II. THE ROLE OF CENTRAL BANKS
IN FINANCIAL STABILITY:
HISTORICAL REVIEW
AND CRITIQUE
Financial Stability: Lessons Learned
from the Recent Crisis and Implications
for the Federal Reserve
Nellie Liang*
Board of Governors of the Federal Reserve System
This paper addresses some lessons learned about financial stability during the recent financial crisis, with a particular focus on the implications for changes at the Federal Reserve. Changes also are driven by the passage of the Dodd–Frank Wall Street Reform and Consumer Protection Act of 2010 (DFA).1 DFA addresses significant shortcomings and gaps in the regulatory structure of large complex financial institutions, and relies heavily on preemptive policies to reduce the likelihood that the firms fail. However, it also limits the tools available to regulatory agencies to address a crisis once it is underway. This emphasis on preemptive actions may reflect an apparent belief among many in the Congress that the actions taken by regulators in 2008 and 2009 to address the financial crisis as it unfolded, while effectively preventing a financial collapse, may have increased moral hazard problems going forward. DFA is silent on reforming short-term funding markets—a central propagation channel in the crisis—and leaves the task of reform to the regulatory agencies. The lessons learned and DFA have altered the ways that the Federal Reserve, other regulators, and firms conduct their business and assess systemic risks to foster financial stability.
This paper contains two major sections. Section 1 presents four important lessons for the Federal Reserve that came out of the crisis; it includes a brief review of the actions taken by the Fed during the crisis, as well as some insights from the literature on the causes and consequences of financial crises. Section 2 addresses how the Fed is reshaping its organization, policies, and governance in the aftermath of the crisis to meet its expanded mandate for financial stability under DFA.
1. Lessons Learned
The financial crisis exposed serious deficiencies in the existing financial regulatory framework and cast doubt on some widely held perceptions about the relationship between financial stability and other economic policy objectives. It also challenged the Federal Reserve to find new ways to fulfill its traditional central banking functions for financial and monetary stability and offered important lessons that should be heeded as the Fed implements DFA and works to improve the resilience and performance of the financial sector.
To set the stage for lessons and implications, the Federal Reserve has several functions relating to financial stability, including: (1) maintaining the stability of the financial system, if necessary by serving as lender of last resort; (2) conducting monetary policy — currently according to the dual mandate, which treats maximum employment and price stability as co-equal policy objectives; and (3) serving as primary supervisor for bank holding companies and state member banks.
Lesson 1. Pre-crisis regulatory policies were not sufficient to ensure financial stability. Perhaps the most important lesson is that there were notable gaps in the pre-crisis regulatory and supervisory structure. In part, the gaps reflect more rapid development of financial practices than in the regulatory structure. In particular, the regulatory system was built largely around ensuring the safety and soundness of individual depository institutions, often evaluated by capital-to-risk-weighted assets ratios, which tend to be lagging indicators of performance. Regulators and regulated firms alike paid insufficient attention to systemic risk externalities that the firms could impose on other parts of the financial system and real economic activity. Risk externalities, such as fire sales driven by de-leveraging, contagion from common exposures or lack of information, and complexity, have the potential to transmit and amplify distress of an individual firm to other parts of the financial system and disrupt real economic activity (see Bernanke and Gertler, 1989; Allen and Gale, 2005; Brunnermeier and Pedersen, 2009; and Caballero and Krishnamurthy, 2008). Overall, risk management practices were not sufficiently forward- or outwardlooking.
In addition, insufficient attention was paid to the rapidly growing shadow banking sector, which had its roots in regulatory capital arbitrage and greater demand for private alternatives to insured deposits from, for example, money market mutual funds and institutional cash pools. Securitizations, asset-backed commercial paper (including SIVs), and broker-dealers often use short-term debt from liquid markets to purchase long-term illiquid assets (see Pozsar et al., 2010; Gorton and Metrick, 2012; and Covitz, Liang, and Suarez, 2013). But shadow banks are susceptible to runs on their short-term debt, which is not covered by deposit insurance, leaving the financial system highly vulnerable.
Although shadow banks are vulnerable to runs, the existing resolution regime for commercial banks did not apply to these institutions. Bankruptcy was the only formal process available for resolving financial institutions other than insured depositories. Because the bankruptcy courts focus not on systemic risk, but on maximizing proceeds to creditors, the idea of addressing financial claims through the bankruptcy process raised concerns about further destabilizing the financial system during the crisis.
