elements of the good bank/bad bank. In October UBS spun $60 billion of toxic assets into a fund backed by the Swiss central bank. America will absorb most of the losses on $306 billion of problem assets at Citigroup and $118 billion at Bank of America. It is also creating a facility, supported by the Federal Reserve, for asset-backed securities which could relieve banks of bad loans. Britain may insure banks against future losses. Like a bad bank, the aim is to isolate toxic assets, encourage private money to come in and discourage banks from hoarding their capital. But no market value has been put on them (although loss-sharing agreements in part serve that purpose). This spares banks from immediately recognising their losses, but it leaves a fog of uncertainty over the system. Banks will not boost lending if they fear that future loan losses will eat through the rest of their capital.
A bad bank could alleviate these concerns by convincing both banks and investors that the problem assets have either been removed from the banking system or will be as they surface. Paul Miller of Friedman Billings Ramsey, an investment bank, says a government bad bank can pay more for assets than a private investor because its cost of funds is irrelevant, it needs no capital and can hold the assets to maturity. It could also develop a professional and uniform approach to valuing and disposing of bad assets while leaving new lending decisions to the good banks.
But a bad bank faces different problems, the most serious of which is setting a price for assets that both it and the seller can agree on. This was less of an issue in the early 1990s, since the assets for the most part came from banks that had already failed or were under government control. Today, if the bad bank pays above the fair-market value, it would raise the cost to taxpayers, imperil its political legitimacy, and deprive the market of badly needed transparency. If it pays fair value or less, banks might be reluctant to participate. Those that did may have to recognise large, immediate losses, depleting their capital. As a result, setting up a bad bank would entail additional capital injections. To reassure itself that the recipient bank can survive, the government would invest only if the bank can simultaneously raise funds from private investors. Any bank unable to raise private capital, perhaps rendering it insolvent, would be taken over.
But such steps would be time-consuming and could be hit by a loss of market confidence at any moment. America abandoned its original plan to buy toxic assets last October in favour of extra capital because the crisis demanded faster action.
An alternative (or perhaps prelude) to a bad bank would be nationalisation. This would at a stroke end the tension between the goals of private shareholders who want to hoard capital and lend less, and government overseers who want banks to lend more and modify mortgages of homeowners facing foreclosure.
But nationalisation carries huge costs of its own. With the world awash in unwanted bank assets, it could take years for the governments to privatise their banks. Meanwhile politicians would be tempted to turn banks into instruments of industrial policy, propping up politically powerful industries such as carmakers and scrimping on more deserving recipients. Politically motivated lending could result in even larger loan losses in the future, and private banks would be put at a disadvantage. At the other extreme, governments might be so fearful of taxpayer losses that they lend even less than their private counterparts.
Economists have long recognised that banks are special. Through decades-old relationships with millions of households and businesses, they normally (though, sadly, not recently) steer savings to productive and lucrative endeavours. Letting banks collapse would wipe out this critical mechanism; nationalising them could, eventually, do it similar damage.