Enterprise Regulatory Capital Framework Rule-Prescribed Leverage Buffer Amount and Credit Risk Transfer

AGENCY:

Federal Housing Finance Agency.

ACTION:

Notice of proposed rulemaking: request for comments.

SUMMARY:

The Federal Housing Finance Agency (FHFA or the Agency) is seeking comments on a notice of proposed rulemaking (proposed rule) that would amend the Enterprise Regulatory Capital Framework (ERCF) by refining the prescribed leverage buffer amount (PLBA or leverage buffer) and credit risk transfer (CRT) securitization framework for the Federal National Mortgage Association (Fannie Mae) and the Federal Home Loan Mortgage Corporation (Freddie Mac, and with Fannie Mae, each an Enterprise). The proposed rule would also make technical corrections to various provisions of the ERCF that was published on December 17, 2020.

DATES:

Comments must be received on or before November 26, 2021.

ADDRESSES:

You may submit your comments on the proposed rule, identified by regulatory information number (RIN) 2590-AB17, by any one of the following methods:

• Agency Website: www.fhfa.gov/open-for-comment-or-input.

• Federal eRulemaking Portal: https://www.regulations.gov. Follow the instructions for submitting comments. If you submit your comment to the Federal eRulemaking Portal, please also send it by email to FHFA at RegComments@fhfa.gov to ensure timely receipt by FHFA. Include the following information in the subject line of your submission: Comments/RIN 2590-AB17.

• Hand Delivered/Courier: The hand delivery address is: Clinton Jones, General Counsel, Attention: Comments/RIN 2590-AB17, Federal Housing Finance Agency, 400 Seventh Street SW, Washington, DC 20219. Deliver the package at the Seventh Street entrance Guard Desk, First Floor, on business days between 9 a.m. and 5 p.m.

• U.S. Mail, United Parcel Service, Federal Express, or Other Mail Service: The mailing address for comments is: Clinton Jones, General Counsel, Attention: Comments/RIN 2590-AB17, Federal Housing Finance Agency, 400 Seventh Street SW, Washington, DC 20219. Please note that all mail sent to FHFA via U.S. Mail is routed through a national irradiation facility, a process that may delay delivery by approximately two weeks. For any time-sensitive correspondence, please plan accordingly.

FOR FURTHER INFORMATION CONTACT:

Andrew Varrieur, Senior Associate Director, Office of Capital Policy, (202) 649-3141, Andrew.Varrieur@fhfa.gov; Christopher Vincent, Senior Financial Analyst, Office of Capital Policy, (202) 649-3685, Christopher.Vincent@fhfa.gov; or James Jordan, Associate General Counsel, Office of General Counsel, (202) 649-3075, James.Jordan@fhfa.gov. These are not toll-free numbers. The telephone number for the Telecommunications Device for the Deaf is (800) 877-8339.

SUPPLEMENTARY INFORMATION:

Comments

FHFA invites comments on all aspects of the proposed rule. Copies of all comments will be posted without change and will include any personal information you provide, such as your name, address, email address, and telephone number, on the FHFA website at https://www.fhfa.gov. In addition, copies of all comments received will be available for examination by the public through the electronic rulemaking docket for this proposed rule also located on the FHFA website.

Table of Contents

I. Introduction

II. Background and Rationale for the Proposed Rule

A. PLBA

B. CRT

III. Proposed Requirements

A. PLBA

B. CRT

C. ERCF Technical Corrections

IV. Paperwork Reduction Act

V. Regulatory Flexibility Act

I. Introduction

FHFA is seeking comments on amendments to the ERCF that would refine the leverage buffer and the risk-based capital treatment for CRT transactions. The proposed amendments would better reflect the risks inherent in the Enterprises' business models and encourage the Enterprises to distribute acquired credit risk to private investors rather than to buy and hold that risk. The dynamic PLBA considered in this proposed rule is intended to achieve FHFA's objective stated in the ERCF of having the Enterprises' leverage capital requirements provide a credible backstop to risk-based capital requirements. Linking the PLBA to the ERCF's stability capital buffer, in conjunction with the proposed rule's refinements to the ERCF's CRT securitization framework, would enhance the safety and soundness of the Enterprises by removing inappropriate capital disincentives to the Enterprises to transfer risk.

FHFA adopted the ERCF on December 17, 2020 (85 FR 82150), with the purpose of implementing a going-concern regulatory capital standard to ensure that each of Fannie Mae and Freddie Mac operates in a safe and sound manner and is positioned to fulfill its statutory mission to provide stability and ongoing assistance to the secondary mortgage market across the economic cycle. In doing so, the ERCF accomplished a statutory requirement that FHFA establish by regulation risk-based capital requirements to safeguard the Enterprises against the risks that arise in the operation and management of their businesses, and implemented a new leverage framework that included both a minimum requirement and a leverage buffer. The ERCF became effective on February 16, 2021.

