Center For Financial Innovation and Stability
Systemically Important Failure versus Financial System Failure
Notes from the Vault
Larry D. Wall
Earlier this month the Financial Stability Oversight Council (FSOC) determined that financial distress at Prudential Financial could pose a threat to U.S. financial stability. This designation by the FSOC, along with its prior designation of American International Group (AIG) and General Electric Capital Corporation, is very important if you think that the recent crisis was a result of individual large financial firms taking excessive risk because of inadequate prudential supervision. These firms will now be subject to Federal Reserve supervision as required by the Dodd-Frank Act (DFA). Moreover, the DFA requires the Federal Reserve to develop enhanced microprudential requirements for large banking organizations and designated nonbank financial firms, including higher capital requirements and regular stress testing.
However, another way of viewing what happened is that it was not the collapse of a few very large financial firms that caused the crisis. Rather, the problem was the widespread exposure of financial firms to the weak credit standards that were pervasive throughout the residential real estate finance market. In this view, the collapse of a few large firms created so much damage because it intensified investor concern about the solvency of almost all of the other large financial firms. As I have argued in the past, the economic consequences of the failure of a few major firms would have been bad but likely not nearly as adverse as what occurred if the rest of the financial system had been clearly solvent.
This alternative view suggests that enhanced microprudential supervision of large bank and nonbank financial firms is only one of the steps necessary to strengthen the financial system. If we want to reduce the risk of future crises substantially, supervisors must also address the potential for common exposures to a large financial market to threaten the solvency of the financial system. As I have previously explained, this requires supervisors to make a systematic effort to identify and address developments in major financial markets that have the potential for causing crises.
A macroprudential approach focused on financial markets
The first step in preventing a buildup of risk from generating potential crisis-causing losses is to identify the buildup in risk. The policy problem here is not the often heard excuse that claims "no one could have seen it coming." As Washington Post columnist Barry Ritholtz noted, that claim is easily refuted in the case of the housing collapse by pointing to many people who made such a prediction. The real problem for policymakers is that at any given time there are far more markets that someone claims are in a "bubble" than will actually suffer a large price collapse. We might as well shut down the financial system if policymakers are going to try to rein in every market that someone is claiming to be in a bubble.
Fortunately, there is a realistic macroprudential approach to reducing risk in important financial markets that does not depend on bubble identification for its success. Such a policy would instead seek first to identify systemically important markets—that is, those markets sufficiently large or otherwise important such that they could cause widespread illiquidity and insolvency in the financial system. Next, such a policy would seek to understand how those markets work all the way from the initial decision to take the risks through to the final holders of the risk. Finally, the supervisors would move to correct any weaknesses that were identified in the market reviews. The workings of such a policy can be illustrated using the residential real estate finance market prior to 2007 as an example.
Identifying markets that could cause a future crisis
The housing finance market is a huge market by any standard, with the historical Federal Reserve Flow of Funds (see the historical data) showing total one- to four-family lending exceeding $10 trillion in the second quarter of 2006. On this basis alone, the housing market would have merited periodic reviews (say, every three to five years) to make sure that the macroprudential authorities were reasonably up to date on market developments.
In addition to mere size, there were several market characteristics associated with greater risk. Large markets that exhibit one or more of these characteristics would be good candidates for more frequent and/or more intensive reviews.
One characteristic consistent with more monitoring is rapid credit growth. In the case of the housing market, the Federal Reserve's Flow of Funds historical data show a more than doubling of one- to four-family residential mortgage debt from the beginning of 2000 to the second quarter of 2006.
A second characteristic I have previously discussed is the rapid growth of new financial instruments. The risks of existing instruments used in the normal way by financial firms and their long-standing clients are likely to be well understood by the firms, the clients, and the supervisors. However, new instruments or old ones used in new ways may bring new risks, including legal risks that may not be fully exposed until there is a downturn. If this criterion had been applied to housing finance, it would have suggested the need for more careful monitoring. Although many of the high-risk mortgage loan products existed prior to the crisis, such as subprime and no-documentation lending, as Chris Mayer and Karen Pence (2008) demonstrate, there was rapid growth over the 1998 to 2005 period—especially in securitized subprime mortgages. The rapid growth of these newer and potentially more risky products should be a potential cause for concern.
A third concern is increasing prices for the assets used as collateral to secure these loans. If the asset market is in a "bubble," loans that depend upon the value of these assets are more likely to default with potentially large losses to lenders. Jeffrey P. Cohen, Cletus C. Coughlin, and David A. Lopez document consistent increases in national home price indices from the early 1990s to 2006, with several common indices showing a doubling of prices between the first quarter of 1998 and the second quarter of 2006.
