Seeing the Forest for the Trees: GSIB Capital and Enhanced GSIB Regulation

This blog post was written jointly by economists at the Financial Services Forum (Sean Campbell) and The Clearing House (Francisco Covas and Bill Nelson).

Each of the banks in the United States designated by the Basel Committee on Banking Supervision as a Global Systemically Important Bank (GSIB) is required to maintain a capital level that is higher than that required of other banks by an amount equal to their “GSIB surcharge.”  As of 2017, the eight U.S. GSIBs make 45 percent of bank loans to businesses and households and support roughly three quarters of the credit intermediation in capital markets, such as initial public offering transactions, as well as the credit needed for international trade.  The GSIB surcharges make these institutions more resilient to losses, but because capital is an expensive source of funding, unnecessarily high surcharges can lead to reduced credit intermediation and a higher cost of credit for businesses and households.

The Basel Committee agreed on a framework for GSIB surcharges that is tied to the regulator’s assessment of systemic importance.  A higher degree of systemic importance results in a higher GSIB surcharge.  The underlying theory behind this approach is that GSIBs are more complex and more costly to resolve in the event of failure than non-GSIBs.  Moreover, the theory holds that the failure of a GSIB would have significant spillover effects on the rest of the economy.  As a result, a systemically important bank should fund itself with a higher percentage of equity capital and therefore be less likely to fail than a non-systemically important bank.

In calibrating the GSIB surcharge in the United States, regulators sought to ensure that the “expected impact” of a GSIB failure (that is, the probability of failure times the social cost of failure) is equal to the expected impact of the failure of a non-GSIB.  Under this approach, the GSIB surcharge was set so that the probability of default of a GSIB multiplied by the estimated cost to society of a GSIB’s failure is equal to the probability of default of a non-GSIB multiplied by the cost to society of the non-GSIB failure.

U.S. regulators finalized this approach to calibrating the GSIB surcharge in July of 2015.  Since that time, GSIBs have been required to comply with several additional regulations that reduce the probability that a GSIB will fail and reduce the cost of resolving a GSIB if it does fail.  Accordingly, these regulations have lowered the expected impact of a GSIB failure (probability of failure times social cost of failure).  As a result, if the GSIB surcharges were appropriately calibrated in 2015, then the surcharges are now too high and should be reduced.  Such reductions would, in turn, increase the supply of credit GSIBs can provide to businesses and households, while establishing capital levels that continue to satisfy the equal expected impact objective.

A key example of a post-crisis reform that reduces the probability of GSIB default and reduces the cost of resolving a GSIB is the Federal Reserve’s Total Loss Absorbing Capacity (TLAC) rule that was finalized in December 2016 – well after the GSIB surcharge was finalized in the United States.  This rule requires that GSIBs issue certain amounts of long-term debt that can be converted to equity if a GSIB were to fail.  The re-capitalization of the GSIB with the conversion of the long-term debt ensures that the GSIB can continue operating as a going concern after its resolution.  This approach avoids a costly and disorganized “collapse” that would otherwise halt credit and could impact the economy more broadly.

An impact study of the TLAC rule that an international committee of regulators, the Financial Stability Board (FSB), conducted in November of 2015 provides empirically based estimates of the reduction in default probability and resolution cost resulting from the new requirements.  The impact study cites research by Afonso et al (2014)that estimates that long-term debt requirements reduce banks’ default probabilities by 30 percent.  Afonso et al (2014) argues that this reduction is the result of better oversight and discipline of long-term debt investors who know that they will have their long-term debt converted to equity in the event of a resolution.  The FSB impact study also presents research suggesting that TLAC reduces the cost of resolutions—and ultimately the cost of GSIB failure-induced financial crises—by roughly 10 percent.  In combination, these two effects therefore reduce the expected systemic impact of a GSIB failure by about one third.

In finalizing the GSIB surcharge in 2015, the Federal Reserve released a whitepaper explaining its calibration that used a specific relationship between capital levels and default probabilities.  Using that relationship, the reduction in expected systemic cost of default from TLAC estimated by the FSB allows for roughly a 1 percentage point reduction in GSIB surcharges while still satisfying the equal expected impact objective.  We show the specific GSIB surcharge that would result from recognizing the benefits of TLAC for each U.S. GSIB in the table below.  Those interested in the algebra underlying this result can find it here.  Intuitively, because GSIBs now have loss-absorbing long-term debt, they need less additional loss-absorbing equity capital to have the same expected social costs of failure as a non-GSIB.

Moreover, the FSB reached the same general conclusion using a completely different technique.  Specifically, the FSB TLAC impact study reports that the additional required long-term debt is roughly equivalent to a 1 percentage point increase in capital.

A one percentage point change in capital requirements may sound modest but the evidence suggests that such changes can have a real impact on credit provision.  For example, a review of academic research conducted by economists at the Central Bank of the Netherlands suggests that a one percentage point increase in capital requirements results in between a 1.0 and 4.5 percentage point reduction in lending.

These results have broader implications for the post-crisis reform framework.  When calibrating financial regulations, each regulation cannot be viewed independently.  Changes that are made to one regulation will often have implications for other existing regulations.  For example, a recent research note by Francisco Covas and Rob Lindgren of The Clearing House finds that, based on historical experience, GSIB compliance with the liquidity coverage ratio (LCR) (a post-crisis liquidity requirement) should reduce the magnitude of future GSIB losses.  GSIBs under pressure will be less likely to sell assets at “fire sale” prices to meet cash outflows, shifting the relationship between losses and capital used to calibrate GSIB surcharges.  Together, the TLAC adjustment to GSIB surcharges reported here and the LCR adjustment found by Covas and Lindgren imply GSIB surcharges should be reduced by between 1 and 1½ percentage point (see table below and here for details).

image of graph

Recently, the FSB, the Basel Committee, the Federal Reserve, and the U.S. Treasury have all either begun or recognized a need for comprehensive consideration of the costs and benefits of post-crisis regulations taken as a whole.  More specifically, Federal Reserve Vice Chairman for Supervision Randal Quarles recently stated that now is “an eminently natural and expected time to step back and assess” what changes should be made to the post-crisis regulatory framework.[1] The relatively simple analysis presented here indicates that there is considerable scope for such reviews to reveal ways that regulations can be adjusted that both maintain bank resiliency at intended high levels while also increasing the availability of credit to businesses and households, supporting economic growth.

Disclaimer: The views expressed in this post are those of the author(s) and do not necessarily reflect the position of The Clearing House or its membership, and are not intended to be, and should not be construed as, legal advice of any kind.

[1] Randal K. Quarles “Early Observations on Improving the Effectiveness of Post-Crisis Regulation”,