fb-pixel Skip to main content

Was Silicon Valley Bank really too big to fail?

Even in a period of relative economic calm, a bank less than one-tenth the size of JPMorgan Chase was not allowed to fail without some special protection for one group of creditors — those large, uninsured depositors.

An SVB Financial Group chart displayed on the floor of the New York Stock Exchange on March 10. Michael Nagle/Bloomberg

The repercussions of the Silicon Valley Bank failure are still rippling through the financial system. While much remains to be learned about the management and regulatory failures that sparked this drama, some lessons are already clear.

There was a serious lapse in supervision by the Federal Reserve and the California Department of Financial Protection and Innovation. Silicon Valley Bank presented an almost textbook example of vulnerability: It had quadrupled in size in just a few years, usually a sign that a bank has outstripped its risk management capacity. Its business was heavily concentrated in a limited sector of the economy — startups and tech companies. At least 85 percent of its deposits were held in accounts with over $250,000, the upper limit for Federal Deposit Insurance Corp. insurance. Knowing their deposits were uninsured, these account holders ran when the bank’s problems surfaced. Meanwhile, the bank had invested a large portion of those deposits in longer-dated US Treasury bonds, which initially paid a higher return but declined in value as interest rates rose rapidly.


While it is unclear how much difference the regulatory rollbacks by Congress in 2018 and the Federal Reserve in 2019 made for Silicon Valley Bank, they surely didn’t help. It’s possible, though again not clear, that the pre-2019 Federal Reserve rules might have required the bank to take earlier account of those “unrealized” losses from depreciating Treasuries. If so, the bank would have needed to shore up its capital sooner.

Most significant, the nation learned over the weekend that Silicon Valley Bank, the 16th largest depository institution in the United States, was deemed by the government to be too big to fail — at least in the sense that the normal rules for allocating losses were set aside. Specifically, all those uninsured depositors with large accounts were made whole rather than being required to absorb some losses, as is contemplated in banking law. The government invoked the “systemic risk” exception in order to protect the uninsured depositors in Silicon Valley Bank and Signature Bank (which also failed) because of fears that other banks would suffer similar runs of uninsured depositors.


Just as the diagnosis of the origins of the failure will be better informed in the weeks to come, the appropriate prescriptions for regulatory change will also evolve. But steps are clearly needed to address each of the three problems just identified.

As the Federal Reserve has already acknowledged, it must determine why such a glaring supervisory failure occurred. Warning bells should have been clanging in supervisors’ ears. Why were they not heard? Or, if they were heard, why was there no supervisory response? Did the fault lie with the supervisory team dedicated to Silicon Valley Bank? Or were those on-site supervisors — and others around the country examining other banks — acting under instructions from Washington to go easier on the banks? Whatever the answer, a change in supervisory policies and monitoring will be needed.

Some changes in federal regulations will be required. Current rules on how banks must manage their liquidity have not prevented the concentrations of uninsured, runnable deposits that we now know exist in some banks. It is also important to revisit the rules on when banks must take “unrealized” losses on their holdings of securities into account for purposes of calculating how much loss-absorbing capital they have. More generally, it would be wise to reassess both the 2018 law and the 2019 Federal Reserve regulation. Especially troubling is the part of the Fed regulation that went beyond what Congress had required by relaxing requirements on banks with over $250 billion in assets. While those changes did not affect Silicon Valley Bank, there is no reason to wait for an even larger bank to get in trouble before reconsidering the 2019 deregulation.


Last weekend’s actions by the banking regulators and the Biden administration have significant implications for the regulatory system that emerged from the global financial crisis of 2007-2009. One key element of the law Congress passed in 2010 was a mechanism for ensuring that any bank — no matter how large — could be allowed to fail without endangering the financial system. That feature of the law was intended to avoid future bailouts of banks and their creditors, and consequently to give those creditors the incentive to make sure the bank with which they were dealing was being responsibly operated.

There has always been considerable doubt that the government would allow one of the nation’s biggest banks to fail in the midst of a financial crisis. But now we have learned that, even in a period of relative economic calm, a bank less than one-tenth the size of JPMorgan Chase was not allowed to fail without some special protection for one group of creditors — those large, uninsured depositors. The reality may be that uninsured depositors will always have to be protected if runs on other banks are to be avoided. If that conclusion proves correct, there needs to be stronger regulation of bank funding practices and, potentially, higher capital requirements as well.


Some have questioned whether the actions by the Federal Reserve, FDIC, and Treasury were necessary — that is, whether there really was systemic risk. In truth, almost any government faced with the prospect of serious financial stress will err on the side of containing the problem. After the stress passes, however, it is incumbent on the administration, Congress, and the bank regulatory agencies to take steps to address the source of the problem. We should all be watching how they respond to the demonstration that too-big-to-fail concerns may now be most acute in a group of banks considerably smaller than Citigroup and Bank of America.

Daniel K. Tarullo, who was oversight governor for supervision and regulation at the Federal Reserve from 2009 to 2017, teaches at Harvard Law School.