Sunday, January 10, 2010

Market discipline is not the answer

In November Barron's published my article "Market discipline is not the answer". Due to space considerations, it has to be edited to fit. Below is my original draft, that covers a bit more:

Market discipline is not the answer

Christopher T. Mahoney

There’s a growing consensus on both sides of the Atlantic that market discipline must be reintroduced into the financial system. Regulators have advocated that steps be taken to make it possible for large banks to fail, at a loss to depositors and other creditors—to reduce or eliminate the concept of “too big to fail.”

This view, while appealing to adherents of free markets, is nonetheless very dangerous. At times of crisis, the failure and default of a systemically important institution will lead to contagion and a run on the financial system, necessitating a system-wide rescue or precipitating a financial collapse with unknowable consequences. And, wishes and dreams notwithstanding, it is simply not possible to uncreate systemically important financial institutions by slimming them down or breaking them up.

The history of financial crises is the history of the threatened failure of financial institutions previously considered systemically unimportant, but nonetheless threatening to cause a contagious run on the entire system. Runs occur on solvent banks during panics because there is insufficient information in the public domain to allow depositors to discriminate between the strong and the weak. Bank accounting is complicated, backward-looking and inexact. Who really knew in September 2008 what Lehman's true solvency was, or Goldman's for that matter?

Every financial crisis in the past two centuries began with the threatened failure of a second- or third-tier institution: Barings in 1890, Knickerbocker Trust in 1907, Bank of United States in 1931, Sanyo and Yomura Securities in 1997, Long Term Capital Management in 1998 and in the most recent crisis: IKB, Northern Rock, Hypo Real Estate, Bear Stearns, AIG, and Lehman.

In every case, either the authorities allowed the threatened institution to fail, resulting in a full-blown systemic crisis (Bank of United States, Sanyo Securities, Lehman) or the authorities blinked, ignored their own rules, and mounted an extra-legal rescue to protect the system.
Leading up to crises, the regulatory emphasis has been on market discipline (indeed, even going so far as to ban bailouts), but by mid-crisis, the entire system ends up under the safety net. No one would ever have imagined in 2006 that every building society in the UK was “too important to fail,” but they were so deemed.

If systemically unimportant institutions have to be rescued to prevent panic and mass contagion, obviously the largest and most complex institutions have to be rescued, whether or not they are banks and whether or not such rescues are allowed by law. Additionally, it is silly to suggest that every bank should be made so small that it can be allowed to fail in the midst of a crisis without systemic consequences. Such an atomized financial system would neither function nor be safe. The US had an atomized financial system in 1930-33, and the result was catastrophe.

There is nothing wrong with maintaining “constructive ambiguity” about the size of the safety net, so long as this ambiguity is jettisoned when a system is in general crisis. It is safer and easier to throw the safety net over the system before the first failure than it is to stop a mass run. That is the critical lesson of the catastrophic failures of the Bank of the US in 1931 and Lehman Brothers in 2008. The banking panic of 1931-33 would have been averted had the Fed rescued Bank of United States, and the panic of 2008 would not have occurred had Lehman been rescued and the safety net been extended to Wall Street, as indeed it was only a month later.

The European attitude toward this matter (notwithstanding toothless EU rules against state support) has been more pragmatic that that of the U.S. In Europe, when the recent crisis loomed, there was little consideration given to the “orderly default” option. Indeed, so far only one small bank has been allowed to default in continental Europe. In his book In Fed We Trust, David Wessel quotes an ECB official as saying “We don't let banks fail. We don't even let dry cleaners fail.” True, refreshing, and wise. The Europeans understand this issue.

In the U.S., the problem is complicated by a legal and regulatory regime designed to allow banks to fail, intentional ambiguity about the size of the safety net, laws that are intended to prevent bailouts, a large regulatory distinction between banks and nonbanks, and strong political opposition to bailouts. U.S. bank regulation is designed to deal with small banks; there is little legal or regulatory distinction between Citbank and the third largest bank in Peoria.

This unfortunate legacy resulted in the (unwise and incorrect) conclusion by the Treasury that it lacked the legal authority to rescue Lehman, a nonbank. Indeed, in the public dialogue concerning regulatory reform, there has been no discussion of the need to widen and strengthen the safety net. Consequently, when the next crisis occurs in the US, history will repeat itself.
The only solution to the systemic risk paradox is that countries must realize that:
--They are on the hook for their entire financial systems whether they like it or not.
--The ultimate solvency of the financial system is a contingent liability of the government.
--Transparency and market discipline have never worked.
--The only practical solution is a strong safety net and competent prudential regulation of the entire financial system.

Since governments can’t get off the hook for the financial system, then the system must be much better regulated. This means insisting on less leverage, less reliance on market funding, more numerous and intrusive regulators, and intellectual parity between regulators and the regulated. In the regulatory arms race, the banks have ICBMs and the regulators have muskets. Big banks have armies of PhD's paid millions to build and defend their complex risk management and capital adequacy models; the regulators have underpaid government employees who lack the tools to deconstruct and second-guess those models.The cost of competent prudential regulation is much less than the cost of systemic bailouts. The Federal Reserve System's total annual expenses are less than $5 billion and the UK FSA's are less than $1 billion.

This not an argument that financial institutions should not be allowed to fail. At times of systemic uncertainty and fear, financial institutions may be allowed to fail so long as they do not default on deposits, derivatives and senior debt. There is no reason why a failing institution cannot be seized, cleaned and sold, wiping out stockholders, management and many employees, as was done in the case of Bear Stearns, and should have been done with Lehman. Credit markets can accept such a form of resolution so long as it does not set off a tidal wave of defaults.

Unfortunately, there is little likelihood that the structure and quality of US or European financial regulation will improve. No one wants to admit that the safety net is so big, because it would cost governments money, and because the financial industry would oppose a regime that would regulate the unregulated, force a reduction in leverage, and unleash armies of smart, tough regulators who can't be intimidated or captured. What is more likely is a return to business as usual and another crisis down the road.

The author is a retired vice chairman of Moody's Investors Service.

No comments: