Economic Brief

The expansion of the federal financial safety net has increased the incentives for financial firms to take on more risk than they would have otherwise. Yet current regulatory reform proposals do not address this root cause of financial instability. Sharply curtailing the financial safety net is a necessary step to achieve enhanced market discipline.

The most recent expansion of the financial safety net that protects debtholders and depositors of financial institutions from losses began on March 15, 2008, with the bailout of Bear Stearns' creditors. The New York Fed assumed the risk of loss for $30 billion (later reduced to $29 billion) of assets held in the portfolio of the investment bank as enticement to JPMorgan Chase to acquire it. Later, the Fed opened the discount window to banks and financial institutions that did not formerly have access to Fed credit.

The rationale for this and subsequent extensions of the safety net was the minimization of "systemic risk" – namely, a reduction in the threat of a cascading series of defaults brought about by wholesale withdrawal of investors from money markets and depositors from banks. Yet economists today recognize that financial safety nets can create a "moral hazard"– that is, an increased incentive to take risks.¹ Given the twin goals of financial stability and mitigation of moral hazard, what financial regulatory regime should emerge as a successor to the current one?

Any proposal must address the consensus that financial institutions took on excessive risk in the period from 2003 to the summer of 2007. These institutions did so through the use of leverage that involved borrowing short-term and low-cost funds to invest in long-term, illiquid, and risky assets. Any new financial regime must limit this sort of risk-taking. But should that limitation come from increased oversight by government regulators or should it come from the enhanced market discipline that would follow from sharply curtailing the financial safety net? One of the authors of this Economic Brief, Robert Hetzel, argues for the latter alternative in the spring 2009 issue of the Federal Reserve Bank of Richmond Economic Quarterly.²

Myths of Financial Market Fragility

A strong case can be made that the financial system would not be inherently fragile in the absence of an extensive financial safety net. History offers evidence that financial markets become fragile mainly because the financial safety net exacerbates risk-taking.

There is ample historical analysis of bank runs before the establishment of deposit insurance in 1934 and the subsequent expansion of the financial safety net. In those studies, bank runs were found to have originated with banks that were actually insolvent. Before deposit insurance, market discipline was effective in closing banks promptly enough to avoid significant losses to depositors. Moreover, the failures that did occur resulted not from "panics" but rather from well-founded withdrawals from banks whose assets suffered declines in value because of aggregate disturbances. An example of such a disturbance was the failure in the 1840s of Indiana banks that held the bonds used to finance the canals rendered unprofitable by the advent of the railroad.³

From the end of the Civil War to the end of World War I, bank failures were relatively few in number and resulted only in small losses because the fear of large losses by both shareholders and depositors were motivated by significant market discipline. This resulted in high capital ratios and prompt closure of troubled banks. Even in the 1920s, when bank failures became more common, runs were uncommon and, when they did occur, funds were re-deposited in other banks.⁴

The waves of Depression-era bank failures before deposit insurance reflected fundamental concerns about banks' solvency as opposed to depositor panic fueled by incorrect information about the actual health of banks.⁵ In an investigation of all Fed member bank failures, apart from January and February 1933, economists Charles Calomiris of Columbia University and Joseph Mason of Louisiana State University found "no evidence that bank failures were induced by a national banking panic."⁶

The Adverse Incentives of an Expanding Safety Net

In theory, regulators could draw a clear line demarcating the boundaries of the financial safety net. Regulators could limit risk-taking of banking institutions and require high capital ratios. Yet, you could also expect that creditors of uninsured institutions who have their own money at risk would also make similar demands. This latter form of pressure can be described as "market regulation of risk-taking."

A tension arises when regulators cannot draw a credible line separating the insured from the uninsured. Institutions on the uninsured side have an incentive to find ways to retain the cheap funds guaranteed by the perception that they are on the insured side while acquiring the risky asset portfolios with high returns of institutions on the uninsured side. The implicit subsidy to a financial institution from the protection afforded by the financial safety net rises with the riskiness of the institution's asset portfolio. Thus, the current financial safety net provides an incentive to banks to acquire risky assets offering a high rate of return without having to increase their capital buffer commensurately.

Government regulation of risk-taking has not sufficiently substituted for the market regulation that would occur if bank creditors bore the full effects of risky behavior by those banks. At least since the 1980s and the bailout of the savings and loan institutions in the United States, the expanding financial safety net has undercut the market regulation of risk-taking. The protection provided by deposit insurance, by the Fed's discretionary discount window lending, and the federal government's assumption that some institutions are simply "too big to fail," have effectively allowed banks to have access to funds that do not become more costly as the bank's asset portfolio becomes riskier.

