Board of Governors of the Federal Reserve System

I want to thank the staff who worked on the proposals and presented today. As I have said previously, a well-capitalized banking system is critical to the resilience of our financial system but we must recognize that capital requirements are not free.¹ We must balance the benefits of capital requirements against the costs to bank customers and to the real economy. I believe the proposals the Board is considering today achieve this balance by improving the risk sensitivity of our capital requirements while not unnecessarily increasing them.

Before getting into the details, I would like to thank Vice Chair for Supervision Bowman for constructing proposals that consider all aspects of our capital framework, including the interactions between the regulatory capital rule and the stress test. To get this right, we need to be mindful of how all of the pieces fit together.

I previously raised concerns regarding potential increases in operational risk and market risk capital requirements from Basel III implementation given we already project these risks in the stress test. This issue has been addressed in the Board's recent proposal to increase the transparency and public accountability of the stress test which, when combined with the steps proposed today, largely neutralizes the overlap.

I also support the proposed changes related to mortgage origination and mortgage servicing. Mortgage-related activities have largely migrated to the nonbank sector and I support any steps we can take to prudently adjust our regulations to reduce disincentives for banks to participate in this area. I also support the broader changes to adjust our standardized approach that applies to most banks in order to better align our requirements with the risk of traditional lending activities.

The G-SIB surcharge proposal would improve the calculation of the surcharge, including modifying several factors used to calculate the capital requirement by indexing them annually based on a three-year moving average of annual level of nominal U.S. gross domestic product (GDP). I support these changes. The other capital proposals would index various thresholds based on adjusting for inflation using the consumer price index for urban wage earners and clerical workers (CPI-W), and the G-SIB surcharge proposal would seek comment on an alternative approach whereby those factors would be indexed based on CPI-W. Let me take a few moments to talk about this indexation debate and why I think it is critical to index bank size-related regulatory thresholds and the G-SIB surcharge to nominal GDP rather than tie it to any CPI measure.

What matters for assessing the systemic importance of a bank is its nominal size relative to the nominal size of the U.S. economy. The absolute nominal size of a bank is not relevant. For example, in 1930, Bank of America was the largest bank in the world with $6 billion dollars in assets. That's the nominal size of a community bank today.

The U.S. economy grows in nominal terms for two reasons: 1) increases in the general price level of all goods and services, and 2) increases in the quantities of goods and services produced in the economy. Prices increase for all goods and services such as consumer goods, investment goods such as equipment and software as well as government provided goods and services. The quantities of goods and services increase because of investment in building and machinery, technological change, and increasing labor input. The key point is that banks' assets and liabilities also increase over time for both reasons. Banks take in more deposits as the population grows and real incomes grow. They make consumer loans to their customers to buy consumer durables. Banks finance lending for firms to invest in machinery and software as well as for firms to develop new technologies such as artificial intelligence, or AI. So, banks are involved in all parts of the economy and naturally grow as the economy grows. If they grow at the same rate as the U.S. economy, then their relative size and systemic importance does not change.

Indexing our regulatory thresholds is critical to ensure that, as the economy grows, banks can grow as well and in a way that does not subject them to increasingly stringent standards for reasons that have nothing to do with their relative size or risk to the U.S. economy. By not indexing thresholds, or indexing them incorrectly, regulators create distortions in the banking system that are unwarranted and unjustified.

The question is what is the appropriate indexation method? The other capital rulemakings propose adjusting bank size-related thresholds for inflation using a consumer price index measure. While I can see the instinct to do this, since many nominal values are indexed to inflation and inflation alone, I do not understand why we would use CPI-W, the price index for urban wage earners and clerical workers, as the indexing method for judging bank size-related thresholds relative to the U.S. economy. Let me give some reasons why I see this as problematic.

Why would we want to only consider consumer goods and services prices and not prices of investment goods and technology, which banks directly finance? CPI-W measures the price level for urban and clerical workers. Are those the only economic actors that banks interact with? Why focus only on consumer inflation and ignore real economic growth as a factor for indexing a bank's size and capital structure?

CPI-W is a measure of the change in prices paid by certain consumers based on a basket of goods and services with a heavy emphasis on the change in prices of three items: housing, transportation and food. Those three items account for around half of the price index. Do we want these three items to be the drivers of a bank's capital structure? On what rational grounds?

Constructing a price index requires a lot of assumptions such as what items to include, what is the appropriate expenditure share on those items, and how to adjust both items as the economy changes over time. It also requires selecting a particular indexation procedure and whether to use a chained or unchained index. These are complicated technical issues for measuring inflation. The beauty of using nominal GDP is it's a simple-sum method of aggregation that does not require making these types of assumptions and it's easy to understand.

By not indexing a bank's size to nominal GDP in the past, we have created distortions to the banking system and have imposed excessively stringent capital requirements that are unwarranted. So, we face two questions. First, do we want to eliminate future distortions to capital requirements going forward by indexing to nominal GDP? Second, do we want to undo all or at least some part of distortions that occurred in the past because of a lack of indexation to nominal GDP? The G-SIB surcharge proposal today says yes to both questions.

In summary, I support these capital proposals and the indexation method used in the G-SIB surcharge proposal. This is the correct way to proceed and I look forward to reading the comment letters on this issue.

1. See Statement by Governor Christopher J. Waller, Board of Governors of the Federal Reserve System, July 27, 2023. Return to text