Chapter 14: Executive Summary

Private Lessons for Public Banking: The Case for Conditionality in LOLR Facilities

Viral V. Acharya and David Backus

Background

As we work our way through the current financial crisis, central banks have shifted their attention from managing short-term interest rates to providing liquidity to the financial system. In the US, for example, the Federal Reserve's balance sheet has expanded rapidly, as it offered funds to banks and accepted securities in return: from under a trillion dollars in August 2007 to over two trillion in November 2008, expanding primarily through its lending to banks against illiquid collateral. This "lender of last resort" (LOLR) role is neither new nor unusual, but its massive scale suggests that it is worth some thought to get the details right. We make below what may seem right now to be a perverse argument: Central banks can learn something from the private sector about how to manage its provision of liquidity.

The Issues

Let us start from the beginning. Walter Bagehot codified the nineteenth century's collective wisdom on central bank provision of liquidity in Chapter VII of Lombard Street (1873). In many respects, the same principles guide modern central banks. It involves the following elements: (i) Central banks should hold large reserves; (ii) In times of panic, the central bank should freely advance these reserves to any private bank able to offer "what in ordinary times is reckoned a good security" as collateral; (iii) These advances should be charged a penalty rate to discourage applications from banks that do not need it. (Bagehot seemed concerned primarily with the practical goal of conserving limited reserves); and, (iv) This policy of using reserves to stem panics should be clearly communicated. Otherwise, uncertainty about central bank actions can themselves contribute to the panic.

These guidelines remain insightful but we think they miss an important aspect of financial crises: it is not easy to tell the difference between an illiquid and an insolvent institution. In fact, that is usually what precipitates matters: no one is sure who is solvent. In those circumstances, a central bank can easily find itself lending to an insolvent institution, perhaps creating an unnecessary delay in its timely reorganization and recapitalization.

Consider an undercapitalized and possibly insolvent bank: call it Lehman Brothers if you like. If it can borrow from the central bank, it faces less pressure to raise more capital privately to solve its underlying problem. It has been a shocking revelation to many that the total capital raised (public and private) is remarkably small when compared to the total losses incurred by financial institutions worldwide (including banks, broker-dealers, insurers and GSEs) from 3Q 2007 to date. Excluding 4Q 2008 which features large-scale capital infusions from governments into the financial sector, financial institutions simply did not raise enough capital, in fact, not even enough to cover their losses. What is more, even firms in difficulty such as Lehman Brothers and Citigroup were paying significant cash dividends - that is taking capital out of their balance-sheets — until they failed or were bailed out.

Could the central banks while providing liquidity to these institutions (many unhealthy at that such as the GSEs, Washington Mutual, Wachovia, Lehman Brothers, and so on) ensure that they restructured — by reducing leverage and risk or converting debt to equity — or recapitalized — by issuing preferred or equity capital in markets?

We believe the answer is yes, provided the liquidity facilities created by central banks granted them the rights to deny liquidity conditional on bank health and characteristics. Such rights are indeed available to private providers of liquidity insurance — namely, the banks — when they allow borrowers to pre-arrange such liquidity facilities from them. The private lines of credit (LCs) serve a similar purpose for borrowers as central banks' LOLR facilities do for banks: they represent contracts pre-arranged by firms with banks for banks to give them liquidity when firms need it. Indeed, LCs often constitute the borrower's last line of defense against an economy-wide shortage of credit, as in the current crisis. The tradeoffs involved are also the same — providing liquidity to avoid deadweight costs of liquidation of a sound enterprise but weighing that against the fact that insurance will reduce the discipline on the enterprise to avoid being in such a situation in the first place. How does the structure of private insurance deal with this tradeoff?

Private lines of credit have the borrower pay a the commitment fee and interest rate once the lines are drawn that are both tied to the firm's credit rating, which allows the lender to respond to changes in credit quality. More importantly, they include covenants (cash-flow based, for example) and a "material adverse change" (MAC) clause that give the lender the ability to refuse the loan if the conditions of the borrower have changed. These terms, and their enforcement observed in practice, suggest that lines of credit are private solutions to liquidity issues, not solvency issues. In some respects, central banks' LOLR facilities resemble private lines of credit. Prices aren't tied to credit rating, but central bank lending is secured against collateral, albeit illiquid. What's missing, however, is anything resembling the material adverse change clause. There's nothing, in other words, to keep an undercapitalized bank from using such a facility. This, we view, as a serious limitation in structure of these facilities.

Policy Recommendations

Our main recommendation is thus that just like private lines of credit, central banks' liquidity facilities should be conditional. In particular, central banks should ascertain while providing liquidity to an institution that they are indeed lending to sound institutions.

A straightforward way to achieve this objective is to require in the LOLR facilities that eligible institutions and firms can borrow from the central bank against eligible collateral only if they meet pre-specified requirements, for example, maximum leverage and minimum capital ratios. Such conditionality would incentivize weak banks to recapitalize when their losses mount so as to have access to the LOLR facilities and thereby limit moral hazard. Conversely, in absence of such conditionality, weak banks may access liquidity facilities and simply play the waiting game — a way for management to avoid being diluted by fresh capital issuance and thereby risk being even more insolvent if things do not improve. As the Federal Reserve expands its liquidity operations to a wider set of institutions and firms in the economy, the role for such conditionality in its liquidity facilities seems imperative.

Return to the NYU Stern White Papers Project