History suggests that the Senate Banking Committee’s proposal to raise the asset threshold for applying heightened regulatory standards to large banks goes too far.
The committee recently approved a bill that would increase the asset threshold for systemically important financial institutions from $50 billion to $250 billion. While Congress would be justified in raising the $50 billion line to $100 billion, the proposed $250 billion figure is much too high and would undermine the Federal Reserve Board’s ability to regulate SIFIs effectively under the Dodd-Frank Act.
Three emergency transactions during the financial crisis showed that financial institutions with assets of $100 billion to $250 billion can be systemically important.
In early 2008, the Fed approved Bank of America’s hasty acquisition of Countrywide — a $200 billion thrift holding company — to prevent Countrywide’s collapse. Countrywide’s subprime and Alt-A mortgage operations subsequently inflicted large losses on Bank of America. Those losses were a significant factor behind the federal government’s decision to provide $40 billion of capital assistance, $120 billion of asset guarantees and $44 billion of debt guarantees to Bank of America.
Later that year, the Fed approved PNC’s emergency acquisition of National City to prevent National City’s failure. National City was a $145 billion bank holding company that suffered crippling losses from its nonprime mortgage activities. PNC received $7.7 billion of federal capital assistance and $5 billion of federal tax benefits to support the acquisition.
Also in late 2008, the Fed approved the emergency conversion of General Motors’ financing arm, GMAC, from an industrial banking company — with $210 billion of assets — to a bank holding company. The conversion placed GMAC under the Fed’s protective umbrella and allowed it to qualify for a full range of federal assistance. Federal agencies subsequently provided $17.2 billion of capital assistance and $7.4 billion of debt guarantees to ensure GMAC’s survival.
In February 2009, as federal regulators prepared to conduct the first stress test for the 19 largest U.S. banks, they announced that the federal government would provide any additional capital needed to ensure the survival of all of those banks. After completing the stress test, the federal government injected new capital into GMAC after it could not raise capital on its own.
Crucially, federal agencies made clear in 2009 that the 19 largest banks — each with assets of more than $100 billion — were systemically important and “too big to fail.”
The proposed $250 billion asset threshold would have excluded Countrywide, National City, and GMAC. It would also exclude a number of large bank holding companies that are likely to be systemically important during the next financial crisis. Big regional banks and the largest money center banks have held highly correlated risk exposures during every U.S. banking crisis since 1980.
Those correlated exposures resulted from very similar business strategies that many large banks pursued during the boom leading up to each crisis. Regulators felt compelled to rescue troubled regional banks in order to prevent contagious runs by creditors on money center banks. There is no reason to believe that “next time will be different.”
Although the Senate bill permits regulators to apply enhanced regulatory standards to some banks smaller than $250 billion at their discretion, the reality is that it would be extremely hard to do in practice. The federal court decision to throw out MetLife’s SIFI label — which remains in force after Treasury abandoned its appeal — would pressure the Fed to justify every such action with a detailed cost-benefit analysis. In addition, the reasoning in the court’s MetLife decision would support close scrutiny of the Fed's cost-benefit analysis. It is therefore very doubtful whether the Fed would be willing to devote the necessary institutional resources to defend efforts to regulate companies falling below the new threshold.
A $250 billion cutoff and the MetLife precedent would also pose formidable challenges for any attempt by the Fed to regulate nonbank financial companies with assets under $250 billion. Dodd-Frank does not prescribe a particular asset threshold for designating nonbank financial companies as SIFIs. However, a decision by Congress to establish a $250 billion asset threshold for bank holding companies would create a powerful analogy. The new threshold would enable nonbank companies with assets under $250 billion to argue that the Fed lacked any persuasive basis to treat them as SIFIs under Dodd-Frank.
The Treasury under President Trump has already recommended that regulators not pursue nonbank SIFI designations except as a “last resort.” However, it is possible the next administration will feel differently and may find it very difficult to designate nonbank companies smaller than $250 billion if this higher threshold is enacted.
We have already seen that larger nonbanks can put the financial system at risk, even when they are below the $250 billion line. Long-Term Capital Management, a highly aggressive hedge fund, had only $130 billion of assets in late 1998, when the Fed arranged an emergency rescue of the firm by 14 of the largest U.S. banks and securities firms. The Fed orchestrated that rescue because it concluded that LTCM’s failure would trigger contagious runs and a systemic meltdown in global markets for derivatives, junk bonds and other risky financial instruments. The company had more than $1.25 trillion of derivatives contracts, but contingent risk exposures under derivatives would not be counted in applying a $250 billion asset threshold. The LTCM debacle showed that a nonbank financial company with over $100 billion of assets can be systemically important if it has extensive interconnections and asset correlations with big financial institutions.
Congress would be justified in raising Dodd-Frank’s threshold for bank SIFIs from $50 billion to $100 billion. As show above, $100 billion was the clear demarcation point for “too big to fail” status at the height of the last financial crisis. However, $250 billion is much too high.
Our experience in recent decades demonstrates that we must err on the side of over-inclusion in regulating the enormous — and costly — risks posed by SIFIs.