Skip to Main Content

Last-Resort Lending for the 21st Century

Econ Focus
Fourth Quarter 2014
Federal Reserve Sidebar

Why the continued interest in Henry Thornton and Walter Bagehot so long after their time? They were two of the first to navigate what today's central bankers accept as a fundamental trade-off of crisis policy: the need to limit panics today without encouraging greater risk-taking in the future.

Their broad principles for striking this balance were to supply ample liquidity in crises but in a way that is sufficiently painful to borrowers — lending only to worthy borrowers at high interest rates and against sound collateral — that they'll want to take measures to avoid vulnerability in the future.

In the middle of a crisis, that can be harder to achieve than one might think. Here are some of the issues that central banks face.

Illiquidity vs. Insolvency

Most central bankers would prefer never to bail out insolvent firms. But crises unfold quickly and it can be unclear who is solvent and who is not. So how can central banks distinguish firms experiencing a temporary liquidity shock from those that are fundamentally insolvent?

"I would say that it's very well near impossible to make that distinction," says Charles Goodhart, an economist at the London School of Economics who has written extensively on lender-of-last-resort policy. "Illiquidity is almost always a function of concern about potential insolvency, even if that concern is misguided."

There's a complicating factor: Are there some cases where insolvent firms should, in fact, be saved — perhaps if their failure would hurt many others? Typically, markets minimize spillover risk by charging premiums to borrowers that are riskier. But economists have modeled scenarios in which firms are not forced to bear the costs of the ways in which their actions would affect others. Such models — many of which describe a far more complex financial system than what existed in 19th century England — suggest the possibility of outcomes where risks become contagious, leading to runs or widespread liquidity crises. The extent to which these characterized the 2007-2008 crisis is still an open question; an alternative view is that a more important component of the crisis was markets adjusting to previously unknown risks emanating from the housing market.

Federal Reserve

Averting Financial Crises

Advice from classical economists, as recounted by monetary history researcher Thomas Humphrey

Either way, there is a moral hazard problem to contend with. If central banks routinely prevent systemic losses, firms will choose to become too systemically linked, increasing the likelihood of contagion. That means market failures may be better addressed with regulatory measures than with emergency lending. And for the lending that does take place, it provides a strong argument for making it costly for firms to borrow in a crisis so they'll want to use it as truly a last resort — for example, with penalty rates.

What Constitutes a Penalty Rate?

In principle, penalty rates — often discussed in terms of interest rates — come down to whether the loan from the central bank is cheaper than private alternatives in a crisis. If it is, the lending might encourage excessive risk-taking because investors won't pay the price, so to speak, of financial market turmoil.

Thornton and Bagehot advocated a "high" interest rate but didn't spend much time defining it. Much of Bagehot's case was based on the need to keep the gold standard functioning, and strict usury laws were in place in Thornton's time, notes monetary economist David Laidler, professor emeritus at the University of Western Ontario. But many scholars agree at least in principle that a penalty rate is funding which is costlier than a firm could get in normal times but cheaper than the panic-induced crisis rate (since a central bank offering loans above the latter would find no takers).

But the right penalty rate can be hard to identify in practice. As noted, the essence of a crisis is often that the true values of assets become uncertain after previously unknown risks come to light. Some research suggests that this problem can be exacerbated by so-called "fire sales" that artificially depress asset prices as firms struggle to raise funding.

But erring on the side of high penalty rates would have costs. It would deplete the borrower's capital further, which might worsen the panic. Another concern is that markets know that only the weakest banks will be desperate enough to pay penalty rates. The classical-era Bank of England dealt with this potential problem by providing loans through institutions known as discount houses that kept the borrowers essentially anonymous. In 2007 and 2008, both the Fed and the Bank of England argued that a "stigma" left their traditional discount window facilities underutilized in the early days of the crisis. In the United States, the Fed launched an alternative facility in which firms bid for funds. The winning bid often landed at sub-penalty rates.

A final challenge is that penalty rates simply may not provide the amount of funds that policymakers wish to funnel to markets. For example, one of the Fed's recent crisis programs gave special loans at sub-penalty rates to banks willing to purchase troubled asset-backed commercial paper from money market mutual funds. Fed policymakers argued at the time that charging a penalty rate would not provide the funds necessary to support the economic activity dependent on those markets.

The program seemed to help calm markets, but to some observers, this type of lending is simply a handout to certain sectors, not a lender-of-last-resort function. Richmond Fed President Jeffrey Lacker has argued that there was no unmet funding need in some markets that were supported — only prices that investors didn't want to pay due to the risky environment.

Money vs. Credit

At the broadest level, no one disagrees that the fundamental goal of last-resort lending is to prevent financial market problems from causing recession and job loss. But among modern observers, there are two views on how the central bank should go about it: Should the central bank expand the supply of money to meet the panic-induced demand for safe assets? Or should it extend credit directly to firms to stop failures and panics at the source?

Laidler describes this "money vs. credit" debate as "a swamp from which few return once they enter it." In other words, the division between the two has not been entirely clean in practice. The 19th century Bank of England, for example, conducted monetary expansion via lending to firms. Today, the Fed conducts monetary policy largely through open market operations that inject liquidity broadly. More recently, the Fed mixed the money and credit functions with "quantitative easing" that expanded its balance sheet — an act of monetary easing — but by purchasing mortgage-backed securities.

