Regulators warn banks on leveraged loans, but data doesn't back them up
Last week, the federal banking agencies released their review of shared national credits — the $4.4 trillion in syndicated loans that are held by three or more regulated financial institutions. This program was created to evaluate large loans held by multiple banks. A key objective of this program is to determine whether the different banks are assessing the risk of their common loan consistently. For this purpose, supervisors conduct two examinations each year that review a sample of syndicated loans at banks that are agents or participants in a SNC.
The last review found that the percentage of loans held by all institution types (U.S. banks, foreign banks and nonbanks) rated below “pass” (that is, loans that pose greater risks) had declined by more than one third over the past year, falling from 9.7% in 2017 to 6.7% in 2018, its lowest level in over a decade. The drop was attributed in large part to improvements in economic conditions in the oil and gas industry. In addition, non-pass loans held by U.S. banks and foreign banking organizations declined an extraordinary 40% over the past year.
In the face of these facts, however, the agencies’ joint press release states that “the review finds increased risks associated with leveraged lending. The agencies remind banks to update credit risk management practices as the risk profile of borrowers and the industry changes.”
The report does not provide any statistical evidence that risks associated with leveraged lending have increased or that banks, as opposed to other holders of such loans, should be cautioned about those risks. As a consequence, investors in banks may falsely conclude that banks are significantly exposed to risk from leveraged loans, supervisors may falsely conclude that there is a safety and soundness risk to banks that justifies a supervisory response, and the Treasury’s Financial Stability Oversight Council may falsely conclude that any policy response should focus on banks rather than on nonbanks.
To understand the risks fully, one would certainly want to know the answers to three basic questions:
1. What share of leveraged loans are rated non-pass and what was that share in previous years?
2. What share of leveraged loans in SNC are held by banks?
3. What is the share of leveraged loans held by banks rated non-pass; what is the share of leveraged loans held by nonbanks rated non-pass?
To be clear, the SNC data can easily be used to answer these questions. Given the high and longstanding emphasis that the agencies continue to place on the risks associated with leverage loans, we would urge the agencies to provide those answers.
In the meantime, after some sleuthing, we were able to get an answer to question No. 1 for 2018 and 2017 by piecing together other statistics provided in the reports and in the press releases. Our analysis shows that the share of leveraged loans rated below pass was 11% in 2018 — down from 15% in 2017.
Note that the percent of loans rated non-pass is precisely the measure used by the agencies to conclude that the “risk in the portfolio of large syndicated loans has declined.” If that is the conclusion for all loans, and given that the share of non-pass loans for leveraged loans has declined, how did the agencies conclude that the risk of leveraged loans had gone up?
One possible explanation would be that the banks now hold a greater share of leveraged loans and so are more exposed to the associated risk. The report explicitly states, however, that “[m]ore leveraged lending risk is being transferred to … nonbank entities.” The report does not provide the information necessary to calculate the share of leveraged loans held by banks — that is one of the questions we encourage them to answer (No. 2) — but we expect that answer is “low.”
Another possibility is that banks hold a relatively higher share of the riskiest leveraged loans relative to nonbanks. Again, that is another question we encourage the agencies to answer (question No. 3). We do know, however, as shown in the chart, that only 2% of SNC loans held by banks are classified as substandard, doubtful or loss, while 12% of the loans held by nonbanks are classified that way.
We therefore suspect that the share of the riskiest leveraged loans held by banks is lower, not higher, than the share of the riskiest leveraged loans held by nonbanks. Of course, overall nonbanks also hold a higher share of leveraged loans relative to banks.
We hope that the Fed will publicize the answer to these important questions.