“We have run an extremely adverse scenario that assumes … [a] contraction of gross domestic product, down as much as 35% in the second quarter and lasting through the end of the year, and with U.S. unemployment continuing to increase, peaking at 14% in the fourth quarter…. If [this] were to play out, the Board would likely consider suspending the dividend….” Jamie Dimon, Chairman & CEO Letter to Shareholders, JPMorgan Chase & Co., April 2020.
In the aftermath of the financial crisis of 2007-09, authorities were intent on making the financial system resilient to a very large shock. Well, COVID-19 is upon us, and the shock looks to be bigger than the most adverse scenarios in supervisory stress tests. Private forecasts are approaching the dire scenario that Jamie Dimon describes to the shareholders of JP Morgan Chase: a severe plunge of activity in the current quarter and an increase of the unemployment rate to a post-Depression high.
Banks will not be able to dodge the financial fallout. Many borrowers are likely to suspend repayment soon, presaging widespread defaults. We will not know the extent of the damage, or who will ultimately bear the costs, for some time. What we do know is that there will be large losses that we need to contain and share.
In the 2007-2009 crisis, everyone’s worry was about the extent to which the financial collapse would damage the real economy. Policymakers’ primary concern was repairing financial institutions and markets so that they could perform their role of efficiently channeling resources in support of general economic activity. Preventing another Great Depression required that authorities backstop large intermediaries, as well as ensure the function of money market funds, commercial paper markets, and the like.
This time really is different (see here). Rather than the financial system undermining the real economy, it is very much the other way around. With few exceptions (like Sweden), advanced economies have entered a form of suspended animation. As a result, households and businesses are losing income that they can never replace. The hope is that the COVID-19 crisis does not trigger a full-fledged financial crisis, exacerbating what is already destined to be the most severe global downturn since the 1930s.
Returning to the post-2009 financial regulatory reforms, authorities had two complementary goals: increase the financial system’s reliance on equity financing and enhance the ability of institutions to recapitalize themselves after a shock. The idea was that these improvements would allow financial institutions, especially banks, to weather an extremely severe storm.
Our view is that we have made limited progress. In a recent post, we emphasized how COVID-19 is eroding banks’ capital buffers (that already were slim in parts of Asia and Europe). Regular readers of this blog know that we believe capital levels should have gone up much more than they did prior to the pandemic. In recent years, we were especially disheartened that the U.S. authorities allowed banks to make payouts (dividends and share repurchases) in excess of their profits, resulting in a falling leverage ratio (see here).
The following chart illustrates the results of these lenient policies. We plot net share repurchases—which represent three fourths of banks’ 2019 payouts—for the eight global systemically important banks (G-SIBs) in the United States from 2012 to 2019. Not only did the aggregate increase by at an average annual rate of roughly 40%, but with the exception of Goldman Sachs and State Street, the buybacks of each individual bank increased year after year.
Share repurchases net of issuance by the U.S. G-SIBs, 2012 to 2019
Fortunately, the U.S. G-SIBs recently suspended buybacks (see here). But how big were the pre-pandemic buybacks compared to banks’ capital buffers? At the end of 2019, the standard measure of their capital―common equity tier 1 (CET1)—totaled $801 billion. Their leverage ratios—comparing CET1 to the Basel III total exposure measure—ranged from about 5 to 6 percent. Had the banks not engaged in buybacks since 2016 and retained a combined $306 billion, they would have gone into 2020 with leverage ratios between 1.5 and 3.2 percentage points higher. In our view, this was an enormous missed opportunity.
We strongly suspect that, as the full impact economic and financial impact of COVID 19 becomes apparent, some banks will need a form of recapitalization. They were not able to do this the last time on their own. In the United States in 2008-09, the federal government, acting on behalf of taxpayers, provided the equity funds. Will this time be different?
In countries where large banks’ price-to-book ratios were well below one at the start of 2020, the answer is likely no. Governments will almost surely have to step in to backstop financial systems in parts of Europe and Asia. What about the United States?
