What was new prior to the crisis was the extent to which maturity transformation and financial intermediation had migrated outside of commercial banks. The growth of what we call the shadow banking system occurred largely without the types of safeguards—robust prudential regulation, deposit insurance, lender of last resort—that have safeguarded the commercial banking system from the types of widespread panics and runs that are capable of destabilizing the financial system. The systemic risk created by this gap in coverage was not well recognized by regulators or the private sector prior to the global financial crisis. Market participants had little incentive to internalize the negative externality of the run-risk created by their collective choice of finance, and they made erroneous assumptions about the liquidity of asset markets and the capacity and willingness of banks to distribute central bank liquidity to the wider financial system during periods of stress. 

Let me summarize. The creation of private money--that is, safe claims that are useful for transactions purposes--is obviously central to what banks do. But safe claims can be manufactured from risky collateral in different ways, and banks are not the only type of intermediary that engages in this activity. What makes banks unique is that they use a particular combination of financial and institutional arrangements--including capital, deposit insurance, and access to a lender of last resort--as well as substantial investments in bricks and mortar, to create liabilities that are not only safe and money-like, but also relatively stable and thus unlikely to run at the first sign of trouble. This is in contrast to shadow banks, who create money-like claims more cheaply, by relying on an early exit option, and who are therefore more vulnerable to runs and the accompanying fire-sale risk. I have argued that there is a synergy between banks' stable funding model and their investing in assets that have modest fundamental risk but whose prices can fall significantly below fundamental values in a bad state of the world. This synergy helps explain both why deposit-taking banks might have a comparative advantage at making information-intensive loans and, at the same time, why they tend to hold the specific types of securities that they do.

Our framework for {the large-scale asset purchase} aspect of our credit policies relied on preferred habitats of investors and imperfect arbitrage. There was ample evidence that private agents had especially strong preferences for safe and liquid short-term assets in the crisis; in those circumstances, sizable purchases of longer-term assets by the central bank can have an appreciable effect on the cost of capital to households and businesses. The marked adjustments in interest rates in the wake of the announcements of such actions, both in the United States and elsewhere, suggest that market participants also saw them in this light.

In the past, systemic risk had been thought of as involving a single large institution. In the recent cases, we invoked the systemic risk exception with respect to the system as a whole, thereby allowing assistance to flow to institutions of all sizes.

For its part, the Federal Reserve has also taken steps to assist smaller institutions. For example, all banks can borrow funds under the Federal Reserve Term Auction Facility, which operates much the same as the discount window, but offers longer terms. And Regulation D was recently modified to allow community banks to earn interest on excess reserves held in bankers' banks on a pass-through basis.

I expect securitization to continue because this form of financial intermediation developed in response to needs and realities that have not disappeared. I was a commercial banker in the 1980s and I remember well the onset of the practice of balance sheet allocation according to return on assets and ultimately return on equity. Commercial banks were, and remain, caught in a dilemma of wanting to serve their clients by providing loans, but not always being able to justify booking very competitively priced loans on their balance sheets.

[I]t must be recognized that there are many types of banks in the United States with different specializations, geographic concentrations, and comparative advantages. Consequently, the extraordinary stress in the financial system, the downturn in the U.S. and global economies, and the associated reductions in asset values have affected each bank differently.

In the U.S., the market took over. The market has flopped...Everybody is running back to Mother, the commercial banking system.

Financial Headwinds in the Early 1990s... 

[C]apital shortfalls meant that banks had to either raise new capital or restrict their asset growth by cutting back on lending. Because of their weak condition, banks could not raise much new capital, so they chose the latter course. The resulting slowdown in the growth of credit was unprecedented in the post-World War II era (Reifschneider, Stockton, and Wilcox, 1997). Because banks have informational advantages in making certain loans (e.g., Mishkin, 2007a), many bank-dependent borrowers could no longer get access to financing and thus had to cut back on their spending.

Community banks today--to a somewhat greater degree than their larger money-center brethren--most clearly retain the traditional commercial banking approach to financial intermediation.  Community banks finance relatively opaque entities, such as private companies and households.  They raise funds primarily by issuing demand deposits.  They deploy capital by underwriting loans, monitoring borrowers, and retaining some loans in portfolio as long-term investments. 

Systemic events in market-based financial systems are perhaps more likely to involve price fluctuations and abrupt changes in market liquidity than are systemic events in depository-based financial systems. But that is not really bad news because such events can more readily be countered by macroeconomic policy instruments than could old-fashioned crises of depository intermediation. Supplying additional liquidity and reducing borrowing costs can greatly ameliorate the effects of market events on the economy, and those types of macroeconomic interventions will carry less potential for increasing moral hazard than would the discount window lending that was a prominent feature of crisis management when depositories funded more credit.

Market-intermediated finance also requires us to live with less control and less knowledge than we had when banks were dominant. Greater uncertainty about where risks are lodged is the flip side of better dispersion of those risks, especially to less regulated sectors, and of more resilience of the whole system. Gathering additional information about the risk profiles of currently less regulated institutions is unlikely to yield insights that can be acted upon and may create a false sense of comfort among market participants, which could make the system substantially more risky. We need to have confidence in the invisible hand. But confidence does not mean blind faith, a thought that brings me to what we can productively do to reduce systemic risks within the boundaries that I just described.

[T]hat is a key role for banks in a crisis: to obtain funds--through the discount window or from open market operations, if necessary--and to channel them to those needing funds, based on an assessment of their creditworthiness. Banks' access to the discount window and the payments system, as well as their ongoing relationships with customers and their credit-evaluation skills, allow them to play this role. During a crisis, those banks that play critical roles in the payments system are especially important. As a result, these banks are expected to be very resilient. Though banks now have a smaller role in transmitting monetary policy, they still help to transmit policy actions by arbitraging between the federal funds market and other money markets.

[A] decline in uncertainty resulting from a substantial increase in real-time information implies a reduction in what might be called "knowledge float"--the ability to maintain proprietary information and earn a rate of return from that information with no cost. As you know, financial intermediaries historically have been successful not only because they diversified to manage risk but also because they possessed information that others did not have. This asymmetry of information was capitalized at a fairly significant rate. But that advantage now is rapidly dissipating. We are going to real-time systems, not only with transactions but with knowledge as well.

I would argue that banks remain quite special in their susceptibility to runs and in the severe consequences that a large-scale banking panic would involve today. Balancing the need for a banking "safety net" to defuse potential bank runs with the need to create the right incentives for banks in assessing and assuming risk is one of the most difficult challenges we face as central bankers.

I would argue that banks remain quite special in their susceptibility to runs and in the severe consequences that a large-scale banking panic would involve today. Balancing the need for a banking "safety net" to defuse potential bank runs with the need to create the right incentives for banks in assessing and assuming risk is one of the most difficult challenges we face as central bankers.

As I hope my previous comments make clear, regulatory and supervisory policies must recognize the dynamic forces at play in the financial sector. Such policies must promote and exploit the competitive process in order to foster efficient delivery of services while encouraging financial and economic stability. These objectives are not likely to be achieved by regulations that arbitrarily identify and rigidly segment bank and nonbank financial markets. Rather, our goal as policymakers should be to establish rules of the game that provide proper incentives for financial institutions to accurately assess and manage the risks inherent in their business decisions. Likewise, we should foster reporting standards and information flows so that the consumers of financial services and products are as well informed as possible about the risks and returns of the financial services and products they buy. To the maximum extent possible, market forces should determine which bundles of financial services and products are provided by banks and other types of financial service providers.