The prevailing assumption in policymaking circles has been that government does not need an explicit stance with respect to bank bailouts or to credibly commit to a specific course of action. A term that has been used to describe current policy is "constructive ambiguity." Although this characterization in principle acknowledges that regulators have the discretion to not bail out all bank creditors, the prevailing practice of regulators – namely, their past actions to prevent uninsured depositors and debtholders from incurring losses in the event of a bank or thrift failure – limits the incentives for market participants to monitor risk-taking by creditors.

A Proposal to Roll Back the Financial Safety Net

Rolling back the financial safety net depends upon the ability of government to commit credibly to allowing creditors to take losses and thus encourage market-based regulation of risk. That requires taking the bailout option out of the hands of regulators.

The feasibility of any such proposal requires thinking about what the financial system might look like with a severely limited safety net. We can make some plausible assumptions. The large amount of investment in government and prime money market mutual funds holding short- term government securities and prime commercial paper is evidence of the extensive demand by investors for debt instruments that are both liquid and safe. In the absence of the safety net, these investors would constitute a huge market for financial institutions marketing themselves as safe because of high capital ratios and a diversified asset portfolio of high grade loans and securities.

In other words, the market could create a parallel narrow banking system. These institutions would constitute a core of run-proof institutions into which, in the event of a financial panic, creditors would deposit the funds they withdrew from the risky institutions. Yes, depositors at the safe banks would earn a low rate of return, but they, not the taxpayer, would then be the ones paying for financial stability.

To make such a system viable, changes to the current regulatory regime would need to be made. First the government would commit to not bail out the creditors of financial institutions, especially those of large banks. Instead, if a bank experiences a run, the chartering regulator would put it into conservatorship. Under conservatorship, regulators would assume a majority of the seats on the bank's board of directors. The directors would then decide whether to sell, liquidate, break up, or rehabilitate the bank. By law, this conservatorship should eliminate the value of equity and impose an immediate loss on all holders of debt and holders of uninsured deposits, while leaving the remainder of these instruments free for transacting.

Second, a mechanism should be put in place that encourages banks and creditors to monitor themselves and each other.⁷ The federal government could divide banks into groups – for instance, the ten largest could be put into one group, the next ten largest in another group, and so on. The individual banks would pay deposit premia into a fund set up for that group in particular. The banks would also be subject to an assessment to replenish the fund if a bank in their group required

recapitalization as a result of a bank run, even after the mandated "haircuts" for creditors. Each group would have an advisory board that would make recommendations to the Federal Deposit Insurance Corporation (FDIC) for its group about regulating risk, setting the level of the insurance premia, and designing a risk-based insurance levy. The FDIC would set individual group capital standards and other regulations to limit risk-taking.

This system would create an incentive for banks in the group to lobby the FDIC to prevent excessive risk-taking by the other banks in their group. As a check, the public would see the cost of the subordinated debt of each group relative to that of the others. At the same time, the large number of banks in the group would discourage collusion. In the event of a run on a solvent bank, the other banks in the group would possess the information needed to lend to the threatened bank to limit the run just as banks did in the pre-Fed era. A demonstrated willingness of banks to support each other would inspire depositor confidence, too.

Finally, an essential step to establishing a more market-based regulatory arrangement is terminating the legal authority of the Fed to make discount window loans. Instead of parceling out credit to specific firms or sectors, the Fed would resort to its traditional tools of providing the entire market with liquidity by undertaking purchases of securities through its open market operations.⁸ It could also use its new ability to pay interest on bank reserves to maintain its funds rate target.

Conclusion

Government regulation of risk-taking has proven not to be a suitable substitute for a market-based system of regulation. Yet many reform proposals currently debated suggest more government regulation is necessary, not less.

Any proposal to fix the current regulatory regime must first cope with the idea that capital markets may not be as inherently fragile as some think. Two related questions must also be asked. First, in the absence of the risk- taking induced by a generous safety net, would market discipline produce contracts and capital levels sufficient to protect all but insolvent banks from runs? And, second, could regulators place the actually insolvent banks into conservatorship (with mandatory losses imposed on debtors and large depositors) without destabilizing the remainder of the financial system? The historical record leads us to answer those questions in the affirmative.

Robert L. Hetzel is a senior economist and policy advisor in the Research department of the Federal Reserve Bank of Richmond. Stephen Slivinski is senior editor of the Bank's quarterly magazine, Region Focus.