Now that every bank officially passes the stress tests, and even the bank’s judged to need more equity are planning to simply convert government preferred shares into common equity, we’ve pretty much forgotten what the point of this exercise was supposed to be. It’s hard to avoid the impression that the whole thing was some kind of marketing exercise for the banking sector.
A far better stress test would be one that actually allowed for the possibility of failure by large banks such as Citigroup and Bank of America. That idea still strikes terror into the hearts of banking regulators under the mistaken impression that the failure to prevent the bankruptcy of Lehman Brothers somehow caused the credit markets to go into cardiac arrests. (In reality, the collapse of Lehman Brothers simply provided a signal to the markets that the financial sector was far less healthy than previously believed, and markets corrected to incorporate this information.)
In a great essay in the Financial Times today Nouriel Roubini and Matthew Richardson explain that keeping insolvent banks afloat is a terrible idea. By far the most important thing we can do is re-establish the system where insolvent banks are permitted to sink. Far better than any government regulation, a working market check on banks would prevent reckless banking. And the key to this is convincing bank creditors that we’re not going to bail them out anymore.
From the FT:
Once again, the question will be how the near-insolvent banks can be kept afloat, to avoid systemic risk. But the question we really should be asking is: why keep insolvent banks afloat? We believe there is no convincing answer; we should instead find ways to manage the systemic risk of bank failures…
That leaves the creditors – depositors, short- and long-term debt-holders and preferred shareholders. For the large complex banks, about half are depositors. To avoid runs on these deposits, the government has to provide a backstop. But it is not clear it needs to cover other creditors of a bank, as the failures of IndyMac and Washington Mutual attest.
Even if systemic risk were still present, the government should protect the debt (up to some level) only of the solvent banks, not the insolvent ones. That way, the risk of the insolvent institutions would be transferred back from the public to the private sector, from the taxpayer to the creditors.
The government may be able to avoid the mess by persuading long-term creditors to swap their debt for equity, at a loss. The recent failed effort with Chrysler suggests this will not be easy. But a credible threat of bankruptcy could scare creditors into negotiation, to avoid bigger losses.
Suppose the systemic risk problem is solved. The other argument against allowing banks to fail is that after a big loss by creditors, no one would be willing to lend to banks – which would devastate credit markets. However, the creative-destructive, Schumpeterian, nature of capitalism would solve this problem. Once unsecured debtholders of insolvent banks lose, market discipline would return to the whole sector.
This discipline would force the remaining banks to change their behaviour, probably leading to their breaking themselves up. The reform of systemic risk in the financial system would be mostly organic, not requiring the heavy hand of government.
Why did creditors not prevent the banks taking excessive risks before the crisis hit? For the very same reason creditors are getting a free pass now: they expected to be bailed out. For capitalism to move forward, it is time for a little orderly creative destruction.
The only problem with Roubini and Richardson’s approach is that it still allows for guarantees of solvent banks. But the presence of any guarantee obscures who is and who isn’t solvent. The judgment would have to be made by bureaucrats, rather than markets. Which brings us right back to the stress tests, where all banks are solvent. A better idea would be to strip out all guarantees, except deposit guarantees, and allow the market to judge solvency.