By Oz Shy
Last month, the Federal Reserve announced that 31 out of 33 U.S. banks had passed its latest “stress test,” designed to ensure that the largest financial institutions have enough capital to withstand a severe economic shock.
Passing the test amounts to being given a clean bill of health by the Fed. So are taxpayers – who were on the hook for the initial US$700 billion TARP bill to bail out the banks in 2008 – now safe?
Yes, but only until the next crisis.
Skeptics of these tests (myself included) argue that passing them will not prevent any bank (large or small) from failing, in part because they’re not stressful enough and the proposed capital requirements are not high enough.
But beyond this, the stress tests highlight a significant shortcoming in how regulators hope to prevent the next wave of bank failures: They’re focusing way too much on size, particularly with the designation of so-called systemically important, “too-big-to-fail” banks.
U.S. lawmakers in search of a solution are currently working on legislation that would make it easier for too-big-to-fail banks to actually fail through bankruptcy. While doing so would be a good thing, it still raises important questions.
Are policymakers right to focus on size in determining whether a bank poses a major risk to the financial system and taxpayers? Would splitting larger banks into smaller ones free taxpayers from the repeated burden of rescuing them during times of crisis? Does calling a bank “too big to fail” even mean anything?
To me, this focus on size and “too big to fail” seems misplaced. I’m among those who advocate replacing our current system with something known as “narrow banking,” which would totally separate deposits from riskier lending activities. This would have the best chance of protecting taxpayers from having to foot the bill for future bailouts, as I’ll explain below.
What’s too big to fail
The 2010 Dodd-Frank Wall Street Reform and Consumer Protection Act requires the Fed to conduct an annual stress test of bank holding companies with $50 billion in assets, often referred to as too big to fail.
Dodd-Frank also says that banks and other companies that “could pose a threat to the financial stability of the United States” if they fail or engage in very risky activities must write bankruptcy plans known as living wills and meet stricter capital requirements.
Some policymakers contend that the increased regulation and capital are not enough and have called for breaking up big banks into smaller ones in order to reduce the probability of having to use taxpayer money to bail them out.
A question of size
In order to make sense of the too-big-to-fail slogan, we first must agree (or disagree) on what we mean by “big.” So, let us examine the biggest bank in the U.S.: JPMorgan Chase.
JPMorgan maintains about $1.39 trillion worth of deposits. Suppose we break up JPMorgan into four banks, each with approximately $348 billion of depositors’ money. Are these “baby” JPMorgan banks now “too small to fail”? Clearly not, since in the event of a bank failure, taxpayers may still be on the hook for up to $348 billion, each.
So if that’s not small enough, let’s divide them another time, into eight banks that each handles $174 billion in deposits. Could we now regard these new “baby-baby” banks as “too small to fail?”‘ Again, should we consider $160 billion not a small sum if the bank goes under and needs to be rescued by the government?
By this logic, we’d of course also need to break up the other big banks, such as Bank of America and Wells Fargo (each with just under $1.3 trillion in deposits) and Citibank ($947 billion).
The idea that breaking up banks into smaller banks reduces risks is an abstraction since our repeated experience (from the Great Depression to the Great Recession) shows that many banks tend to fail at the same time like dominoes, which by definition we call a “financial crisis.”
Would it really matter to the taxpayer whether a large JPMorgan fails or several “baby JPMorgans” collapse at the same time?
So size does not matter after all. If the term “big” does not make a lot of sense, then why do regulators and central banks keep repeating this slogan?
In fact, my reading of too big to fail is that, intentionally or unintentionally, it sends banks the wrong message. That is, regulators and central banks are basically telling the “big” banks that they should not worry very much because taxpayers and the Fed will always rescue them because they are too big to fail.
Therefore, intentionally or not, the assertion of too big to fail and passing artificial stress tests may actually generate exactly the opposite incentive for financial institutions, which may encourage large banks to continue taking risks with depositors’ money. In fact, using data for more than 200 banks in 45 countries, a New York Fed paper found that banks classified by rating agencies as more likely to receive government support engage in more risk-taking. The authors also show that riskier banks are more likely to take advantage of potential government support.
The stated logic behind the classification of too big to fail was to impose more restrictions on these financial institutions and conduct stress tests so they will become less likely to fail. The additional restrictions that central bankers are now considering imposing on banks, such as maintaining capital ratios as high as 16 percent to 18 percent, still do not free the taxpayer from bearing the remaining 80 percent of the bailout cost.
Although 16 percent seems like a high number compared with the 12.3 percent capital that large banks were assumed to maintain during the recent stress test, even 20 percent capital requirement leaves banks with substantial leverage. (A capital ratio just shows how much of a bank’s assets are backed by cash or cash-like instruments, as opposed to leverage, or debt.)
Let banks fail
If the TBTF and stress tests send the wrong message to banks, then I would suggest replacing too big to fail with “all banks, regardless of size, should be allowed to fail” with no taxpayer bailout.
The problem is such a statement suffers from what economists call “time inconsistency.” That is, policymakers may be able to take a hard line now, but once the banking system collapses, there is no government in the world (democratic or otherwise) that would be able to resist the political pressure to pay the banks whatever they want just to bring them back to life.
To solve the time inconsistency problem – and ensure policymakers let banks fail – we need to prepare in advance for the next wave of bank failures by protecting depositors’ money, instead of just focusing on stress tests or size reductions.
And this is actually easier to implement than one may think.
Depositors should be simply allowed to have access to accounts that maintain 100 percent reserves. That is, every cent of their savings would be backed by hard currency. Research of mine has shown that moving in this direction improves social welfare relative to the current system of purely fractional banking, in which banks only hold a fraction of their deposits. Currently they are allowed to lend up to 90 percent (keeping 10 percent as reserves).
There are several ways to implement that. One is to allow each depositor to open an online account with the central bank. This could be done via existing commercial banks or directly with the central bank. In addition, if central banks, such as the Fed, provide direct access via nonbank money transmitters (such as PayPal, Square, Western Union or even Wal-Mart’s Bluebird), depositors would be able to secure their money in advance against any loss, even in the event many banks and money transmitters fail. In fact, Mark Carney, governor of the Bank of England, recently announced that the U.K. central bank will expand direct access to nonbank payment providers.
Why haven’t I mentioned the Federal Deposit Insurance Corporation (FDIC), which was set up explicitly to protect deposits in the case of bank failure? Simple: In 2015, the FDIC fund had just $67.8 billion. Dividing this amount by total U.S. deposits implies that the FDIC can bailout only 1.06 percent. In fact, even if JPMorgan is the only bank that fails, the FDIC can only bailout 5 percent of this bank’s deposits, thereby making the FDIC totally irrelevant during a financial crisis.
On the other hand, a 100 percent reserves policy would break our current system’s bundling of risk-taking with the job of keeping accounts safe and offering payment services. Only then, by ensuring depositors (and voters) aren’t at risk when there’s a crisis, would governments have the will to let banks fail – without any regard to their size – and at no cost to taxpayers.