Again, we are concerned. Before explaining our reasons, recall that, in addition to common equity, current regulations require that G-SIBs issue bail-in-able long-term debt that can replace lost equity without government financial support. These bonds, which make up more than half of each bank’s Total Loss Absorbing Capacity (TLAC), are subordinate to other debt, including uninsured deposits. Put differently, these bonds are the closest thing to equity among the banks’ fixed liabilities. The TLAC debt standard is roughly equal to the equity requirement. This means that bailing in the TLAC debt would be sufficient to recapitalize a bank at the required level should a shock wipe out its equity and force the bank into resolution.
Consequently, to assess the resilience of the banks, we need to consider their TLAC. The following chart shows the amount of TLAC debt as a fraction of each U.S. G-SIB’s CET1, both at the end of 2019. In every case, the number exceeds 100%, so all the U.S. G-SIBs have sufficient TLAC debt outstanding to replace completely their common equity (should that be necessary).
Estimated TLAC Debt of the eight U.S. Global Systemically Important Banks (Percent of CET1), end-2019
This looks encouraging. Cumulatively, these banks have $800 billion in CET1 and more than $1 trillion in TLAC bonds outstanding. Knowing that these loss-absorbing resources exist should be stabilizing, limiting runs on more senior liabilities, including uninsured deposits.
So far, so good. But, the ability to use these bonds to recapitalize the banks requires meeting two conditions. First, the bonds must still be outstanding when the time comes. Second, authorities must have a mechanism for compelling conversion, which is not automatic. On both counts, we are worried.
Starting with the outstanding quantity of bonds, there is rollover risk. To see how big that risk might be, consider the following chart that shows the maturity structure of Citigroup’s outstanding TLAC debt. (With virtually unique transparency, Citigroup publishes detailed TLAC information on its website.) To be TLAC-eligible, a bond’s maturity must exceed one year, requiring banks that are close to the 100% TLAC/CET1 ratio to roll over their 2021 maturity bonds during the course of 2020.
Citigroup TLAC-eligible debt by year of maturity (billions of dollars), 2021-2098
Assuming that all the G-SIBs have a TLAC debt maturity structure similar to Citigroup’s, and that the banks do not wish to see a decline in their TLAC debt/CET1 ratios, we estimate a 2020 rollover need of between $100 and $150 billion in TLAC-eligible bonds. Since TLAC debt has equity-like risks, it is easy to imagine a scenario in which one or more G-SIBs will struggle to rollover their TLAC (at least in the absence of a government backstop). With the S&P bank index down by more than 40 percent year-to-date, that day may not be so far off.
Turning to the second issue, our understanding of the U.S. arrangements is that the conversion of TLAC into equity occurs in resolution. Some years ago, we discussed the importance of setting up a system in which recapitalization can occur outside of resolution. Rather than conventional living wills—which aim to facilitate re-organization (or liquidation) in resolution—we suggested banks have phoenix plans that would allow them to rise out of the ashes. Forcing a bank into resolution seems unnecessarily risky and costly. It would surely be better to have a mechanism in place that provides for automatic recapitalization. However, TLAC conversion is not automatic, and authorities would be reluctant to trigger it outside of resolution, even if it were legally feasible.
As is true in a number of cases―the recent evidence of runs on money market funds and on illiquid, open-end bond mutual funds comes to mind―we really did know better. While banks had improved their capital positions substantially since 2009, many scholars argued that they needed more. Yet, since 2017, U.S. authorities facilitated capital distributions and declining leverage ratios. We knew that banks had improved their resolution planning, but we also knew that they would have incentive to delay recapitalization if a shock eroded their equity buffer.
In sum, the financial regulatory reforms of the past decade remain insufficient to ensure resilience of the system. We hope that the COVID-19 experience will usher in a new zeal for further strengthening the current framework. This means accomplishing three things. First, equity buffers must be larger. Second, when those buffers are likely to erode—as current equity prices indicate they are now—there must be an automatic trigger to ensure earnings retention. Suspension of dividends, stock repurchases and management bonuses is critical to rebuilding a bank’s equity. Third, we need a mechanism for automatic recapitalization of going concerns that does not rely on a taxpayer backstop.
We hope that when it comes to financial regulatory reforms, this time will be different, too.