Moreover, Goodhart expresses doubt that there is sufficient time in a crisis for a central bank to provide money and for that expansion to spread to illiquid but solvent institutions. "People will be thinking, ‘Who is next in line to fail?' and run from them. You've got to stop contagion very, very quickly." Once again, this interpretation depends on the view that market failures make it impossible for firms to adequately protect themselves from contagion.

Another view turns the complexity of today's financial markets on its head: Firms have more alternatives to central bank funding than ever before, and will find ways of directing money to sound borrowers if only the perverse incentives provided by the central bank's backstop would get out of the way. A 1988 article by Marvin Goodfriend, a former research director of the Richmond Fed who is currently at Carnegie Mellon University, and Robert King of Boston University argued for doing away altogether with the Fed's ability to lend directly to firms. That would leave broad open market operations as its only means of pumping liquidity into the economy.

More recently, Goodfriend has argued against a credit role for central banks on the ground that they face an incentive to err on the side of lending perhaps too broadly. That wasn't the case for the 19th century Bank of England; it was held by private shareholders, so the profit motive created a natural inclination to lend conservatively. That may be one reason Bagehot felt the need to encourage liberal lending.

Modern central banks, in contrast, lend with public funds. They also face intense political pressure to protect the economy at all costs — whereas central banks in classical times faced no macroeconomic objectives. On balance, modern central banks are naturally likely to overlook the longer-term moral hazard costs and lend too liberally, according to Goodfriend. The Fed has expanded the scope of its emergency lending since the 1970s, which some observers argue is one reason firms have made themselves so vulnerable to systemic events in the first place.

These issues are far from resolved. For better or for worse, central banks largely chose the credit function in 2007 and 2008. Doing so creates significant long-term challenges, but as Laidler puts it, central banks facing crisis have tended "to swallow hard and get on with it."

Where Do We Go From Here?

Without a clearly defined crisis policy in advance, "by history and tradition, the central bank has always leaned toward liquidity provision," Chairman Bernanke noted to his fellow policymakers in 2009. This leaves regulatory reform to clean up the moral hazard repercussions after the crisis has passed.

That is just what Congress attempted to do with the regulatory provisions of the 2010 Dodd-Frank Act. The Act also sought to restrict the Fed's emergency lending powers. Now the Fed cannot bail out one particular firm; its emergency credit programs must have broad-based eligibility.

Dodd-Frank also required the Fed to get more specific about its crisis procedures. According to an August 2014 letter from a bipartisan group of 15 members of Congress to Fed Chair Janet Yellen, "By directing the Board to establish a clear lender-of-last-resort policy, where both policymakers and the marketplace know the rules of the game beforehand, Congress sought to ensure that banks fully internalized both the risks and rewards of their decisions." The letter argued that the Board's first attempt at such a policy did not achieve that end. In response, they requested further crisis rules that sound similar to the methods proposed by the classicals to avoid moral hazard.

Among their requests: for the Fed to establish a clear timeline for a financial institution's reliance on emergency lending, with concrete limits on the duration of each facility; preset guidelines for winding down lending facilities to ensure they are truly temporary; a broader definition of "insolvent"; a method for ensuring that lending is intended to help financial markets broadly instead of being designed for one specific institution; and a commitment to lend only at penalty rates. (As this issue went to press in the spring of 2015, legislation had been introduced dealing with some of these concerns).

Plenty of observers have offered broad principles on crisis lending. But no one has definitively figured out how to implement them in practice. To some, that is an argument for central banks erring on the conservative side, lending to as few parties as possible to enhance market discipline. In practice, as Bernanke has said, central banks have tended to err liberally to prevent financial and real losses. The 2007-2008 financial crisis provides the largest modern case study of crisis lending, warts and all, for the pursuit of clearer answers.


Goodhart, C.A.E. "Myths about the Lender of Last Resort." International Finance, November 1999, vol. 2, no. 3, pp. 339-360.

Goodfriend, Marvin. "The Elusive Promise of Independent Central Banking." Monetary and Economic Studies, November 2012, vol. 30, pp. 39-54.

Goodfriend, Marvin, and Robert G. King. "Financial Deregulation, Monetary Policy, and Central Banking." Federal Reserve Bank of Richmond Economic Review, May/June 1988, vol. 74, no. 3, pp. 3-22.

Lacker, Jeffrey M. "Economics After the Crisis: Models, Markets, and Implications for Policy." Speech at the Center for Advanced Study in Economic Efficiency, Arizona State University, Feb. 21, 2014.

Laidler, David. "Two Views of the Lender of Last Resort: Thornton and Bagehot." UWO Department of Economics Working Papers No. 20029, September 2002.

Madigan, Brian F. "Bagehot's Dictum in Practice: Formulating and Implementing Policies to Combat the Financial Crisis." Speech at the Federal Reserve Bank of Kansas City's Annual Economic Symposium, Jackson Hole, Wyo., Aug. 21, 2009.

Subscribe to Econ Focus

Receive an email notification when Econ Focus is posted online.

Subscribe to Econ Focus

By submitting this form you agree to the Bank's Terms & Conditions and Privacy Notice.

Phone Icon Contact Us

David A. Price (804) 697-8018