My op-ed
But the addendum to that is, in exchange for meeting that capital benchmark, which prudential rules can be taken off of banks’ plate that won’t then harm financial stability?
Marianne Lake, the JPMorgan Chase chief financial officer, said recently, according to the Financial
An important question is: What rules and processes might be rolled back productively, versus others that are and will remain key components of prudent bank regulation in the second decade of the 21st century?
I have not made a comprehensive survey of which aspects of supervision should remain and which should be eliminated or curtailed. That would be an undertaking for after the new stress-test regime is in place and the continuing validity of living wills has been assured.
But here’s a rough idea of forms of regulatory relief that a stress test benchmark could enable, versus rules that should remain in place to ensure the safety of the financial system.
Leveraged loan guidelines
Starting in 2013, the regulators imposed guidance on leveraged lending for financial institutions, a response to concerns that leveraged loan underwriting had become too relaxed. As Lisa Abramowicz pointed out in a recent Bloomberg.com
The leveraged loan guidelines provide a good illustration of how stress testing satisfies prudential regulatory goals—and how it does so in ways that a historically based capital benchmark would not.
If we look at a bank’s capital strength post-stress test, any significant weaknesses in the credit quality of its loans will have been exposed by the severe stress test scenario, and such weaknesses will have been put into the context of the individual bank’s capital strength and overall risk profile. If the leveraged loans made by a particular bank are riskier than the bank’s overall risk, and the bank’s capital profile cannot adequately absorb that risk in a recession, then the bank will flunk the stress test.
But if the leveraged loans are not riskier than the bank’s overall risk profile can withstand, then the bank can retain sufficient capital to pass the test despite losses on leveraged loans.
Any rule or guidance dealing with the underwriting or credit quality of a particular category of loans could be repealed for banks that are subject to rigorous stress testing and that meet stringent post-stress capital criteria.
Once a bank passes a rigorous stress test, restrictions on particular classes of loans add nothing to the fulfillment of the proper goals of prudential supervision, which are, in the case of individual banks, to try to assure that their capital strength is sufficient that even in a severe recession, their short-term liabilities will not run and they can continue to make loans and serve their customers.
On-site examinations
One should go further, however. For example, with respect to a bank that passes the rigorous stress test, what is the point of an on-site examination? A great deal of on-site examination time is spent reviewing loans and classifying them, which may result in write-downs or additions to loan loss reserves.
Based on the examination, the bank is assigned a Camels rating. But the most substantial areas affecting the Camels score are arguably capital strength (the C in Camels) and asset quality (the "a"). Those two categories really boil down to capital strength, since asset quality outside the context of capital strength has little meaning.
Moreover, the asset quality and capital strength that are measured in the Camels process use historical financial statements, which tell us far less about asset quality and capital strength than stress tests do. For banks with a high stress test performance (those that, subject to rigorous testing, emerge with continued strong capital), on-site examinations are of little utility, compared with their expense and use of the time of bank personnel.
What supervisory steps should remain, or even be strengthened
However, there are aspects of supervision that are still necessary and should be preserved in the regulatory reform initiative. Those include methodologies that test and influence a bank holding company’s culture, and those that permit regulators to see into a bank’s balance sheet to ascertain whether there have been significant changes since the latest stress test.
Forward-looking capital requirements and an ethical corporate culture are the two necessary pillars of bank supervision.
Cultural oversight seems especially important for holding companies that have significant securities business subsidiaries because the culture of such subsidiaries can be so different from the culture of the commercial banking subsidiary.
More intensive oversight of the securities subsidiaries may be warranted since, by their nature, those subsidiaries are more difficult to stress test than commercial banks.
Annual meetings between a bank’s principal regulator and the bank’s management and, separately, meetings with the board of directors, are also important so that the regulator has an opportunity to communicate with both management and the board, but also so that the regulator can learn more about the way the bank or holding company is being managed.
Financial instability always remains a threat
I am not advocating a general elimination of some aspects of supervision simply because the time has come to scale back regulations put in place since the financial crisis. Far from it. I am advocating reexamination of the methodologies to assess the effectiveness both of regulations put in place since the crisis, and of regulation put in place over the prior 40 years.
There is not less need for regulation than there was in 2010. The need is just as great—and it always will be—because financial systems are by their nature unstable. But as long as the Fed pursues its policy of rigorous stress testing (for complex bank holding companies, supplemented by credible living wills), there is room to pare back earlier regulation very substantially.
If the U.S. instead pursues noncredible stress testing (like that favored by some European regulators) then before very long instability will return, as it always has in the past.
