In recent weeks, public debate in the banking industry has centered on loosening stress testing rules for the largest banks. Democrats in Congress have discussed the prospect of removing the burden of Comprehensive Capital Analysis and Review (CCAR) for all banks with assets under $250 billion. Treasury Secretary Steven Mnuchin has gone further, suggesting that banks with assets between $10 billion and $50 billion be freed from all regulatory stress test scrutiny.
The Dodd-Frank rules draw sharp distinctions between adjacent banks. Technically, a $9 billion bank does not need to submit to the annual Dodd-Frank Act Stress Test (DFAST) while an $11 billion bank does. A similarly sharp cutoff exists in determining which institutions are subject to CCAR. The new rules under consideration would create a new class of super banks that limits the size of the CCAR cohort to roughly a dozen institutions.
So now seems like a good time to assess how banks of different sizes have performed during the Dodd-Frank era.
Using call report data from the FDIC, we are able to isolate consistent data on the performance of banks of different sizes going back to the 1990s. Our aim is to aggregate data on distinct segments of the banking industry to determine the impact of Dodd-Frank on their performance.
We resolve the issue of consolidation by considering the institutions as they were constituted in 2017. For example, data for Wachovia, which failed in 2008, are combined with data for Wells Fargo, its new parent, at all historical data points.
Let’s begin with a look at tier one capital ratios, one of the gold standards for capital adequacy. Chart 1 (below) shows aggregate ratios for these three market segments, plus the remaining cohort of small banks. In the worst days of the recession, representative CCAR banks held a 9% tier one capital ratio, the DFAST banks held 11%, and the smaller community banks held 13%.
Systemic Risk Impact
As one might expect, the Dodd-Frank rules have had a significant impact on large banks’ (i.e., banks with more than $250 billion in assets) capital levels. However, the biggest increase is seen for smaller CCAR banks, with assets between $50 billion and $250 billion. The increase for this group amounted to 500 basis points, well over the 380-point rise for the largest banks. Smaller banks and banks in the DFAST category witnessed much slighter rises, meaning that capital ratios are now converging across the industry.
In the context of eradicating systemic risk, one would think that capital levels would be rising fastest for the largest banks. That the smaller CCAR institutions are seeing larger capital increases suggests an overzealousness from regulators in their treatment of the lower reaches of the cohort.
Table 1 shows market share statistics (from 2001 – 2017) for the four market segments. We include data on all loans and leases, all mortgage products and all commercial and industrial (C&I) loans. The data represent average values across the periods defined.
Table 1: Market Share of Loans and Impaired Loans for Segments of the U.S. Banking Industry
Sources: FDIC Statistics on Depository Institutions, Moody’s Analytics
The lion’s share of assets are clearly held by the largest banks. Indeed, since the recession, this cohort has controlled 46% of all loans and leases in the U.S. banking system. This figure, however, is substantially lower than it was during and before the Great Recession.
The decline in share for the largest banks mirrors a similar rise in prominence for the lightly regulated $10 billion to $50 billion companies. These patterns, moreover, are consistent across different lending products: post-recession, big banks have lost six percentage points in share of mortgages, and seven percentage points in share of C&I loans.
By comparison, small banks and the lesser CCAR institutions have maintained the market positions they held prior to the recession. Small banks have gained some market share in mortgages while small CCAR banks have gained prominence in business lending.
Spreading the Risk: A Sea Change
These numbers suggest that the Dodd-Frank rules are squeezing banking activity out of the largest banks to the benefit of their DFAST brethren. These banks have the advantage of being large enough to stomach an increase in compliance costs while sidestepping the more onerous restrictions imposed on the larger CCAR institutions. In other words, the Dodd-Frank rules are succeeding in spreading banking activity more broadly across the industry.
To develop a measure of the riskiness of the cohorts’ respective portfolios, we can compare these market share statistics with equivalent values for the stock of impaired loans. We define impaired loans as all assets that are 30 or more days delinquent at any time. If risks are spread evenly across the industry, we would expect the share of impaired loans to align closely with each cohort’s overall market prominence.
Immediately prior to the recession, all four cohorts were close to par in this regard. Small banks had a slightly riskier profile than expected, mainly because this sector historically engaged in riskier lending to local households and businesses. In contrast, DFAST and “super-regional” small CCAR banks were slightly more conservative than the representative bank.
Since the recession, we have seen a sea change in these historical market positions. In mortgages, for example, small banks hold a mere 8% of all impaired loans – much lower than the 24.5% of all assets held by the sector.
Banks in the $50 billion-$250 billion range have also witnessed a sharp drop in the prevalence of bad mortgages. In large part, this difference is due to big banks taking on the toxic portfolios of banks that failed during the subprime crisis. DFAST banks, meanwhile, have seen impaired loans grow at a consistent rate throughout the sample period, in line with their overall growth in the mortgage sector.
Chart 2 (below) shows the evolution of loans/impaired loans for the two CCAR cohorts. The critical point illustrated here is that the sectors are gradually returning to the inevitable equilibrium – e.g., the share of bad loans in the largest cohort is declining toward the now lower market share.
Risk, at present, remains concentrated in the largest cohort but is slowly in decline. In part, this risk is being taken on by the DFAST banks, but a case could be made that regulations are curtailing the performance of the housing finance industry generally.
In C&I lending, which has not been affected by the same conditions that impacted the mortgage sector, we find that smaller CCAR banks’ portfolios have become far safer. While their overall market share has increased, their share of impaired loans has fallen. The C&I portfolios of the larger CCAR banks, meanwhile, have become relatively riskier since the recession: their share of impairments has fallen, but at a slower rate than their overall C&I market prominence.
There are three key takeaways from this analysis. The first is that the $50 billion threshold for CCAR has caused a disparity in performance between banks just below and just above the cutoff. DFAST banks are growing their market share in a risk-neutral manner (the $10 billion threshold is having little impact on the performance of DFAST banks), while smaller CCAR banks are raising more capital and draining risk in the context of a moribund market share. Raising the CCAR threshold from $50 billion to $250 billion could spark improved performance for smaller CCAR banks.
The second implication is that more consideration must be applied to the fate of the smallest community banks. In C&I, their performance has been solid, but in mortgage they have become extremely safe. Though pressure should be maintained on the largest banks to continue to shed risk in favor of smaller lenders, mortgage rules could be softened dramatically for banks below $10 billion without risking system safety.
The third takeaway is that Dodd-Frank has been successful thus far at shifting risk around the various parts of the banking system, effectively lowering the level of systemic risk. Whether these shifts are also curtailing economic growth, as suggested by some commentators, remains a question for another day.