Lesson 2. The pursuit of price and economic stability is not sufficient for financial stability. The extended period of price and economic stability that preceded the financial crisis lulled many into believing that stable macroeconomic conditions were sufficient to foster financial stability. Monetary policy during this period was geared largely toward sustaining the so-called “Great Moderation,” and based on the argument that bubbles are hard to identify and crises hard to predict, it responded to financial factors only to the extent that they affected the outlook for inflation or real activity. Because identifying bubbles ex ante with certainty is virtually impossible, having to clean up after a bubble burst was viewed as preferable to acting too frequently and unduly restricting credit.
However, financial institutions responded to the period of stability by increasing leverage (Adrian and Shin, 2008). And because the financial regulatory structure is procyclical, in part because capital requirements are often based on recent historical experience, it too facilitated the build-up of risks (Kashyap and Stein, 2004). In any event, we learned that highly leveraged institutions or asset values driven up by leverage and maturity mismatches in assets and liabilities can pose substantial risks to financial stability.2 Furthermore, cleaning up after the crisis has been costlier and has taken longer than had been expected.
Lesson 3. The traditional tools of the Fed are not sufficient for the worst crises. The first traditional tool in the Fed’s central banking toolkit is monetary policy. As financial strains increased in short-term funding markets in the fall of 2007, the Fed started to lower short-term interest rates. However, the federal funds rate is subject to a zero lower bound, which was encountered relatively early in the financial crisis. Of course, additional stimulus could be — and was — provided through large-scale asset purchases (LSAPs); the LSAPs seem to have eased pressures in short-term funding markets somewhat, but they likely entailed costs as well. While research suggests that there are benefits to reducing the federal funds rate sooner to combat a crisis (Reifschneider and Williams, 2000), some recent research suggests that conventional monetary policy may be less effective during periods of stress than in more normal times. Hubrich and Tetlow (2011) use a Markov-switching VAR model to identify “stress events,” which the model associates with periods of elevated financial stress, highly accommodative monetary policy, and a weak outlook for economic activity. They found that although monetary policy was eased during each “stress event” since the 1970s, the policy actions seem to have had only small effects on real economic activity.
The second traditional tool is the discount window, which serves as a backstop source of short-term funds to insured depository institutions (DIs). However, insured depositories are typically reluctant to borrow from the discount window, apparently because a liquidity constraint might be taken as a sign of a solvency problem, and indeed the stigma appeared to intensify in the crisis. The Fed’s initial efforts to spur the use of the discount window (through a lowering of the spread between the primary rate and the federal funds rate and an extension of the terms of the lending) apparently did little to ease the stigma. Accordingly, in December 2007, the Fed introduced the Term Auction Facility (TAF), under which predetermined amounts of discount window credit were auctioned against a broad range of collateral. TAF apparently helped to overcome the stigma somewhat, presumably because of both the competitive auction format and the availability of large fixed amounts of credit.
Shadow banks, however, did not have access to the discount window, and these institutions were subject to investor runs. To address the problems of broker-dealers and the dysfunction in the tri-party repo market, the Fed provided discount-window-like lending to non-DIs under the determined “unusual and exigent” circumstances. This was done through the Term Securities Lending Facility, which allowed primary dealers to exchange less-liquid securities for terms of 28 days at an auction-determined fee, and the Primary Dealer Credit Facility, which allowed primary dealers to borrow at the same rate at which DIs could access the discount window. The Fed also provided backstops for money market mutual funds through the Asset-Backed Commercial Paper Money Market Mutual Fund Liquidity Facility, and for issuers of commercial paper through the Commercial Paper Funding Facility.
The Fed’s performance as lender of last resort under “unusual and exigent” circumstances during the crisis was consistent with Bagehot’s dictum to “lend freely at a high rate, on good collateral.” Lending freely on good collateral during a financial crisis can be useful because it can help prevent solvent financial institutions from selling assets at fire-sale prices in order to meet the demands of depositors and other sources of funding; the creation of several new lending facilities under Section 13(3) of the Federal Reserve Act largely followed this dictum. Moreover, the new facilities were priced to be unattractive in normal times, and thus to unwind automatically as market conditions improved. Of course, whether collateral will turn out to be good or bad depends in part on the success of the actions; for example, the collateral is likely to be good if fire sales are prevented, but may not be if they are not. Also, fire-sale effects can last a long time should financial risks spill over to economic activity more broadly.
An interesting consequence of these activities was the emergence of timely, sophisticated valuation of collateral as an important new area of central banking. Indeed, for the Supervisory Capital Assessment Program — the original stress tests — more than 150 Federal Reserve staff worked for many months to value the legacy assets on the banking and trading books of the 19 largest bank holding companies, based on unprecedented detail about the specific assets collected ...