The ERCF evolved from FHFA's proposals for Enterprise Regulatory Capital Frameworks in 2018 and 2020, which were based on the FHFA Conservatorship Capital Framework (CCF) established in 2017. The ERCF successfully addressed issues identified through the notice and comment process on the pro-cyclicality of the proposed risk-based capital requirements, the quality of Enterprise capital used to meet the capital requirements, and the quantity of capital requirements.

However, FHFA is concerned that certain aspects of the ERCF might create disincentives in the Enterprises' CRT programs that may result in taxpayers bearing excessive undue risk for as long as the Enterprises are in conservatorships and excessive risk to the housing finance market both during and after conservatorships. This concern is heightened by the fact that the Enterprises presently are severely undercapitalized and lack the resources on their own to safely absorb the credit risk associated with their normal operations. In conservatorships, the Enterprises are supported by Senior Preferred Stock Purchase Agreements (PSPAs) between the U.S. Department of the Treasury (the Treasury) and each Enterprise, through FHFA as its conservator. Until recently, the PSPAs significantly limited the Enterprises' ability to hold capital, and only in January 2021 were the upper bounds on retained capital removed. During this period where the Enterprises are building capital, the taxpayers continue to be at heightened risk through potential PSPA draws in the event of a significant stress to the housing sector. The Enterprises have developed their CRT programs over the last several years under FHFA's oversight through guidelines, instructions, strategic plans, and scorecard objectives. FHFA views the transfer of risk, particularly credit risk, to a broad set of investors as an important tool to reduce taxpayer exposure to the risks posed by the Enterprises and to mitigate systemic risk caused by the size and monoline nature of the Enterprises' businesses. If the Enterprises were to substantially shrink their risk transfer programs for an extended period, either in response to regulatory policies or macroeconomic conditions, potential taxpayer exposure and systemic risk may increase as a result.

The refinements in this proposal would lessen the potential deterrents to Enterprise risk transfer. Specifically, the proposed rule would amend the ERCF to:

• Replace the fixed PLBA equal to 1.5 percent of an Enterprise's adjusted total assets with a dynamic PLBA equal to 50 percent of the Enterprise's stability capital buffer as calculated in accordance with 12 CFR 1240.400;
• Replace the prudential floor of 10 percent on the risk weight assigned to any retained CRT exposure with a prudential floor of 5 percent on the risk weight assigned to any retained CRT exposure; and
• Remove the requirement that an Enterprise must apply an overall effectiveness adjustment to its retained CRT exposures in accordance with the ERCF's securitization framework in 12 CFR 1240.44(f) and (i).

The proposed rule would also make technical corrections to various provisions of the ERCF that was published on December 17, 2020.

The PSPAs between the Treasury and each Enterprise, through FHFA as its conservator, as amended by letter agreements executed by the parties on January 14, 2021, include a covenant at section 5.15 which states: “[The Enterprise] shall comply with the Enterprise Regulatory Capital Framework [published in the Federal Register at 85 FR 82150 on December 17, 2020] disregarding any subsequent amendment or other modifications to that rule.” Modifying that covenant will require agreement between the Treasury and FHFA under section 6.3 of the PSPAs.

II. Background and Rationale for the Proposed Rule

A. PLBA

Background

The ERCF requires an Enterprise to maintain a leverage ratio of tier 1 capital to adjusted total assets of at least 2.5 percent. In addition, to avoid limits on capital distributions and discretionary bonus payments, an Enterprise must also maintain a fixed tier 1 capital PLBA equal to at least 1.5 percent of adjusted total assets.

The primary purpose of the combined leverage requirement and PLBA is to serve as a non-risk-based supplementary measure that provides a credible backstop to the combined risk-based capital requirements and prescribed capital conservation buffer amount (PCCBA), where the PCCBA comprises the stability capital buffer, the stress capital buffer, and the countercyclical capital buffer. This type of simple, transparent, and independent measure of risk provides an important safeguard against model risk and measurement error in the risk-based capital requirements and acquisition strategies of the Enterprises. FHFA's rationale for the leverage requirement and buffer is consistent with that of U.S. and international banking regulators, although the size of each regulator's leverage buffer varies by regulatory regime. In the U.S., large banking organizations must maintain an enhanced supplementary leverage ratio (eSLR) of 2 percent of total leverage exposure on top of their 3 percent supplementary leverage ratio (SLR) to avoid restrictions on distributions and discretionary bonuses. Internationally, systemically important banks are required to hold a leverage buffer that varies by the bank's systemic importance.