Thus, the housing industry not only was large enough to merit periodic review, but it had three markers of potentially increasing risk that would merit more frequent and more intensive reviews. Of course, rapid growth, the expanded use of potentially risky instruments, and increasing collateral values are not necessarily signs of a market headed toward a crisis. What action, if any, the prudential supervisors take should instead depend upon the findings of their end-to-end review of the market.
End-to-end market reviews
A thorough end-to-end market review would start with the initial underwriting process and trace the transmission of risks from the initial lender through to the ultimate holder of the risk. In the case of a securitized product, this would include understanding how the process of bundling worked in practice, how the securities were rated, and the chain connecting the issuer of the securities to their ultimate holder. Finally, if the ultimate holders of the securities claimed to be hedging their exposure, it would include a review of the relevant hedging markets. Had such a review been done, many anecdotes about high-risk lending practices would have been shown not to be the isolated cases supposed by some policymakers but rather to have become common practice in large parts of the housing finance industry. Some examples of the many higher-risk practices that could have been revealed by a careful review include:
- A significant fraction of the no-document loans appeared to be living up to (or down to) their nicknames of "liars' loans" where the loan application contains material misstatements (Wei Jiang, Ashlyn Aiko Nelson, and Edward Vytlacil).
- Investors were buying mortgage-backed securities (MBS) that they knew were otherwise high risk based on their expectations for continued appreciation of real estate prices (Kristopher Gerardi, Andreas Lehnert, Shane M. Sherlund, and Paul Willen).
- The ratings agencies were decreasing their standards for private-label mortgages later in the housing cycle (Adam B. Ashcraft, Paul Goldsmith-Pinkham, and James Vickery).
- Foreign and domestic SIFIs were creating thinly capitalized MBS conduits (Viral V. Acharya, Philipp Schnabl, and Gustavo Suarez) and switching from loans to highly rated MBS (Viral V. Acharya and Philipp Schnabl) as ways of arbitraging accounting and regulatory capital rules.
To be fair, many of these practices were known by some microprudential supervisors from the examination of individual firms. However, such institution-by-institution analysis will not necessarily indicate when a variety of high-risk lending practices have become market-wide norms. This is especially the case if, as was true in housing finance, some of the important players in the market are not subject to routine prudential examination. In order to understand market practices, it is necessary systematically to go from one end of the market to the other end.
Addressing potential causes of a future crisis
If a buildup of risks is observed, the next step is taking actions to prevent or at least mitigate a future crisis. The first step would be to use existing prudential authority to require regulated firms to adhere to sound credit practices. For example, if supervisors determine that no-document loans are indeed liars' loans, the supervisors could restrict their regulatees' ability to make no-documentation loans.
A second possible step by prudential supervisors arises if they observe SIFIs making loans whose performance appears to depend upon rising collateral values. In this case, the supervisors could design a stress test scenario that measures large banking organizations and FSOC designated nonbank financial firms' dependence on increasing collateral values. The results of these stress tests could inform a variety of supervisory policies, including whether the Federal Reserve accepts the firms' capital plans.
The third possible step, addressing the buildup of risks outside the prudential supervisory framework, may be more difficult. One possibility would be for FSOC to designate major players in a market as systemically important and therefore subject to prudential supervision and stress testing. A second possibility would be to address the potential risks through those aspects of the market that do touch the regulated financial system. For instance, hedge funds are not subject to prudential regulation. However, hedge funds depend upon regulated firms as sources of various types of credit, including margin credit. Thus, as Federal Reserve Chairman Ben Bernanke described, the supervisors could influence the systemic risk being generated by hedge funds through the supervisors' control over the firms they do supervise.
A fourth possible response would be to use tools intended for other purposes to support financial stability. For example, the Federal Reserve had some regulatory power over housing finance as a part of its ability to write consumer protection regulations. Much of the bank supervisors' consumer protection power was shifted outside the prudential supervisors and given to the Consumer Financial Protection Bureau (CFPB) by the DFA. However, the DFA also gives the FSOC the authority to recommend new or heightened standards if the FSOC determines that an activity poses significant risk to U.S. financial firms or markets.
Finally, the FSOC is also required to make an annual report to Congress in which it identifies emerging threats and includes appropriate recommendations. If an end-to-end market review reveals a potential threat that cannot be adequately addressed with existing regulatory tools, the FSOC's report and the testimony of its members could be used to encourage appropriate congressional action.
Larry D. Wall is the executive director of the Center for Financial Innovation and Stability at the Atlanta Fed. The author thanks Kris Gerardi and Paula Tkac for helpful comments on the paper. The views expressed here are the author's and not necessarily those of the Federal Reserve Bank of Atlanta or the Federal Reserve System. If you wish to comment on this post, please e-mail email@example.com.