Part IV in Our Series on US Bank Capital Requirements

• In Part II of this blog series, we discussed how the market risk component of the expected Basel III Endgame reforms – known as the Fundamental Review of the Trading Book (“FRTB”) – will effectively duplicate aspects of the Global Market Shock (“GMS”) component of the Federal Reserve’s Stress Capital Buffer (“SCB”) requirement.
• In this blog, we take a deeper dive into the reasons behind this double count. As we discuss below, the duplication arises because the trading and counterparty losses flow through both the SCB’s numerator (via GMS losses) and denominator (via market risk Risk-Weighted Assets or “RWA”). As a result of this double count, market risk capital requirements will be significantly greater than would otherwise be expected given the underlying risks it is seeking to capture.
• We propose simple adjustments that could meaningfully mitigate the impacts of the duplication between the two capital requirements. Those mitigations could include removing the GMS from the SCB, given that the market risk losses it was designed to measure are already captured by the FRTB; redesigning the GMS to be reasonably plausible; and setting capital requirements based on the larger of the FRTB and the GMS rather than the sum of the two components.

Identification of the FRTB and the GMS Overlap

Under the current US capital rules, for the purposes of the Standardized Approach, the minimum Common Equity Tier 1 (or “CET1”) capital requirements (in dollar amounts) for a US GSIB is calculated as the sum of 4.5% (the Basel Committee’s minimum CET1 requirements), the SCB, the Countercyclical Buffer (or “CCyB”, 0% currently in the US), and the GSIB surcharge (or “GSIB”), multiplied by the bank’s Standardized Approach RWA (or “RWA_SA”) which captures only credit risk and market risk losses, i.e.:

The SCB is a ratio of stress losses calculated through the supervisory stress test divided by a bank’s .  Abstracting from the various adjustments needed to calculate changes in regulatory capital,^[1] the stress losses can essentially be calculated as the sum of the provisions for credit losses (including credit losses on available-for-sale “AFS” and held-to-maturity “HTM” securities) under the Current Expected Credit Loss methodology (or “CECL”) and trading and counterparty losses under the GMS,^[2] minus the projected Pre-Provision Net Revenue (or “PPNR”,^{[3] which captures operational risk losses), i.e.:}

Focusing on credit and market risks, the key components for the SCB calculation are:

• CECL which captures the projected expected credit losses of banking book exposures;
• GMS which captures the projected trading and counterparty losses; and
• RWA_SA which captures the unexpected credit losses (RWA_credit ), and the trading and counterparty losses ( RWA_market), i.e., RWA_SA = RWA_credit + RWA_market

Therefore, the minimum CET1 capital requirements is approximately equal to:

The trading and counterparty losses are capitalized separately by the FRTB (via ) and by the GMS losses (inclusive of Largest Counterparty Default, or “LCD”, and CVA losses).  Both the FRTB and the GMS are designed to be stress testing frameworks with largely overlapping objectives, risks capture, and modelling methodologies.  Therefore, they essentially double count market risks, leading to capital requirements significantly greater than would be expected given the underlying risk.

Although the double counting between the FRTB and SCB is the most pronounced, it is not the only area where the SCB will overlap with the Basel III Endgame reforms, if the SCB is applied to the Endgame capital requirements.  While there is limited overlap between the SCB and the Endgame’s credit risk capital requirements (because the  captures the unexpected credit losses while the provisions for credit losses captures the expected credit losses), there will be duplication between the Basel Endgame’s operational risk and CVA requirements and the SCB.   This is because operational risk losses and CVA risk losses are captured by the PPNR and the GMS respectively (since the SCB denominator, i.e., , includes capital requirements for the operational risk and CVA risk losses along with the credit risk and market risk losses).  This is another example of potential double counting that will need to be addressed by regulators as they implement the Basel III Endgame reforms.

Impacts of the FRTB and the GMS Double Count

To illustrate the impacts of the FRTB and the GMS overlap using a simple equity trading portfolio, Table 1 presents the applicable risk-based capital requirements (in terms of CET1) as a percentage of the portfolio’s market value, and the portfolio’s corresponding RWA density.^[4]  A larger RWA density means higher capital requirements.

Table 1.  CET1 Capital Requirements as Percentage of The Portfolio’s Market Value.

For this stylized trading portfolio consists publicly traded and AA+ rated equity exposures and a vanilla option on the equity, and accounting for all applicable risk-based capital requirements, the resulting minimum required CET1 capital amounts to nearly 60% of the portfolio’s market value.  The corresponding RWA density equals 758%, which is nearly double the 400% risk weight assigned to “an equity exposure … that is not publicly traded” under the current US capital rules,^[9] and the risk weight assigned to “speculative unlisted equity exposures” under the Basel III Endgame.^[10] This is a clear example of how the combination of the two capital requirements results in unnecessarily punitive treatment that is incommensurate with risks of the portfolio.

The fact that FRTB capital requirements and the GMS losses are comparable is not surprising given that both the FRTB and the GMS heavily overlap in terms of design objectives, risks capture, and modelling methodologies, as detailed in Table 2 below.

Table 2.  Key Overlaps of the FRTB and the GMS.

