Fed emergency lending

Earlier this week, the Federal Reserve’s Board of Governors approved a rule implementing restrictions on its emergency lending powers that were mandated by Congress in the 2010 Dodd-Frank Act. On the whole, the rule is a sensible compromise which clarifies the procedures for Fed lending in a panic while responding to critics’ concerns. Taking into account the many changes made in recent years to increase the resilience of our financial system, we are in a much better place today than before the crisis. Going forward, however, we should be wary of any further changes that might have the effect of deterring financial firms from borrowing from the Fed during a financial panic. Unwillingness to borrow may not sound like a problem, but in fact it could greatly reduce our ability to control panics and to prevent the resulting damage to the economy.

In a financial panic, providers of short-term funding to financial institutions refuse to renew their lending, out of fear that an institution might fail. In the days before deposit insurance, this pullback usually took the form of runs on banks by ordinary depositors. In the 2007-2009 crisis, depositors did not often line up at teller windows, because they knew their accounts were federally insured. But broad-based runs still occurred, albeit electronically, as spooked investors were unwilling to lend short-term, even overnight, to many banks, securities dealers, and other financial firms. When banks or other financial firms cannot obtain funding, they may fail or come close to failing, creating further fear and panic. At a minimum, they will try to raise cash by dumping assets on the market (a “fire sale”), which forces asset prices well below what they would be under more normal conditions, and they stop extending credit to households and businesses, which can bring the economy to a halt.

The most important tool that central banks (like the Fed) have for fighting financial panics is their ability to serve as a lender of last resort: The basic approach was set out by the British economist, Walter Bagehot, in his 1873 book Lombard Street. Bagehot’s advice to central banks can be summarized by what has become known as Bagehot’s dictum: In a panic, lend freely at a penalty interest rate to solvent borrowers on good collateral. Bagehot’s dictum advises central banks to fight panics by standing ready to lend cash to solvent financial institutions, taking as collateral their loans and other assets. By doing so, the lender of last resort can help calm the fears of depositors and other providers of short-term funding, avoid disruptive fire sales of assets, and allow financial institutions to continue lending to the private sector. As the panic eases and private funding returns, financial institutions are able to repay the central bank’s loans with interest. 

Bagehot’s dictum underlay many of the Fed’s programs in its battle against the global financial crisis of 2007-2009. Our system had evolved radically since Bagehot’s time, of course, with a substantial part of credit flows outside the banking system, in nonbank financial institutions and through various securities markets. Accordingly, to follow Bagehot’s principles and to counteract a systemwide funding squeeze, the Fed lent to a wide range of financial institutions and provided liquidity to a number of key markets, such as the commercial paper market. Because the Fed’s ordinary powers permit lending to banks only—a relic of the fact that banks dominated the financial system at the time the Fed was founded—its lending during the recent crisis had to be done under emergency lending authorities, given by Congress in the Great Depression and used for the first time since the 1930s. (It’s perhaps noteworthy that other major central banks already had the flexibility to lend beyond the banking system, and thus, unlike the Fed, did not have to invoke emergency authorities to take similar actions.)

The recent reforms of the Fed’s emergency lending authorities seek to achieve several worthwhile goals: to increase the clarity and consistency of the Fed’s lending policies; to reduce “moral hazard” (the tendency of financial firms to take excessive risk because they expect to be rescued if they get into trouble); and to improve the Fed’s transparency and accountability. Each of these goals has been advanced. For the reforms to be judged a success, however, they must also allow the Fed to continue to serve as an effective lender of last resort. For the most part, I believe they do. Crucially, the Fed retains the authority to lend freely in a panic. In “unusual and exigent” circumstances and with the agreement of at least five Fed Board members and the approval of the Treasury secretary, the Fed may create broad-based lending facilities to assist a market or sector of the financial system that faces a liquidity crisis. (“Broad-based” is defined as meaning at least five eligible borrowers, and not aimed at assisting a specific firm to avoid bankruptcy.) The final rule also provides reasonable and workable definitions of other terms used by Bagehot: “solvent borrowers,” “good collateral,” and “penalty interest rate.” Importantly, the penalty interest rate is defined to be a rate above what would be charged for the particular type of credit in normal times. Charging a rate above the (typically very high) rate available to borrowers in the crisis (as some suggested) would be self-defeating, since the point of central bank lending programs is to help restore normal liquidity conditions.

My biggest concern about the collective impact of the reforms is related to what economists call the stigma of borrowing from the central bank. For lender-of-last resort policies to work, financial institutions have to be willing to avail themselves of the central bank’s loans. If they fear that by doing so that they will be identified by the marketplace as weak, and thus subject to even more pressure from creditors and counterparties, then they will see borrowing from the Fed as counterproductive and will stay away. This is the stigma problem, and it affects everyone, not just the potential borrowers. If financial institutions won’t borrow, then the central bank won’t be able to inject the liquidity necessary to stop the panic, at least not until conditions are extreme. Deprived of access to funding, financial firms will instead hoard cash, dump assets, cut credit, and call in loans, with bad effects on the whole economy.

We faced a serious stigma problem during the recent crisis, and, collectively, the reforms to the Fed’s lending authorities have probably made the problem worse. An example is the effect of new reporting requirements. Dodd-Frank requires that the identities of all borrowers (including non-emergency borrowers through the discount window) be disclosed within two years, or within one year after the termination of a lending program, whichever is earlier. In addition, the rule approved this week confirms that the Fed must provide detailed information to Congress about broad-based lending programs, including the names of borrowers, within seven days—although, very importantly, the names of borrowers can be kept confidential at the request of the Fed chairman. These provisions serve the important purposes of advancing transparency, accountability, and democratic legitimacy, and I am not advocating that they be changed. But we should be aware that, by increasing the risk of early disclosure of borrowers’ identities, these requirements will probably reduce the willingness of firms to borrow from the Fed in a panic and thus potentially impair the effectiveness of the government’s crisis response. 

I don’t see an easy remedy for this problem. As is often the case, policymakers must trade off competing goals. However, in contemplating possible future changes to the Fed’s authorities, we should be very careful to avoid anything that might worsen further the stigma problem and erode the ability of the Fed to serve as an effective lender of last resort.

Comments are now closed for this post.