To: Honorable Christopher Dodd and Honorable Richard Shelby
From: Hal S. Scott, Director
Date: May 4, 2010

Re: Systemically Important Institutions

As the Senate begins serious floor debate over financial reform, one major issue that it faces is whether to identify “systemically important” institutions.  The Committee on Capital Markets Regulation (CCMR), in its April 26 letter to Congress, entitled A Blueprint for Compromise, suggested that this not be done.  Instead, the Federal Reserve should supervise financial institutions over a certain asset threshold, whether or not banks.[1]  This would ensure that systemically important institutions, mainly those whose interconnectedness pose risk for the financial system, are appropriately regulated without the need to label them as such.

A. The need to regulate systemically important institutions without increasing moral hazard and distorting competition.

Under the current version of the Restoring American Financial Stability Act of 2010 (Senate bill), the Federal Reserve is to regulate all banks with $50 billion or more in assets as well as any non-bank financial institutions that the Financial Stability Oversight Council (FSOC) determines are systemically important.[2]  The CCMR opposes labeling financial institutions as systemically important because this increases moral hazard.  Once labeled “systemically important,” it is more likely that an institution will be bailed out and that creditors will, therefore, not adequately police their risks.  In addition, these systemically important institutions will enjoy a cheaper cost of funds than their non-systemically important competitors, distorting competition and the allocation of funds in the capital markets.  While institutions designated “systemically important” will also be more heavily regulated, this is only likely to diminish, not eliminate, their advantage.  Furthermore, identifying systemically important institutions will be extremely difficult.

On the other hand, it is clear that institutions whose failure can put the financial system and, ultimately, the economy, at risk, must be adequately regulated to minimize the possibility of costly taxpayer bailouts.  Lack of regulation is not the answer.

B. Absent regulatory consolidation, legislation should give the Federal Reserve jurisdiction over non-banks based on asset thresholds.

In its May 2009 report, entitled The Global Financial Crisis: A Plan for Regulatory Reform, the CCMR suggested that all financial institutions could be regulated by one agency, which the CCMR dubbed the U.S. Financial Services Authority (USFSA).[3]  This agency could be modeled on similar agencies in the United Kingdom and Japan.  Since one agency would regulate all financial institutions, there would be no need to identify particular institutions as systemically important.  Of course, important or large institutions would be treated differently by the USFSA, but this would not necessitate labeling.  Just as importantly, this approach would address the more general problem of our fragmented regulatory structure.  Regrettably, the CCMR’s recommendation to create a consolidated financial regulator has been rejected, and even Senator Dodd’s own initial proposal to consolidate the banking agencies has been withdrawn.

Now, the best alternative is to extend the asset threshold test approach for determining Federal Reserve jurisdiction over banks to non-bank institutions.  With respect to banks, it is clear that not all 36 banks with $50 billion or more in assets[4] are systemically important. We assume this cutoff was chosen to make sure the Federal Reserve had its hands on the pulse of the banking system.  It has the added advantage, however, of avoiding the need to label particular banks as systemically important.  The same approach could be applied to non-banks.  The FSOC, upon the recommendation of the Secretary of the Treasury, would establish appropriate asset thresholds for various types of non-bank financial institutions.  As with banks, the asset thresholds should be set to include all systemically important institutions as well as some institutions that are not systemically important.  Institutions above the threshold would be supervised by the Federal Reserve.

By way of illustration, a $50 billion cutoff for life and health insurance companies would pick up 23 companies and 77% of sector assets, while a $20 billion cutoff for property and casualty providers would cover 15 companies and 59% of sector assets. All such insurance companies operate in each of the 51 U.S. jurisdictions.[5]  For the hedge fund industry, a $12 billion threshold would include the 20 or so largest advisors.[6]  It might well be that the FSOC concludes that some financial industries (e.g., private equity) include no firms that are systemically important.  In this case, it would be unnecessary for the Federal Reserve to supervise firms in those industries.

Some have suggested that a significant disadvantage of using asset thresholds to

determine which insurance companies, hedge funds, and other non-banks would be subject to Federal Reserve supervision, is that there could be a systemically important firm below the relevant asset threshold.  But while this is a possibility, the Senate bill already uses an asset threshold to determine which banks are subject to Federal Reserve supervision.  It is unclear why the prospect of risky and interconnected non-bank institutions falling below the cutoff presents a greater concern than that of risky and interconnected banks below the cutoff.  As with banks, the threshold can be set low enough to make sure this will not happen.

This approach should not result in the Federal Reserve regulating or supervising non-banks in the same manner as banks, given the differences between banks and other financial institutions.  The Federal Reserve would have to determine, with wide public input, how best to regulate the non-banks.  Of course, the more risky and interconnected non-banks might have to be regulated differently than their peers.  But these differentiations would be made within the Federal Reserve.  Using an asset threshold to set the Federal Reserve’s jurisdiction would allow the Fed to calibrate regulatory requirements along a continuum for particular firms based on their interconnectedness and other indicators of systemic risk.  Thus, the Federal Reserve could set higher capital requirements for firms that pose greater systemic risk than for firms that pose less.

Under this proposal, a desirable level of ambiguity would remain as to how systemically important the Federal Reserve actually considered a given institution to be and creditors would, therefore, have to be more on guard against failures.


Hal S. Scott, Director

[1] Letter from the Comm. On Capital Mkts. Regulation to Christopher Dodd, Chairman, Richard Shelby, Ranking Member, S. Comm. on Banking, Housing & Urban Affairs and Blanche Lincoln, Chairman, Saxby Chambliss, Ranking Member, S. Comm. on Agriculture, Nutrition & Forestry 2 (Apr. 26, 2010).

[2] Restoring American Financial Stability Act of 2010, 111th Cong. § 165 (2010).

[3] Comm. on Capital Mkts. Regulation, The Global Financial Crisis: A Plan for Regulatory Reform 204 (2009).

[4] See Nat’l Info. Ctr., U.S. Fed. Reserve Sys., Top 50 BHCs (Mar. 31, 2010), available at

[5] Nat’l Ass’n of Ins. Comm’rs, Top 20 Property/Casualty Groups in Terms of Assets (as calculated by the Ctr. For Risk Mgmt. & Ins. Research) (Dec. 30, 2008).

[6] See Damian Alexander, Global hedge fund assets rebound to just over $1.8 trillion, GLOBAL REV. 2010 (Mar. 2010), available at from Pensions and Investments.