Mitigation of the FRTB and the GMS Overlap

To mitigate the impacts of the FRTB and the GMS overlap, the following adjustments could be made to the capital framework:

• Remove the GMS from the SCB. Because the potential losses on trading activities during market stress are already capitalized adequately by the FRTB (via ), the GMS becomes duplicative and redundant. The most effective way of addressing this problem would be to remove the GMS component from the SCB altogether.
• Redesign the GMS to be reasonably plausible. In 2019, SIFMA conducted a careful in-depth study of the empirical plausibility of the range of GMS shocks individually and collectively from the inception of the Comprehensive Capital Analysis and Review (or “CCAR”) through the 2019 CCAR cycle.  The study finds that the severity of single-factor GMS shocks and the correlation assumptions, which underpin the construct of annual GMS shocks, “cannot be empirically justified as reasonably plausible”.^[11]  Making the GMS shocks reasonably plausible would help mitigate some of the impacts of the combined GMS and FRTB requirements.
• Set capital requirements for trading activities based on the maximum of the FRTB and the GMS. Instead of applying the FRTB and the GMS separately and summing up the resulting capital requirements, the capital requirements for trading activities could be determined as the maximum of the FRTB and the GMS.  As a result, the minimum CET1 capital requirements would approximately equal:

Table 3 reports the resulting CET1 capital requirements on the stylized trading portfolio following the maximum of the FRTB and the GMS (or “Max Of”) approach.  Since the FRTB requires higher capital requirements than the GMS – unsurprising as the FRTB is designed to be a stress testing framework, the final capital requirements are set by the FRTB.  Even with the adjustment, the minimum required CET1 capital is nearly 1/3 of the portfolio’s market value.  The RWA density of 419% is roughly on par with the 400% risk weight assigned to “speculative unlisted equity exposures” under the Basel III Endgame – an indication that the FRTB may significantly overcapitalize certain risks and market segments.

 Table 3.  CET1 Capital Requirements as Percentage of The Portfolio’s Market Value.

Conclusion

The FRTB and the GMS are both stress testing frameworks.  The designs of the two frameworks largely overlap, and the risk losses estimated by both frameworks are generally comparable (as illustrated in the stylized equity trading portfolio analyzed in this note).  The result will be a double counting of risks, and thus significantly higher capital requirements that are incommensurate with risks. This double count should be mitigated prior to the implementation of the Basel III Endgame package.  Those mitigations could include removing the GMS from the SCB, given that the market risk losses it was designed to measure are already captured by the FRTB; redesigning the GMS to be reasonably plausible; and setting capital requirements based on the larger of the FRTB and the GMS rather than the sum of the two components.

Additionally, applying the SCB to the Basel 3 Endgame would give rise to overlapping capital requirements for both operational risk losses and CVA risk losses as well, thereby exacerbating the overall impacts of the Basel 3 Endgame.  This is another example of potential double counting that will need to be addressed by regulators as they implement the Basel III Endgame reforms.

Dr. Guowei Zhang is Managing Director and Head of Capital Policy for SIFMA

Dr. Peter Ryan is Managing Director and Head of International Capital Markets and Strategic Initiatives for SIFMA

Mr. Carter McDowell is Managing Director and Associate General Counsel for SIFMA.

---

References

^[1] The adjustments include: taxes, income attributable to minority interest, change in valuation allowance, payments on non-common capital, other comprehensive income, change in adjustments and deductions from regulatory capital, other additions to regulatory capital, and planned common stock dividends (see 2022 Supervisory Stress Test Methodology).

^[2] The GMS scenario is a set of hypothetical shocks to a large set of risk factors reflecting general market stress and heightened uncertainty.  Since the GMS losses are front loaded to Q1, the calculation assumes the Q1 projected stress losses are the largest amongst all 9 Quarters of the supervisory stress test horizon.

^[3] PPNR is “defined as net interest income (interest income minus interest expense) plus noninterest income minus noninterest expense.”  The projected losses due to operational-risk events are included in the projection of PPNR.  Additionally, for banks subject to the GMS, PPNR includes projected trading revenues which include both changes in the market value of trading assets and fees from market-making activities.  The changes in market value of trading assets also flow through the GMS losses.  Though, the Federal Reserve’s “modelling approach for trading revenue limits the influence of severe market events that are separately captured in the global market shock”. (see 2022 Supervisory Stress Test Methodology).

^[4] The RWA density is defined as the ratio of RWA to the leverage ratio exposure measure.  It denotes a bank’s average risk weight per unit of exposure (see https://www.bis.org/publ/work586.pdf).

^[5] The counterparty credit risk capital requirements are calculated according to 12 CFR Part 217 Subpart E §217.132.

^[6] The market risk capital requirements are calculated using the FRTB SBM as prescribed in the Basel 3 Endgame MAR21 (applying the spot shock of 35% and volatility point shock of 77.78%) and FRTB DRC as in the Basel 3 Endgame MAR22 (setting cash equity maturity at 1-year).

^[7] The GMS losses are calculated using the spot shock of 26.3% and the volatility point shock of 26.5 as prescribed by the GMS of 2023 DFAST.

^[8] Effective October 1, 2022, the GSIB surcharge for the 6 non-custodian GSIBs ranges from 1.5% to 3.5%.  The simple average is 2.5%.

^[9] https://www.ecfr.gov/current/title-12/chapter-II/subchapter-A/part-217#217.52

^[10] https://www.bis.org/basel_framework/chapter/CRE/20.htm?inforce=20230101&published=20221208

^[11] https://www.sifma.org/wp-content/uploads/2019/09/SIFMA-GMS-LCD-Study-FINAL.pdf

Related Resources

• Pennsylvania + Wall Blog

  Blog Series on US Bank Capital Requirements