The Enterprises are chartered to fulfill a countercyclical role in the housing finance market. The COVID-19 pandemic, while unique and not the basis for this proposed rule, has effectively illustrated why a dynamic leverage buffer may be appropriate for the Enterprises. During the pandemic, as many mortgage market participants pulled back from the market due to capital and liquidity constraints, the Enterprises stepped in to fulfill their countercyclical role, leading to greater reliance on Enterprise execution for conforming mortgages. This, combined with the Board of Governors of the Federal Reserve System's (Federal Reserve) monthly purchases of $40 billion in Agency mortgage-backed securities (MBS), caused the Enterprises' balance sheets to expand considerably. As a result, the PLBA represents an increasingly large component of the Enterprises' capital requirements and capital buffers relative to when FHFA calibrated the PLBA in 2019. In addition, the combined leverage requirement and PLBA exceeds the combined risk-based capital requirement and PCCBA at some level for both Enterprises. The leverage requirement and current PLBA are based on adjusted total assets, which is a relatively stable measure over time. Given this calibration, FHFA expects the current relationships between leverage and risk-based capital at the Enterprises will continue for the foreseeable future. When leverage capital is consistently the binding capital constraint, it provides an incentive for an institution to increase risk taking because taking on more risk is not reflected in commensurately higher capital requirements, while greater risk may generate greater returns. When leverage capital sufficiently exceeds risk-based capital, high risk exposures and low risk exposures have the same capital requirements, so an Enterprise has an incentive to acquire higher-risk, higher-yielding mortgages, all else equal.

As of March 31, 2021, Fannie Mae's tier 1 leverage capital requirement plus PLBA of 4 percent was the binding capital constraint relative to their estimated common equity tier 1 (CET1) capital requirement plus PCCBA of 3.3 percent and their estimated tier 1 risk-based capital requirement plus PCCBA of 3.8 percent, all relative to adjusted total assets. Fannie Mae's estimated adjusted total capital requirement plus PCCBA of 4.5 percent (relative to adjusted total assets) was their only risk-based capital requirement that exceeded their leverage capital requirement plus PLBA. At Freddie Mac, the leverage capital requirement plus PLBA was the binding capital constraint relative to every risk-based capital metric. Freddie Mac's estimated CET1 capital requirement plus PCCBA of 2.8 percent, estimated tier 1 risk-based capital requirement plus PCCBA of 3.2 percent, and estimated adjusted total capital requirement plus PCCBA of 3.8 percent, all relative to adjusted total assets, were each smaller than their tier 1 leverage capital requirement plus PLBA of 4 percent.

For the Enterprises combined, the tier 1 leverage capital requirement plus PLBA was approximately 12 percent larger than the combined tier 1 risk-based capital requirement plus PCCBA (relative to adjusted total assets) as of March 31, 2021. This excess of total leverage capital over tier 1 risk-based capital has grown from 10 percent when FHFA calibrated the ERCF near the end of 2019—a 20 percent increase in only two years. The leverage requirement and PLBA are met with tier 1 capital, while the tier 1 risk-based capital requirement and PCCBA are met with tier 1 capital and CET1 capital respectively, which allows for the most direct comparison of leverage capital to risk-based capital. In addition, CET1 capital and tier 1 capital represent the highest quality and second-highest quality forms of capital, respectively, so examining the binding nature of the tier 1 leverage requirement relative to the tier 1 risk-based capital requirement is prudent when considering the safety and soundness of the Enterprises.

Rationale for Revisiting the PLBA

The primary purpose of the ERCF's leverage requirement and PLBA is to serve as a credible backstop to the risk-based capital requirements and risk-based capital buffers. This is consistent with the stated purpose of the SLR and eSLR in the U.S. banking framework. FHFA is proposing a recalibration of the PLBA because a leverage ratio that exceeds risk-based capital requirements throughout the economic cycle could lead to undesirable outcomes at the Enterprises, including promoting risk-taking and creating disincentives for CRT and other forms of risk transfer. Evolutions in the international and U.S. banking frameworks and public comments on FHFA's 2020 re-proposed capital rule support the proposed PLBA recalibration.

Financial regulators and policymakers have consistently investigated ways to lower the quantity of leverage required for banks, with a specific focus on the SLR and eSLR. In the U.S., banking regulators require global systemically important banks (GSIBs) to hold tier 1 capital in excess of 5 percent of total on-and-off balance sheet assets (measured using total leverage exposure, which is comparable to adjusted total assets at the Enterprises) consisting of a 3 percent minimum SLR and a 2 percent leverage buffer (the eSLR). Internationally, Basel III standards require systemically important banks to hold a tier 1 capital leverage ratio buffer in excess of a 3 percent leverage requirement equal to 50 percent of a GSIB's higher loss-absorbency risk-based requirements. This dynamic leverage buffer tailors leverage requirements to business activities and risk profiles, aiming to retain a meaningful calibration of leverage ratio standards while not discouraging firms from participating in low-risk activities. The higher loss-absorbency risk-based requirements is a measure similar to the U.S. banking framework's GSIB surcharge, which varies in size depending on a bank's systemic importance, as measured using a bank's size, interconnectedness, cross-jurisdictional activity, substitutability, complexity, and use of short-term wholesale funding. In April 2018, the Federal Reserve and the Office of the Comptroller of the Currency (OCC) released a similar proposal that would tailor the eSLR for GSIBs by modifying the fixed 2 percent eSLR buffer to equal one half of each firm's GSIB capital surcharge. This proposal would have a significant impact on the leverage ratios of U.S. GSIBs, decreasing the fixed 2 percent eSLR to, on a median basis, approximately 1.25 percent.

In addition, there have been various proposals in recent years from the U.S. Department of the Treasury and the U.S. Congress for a more targeted approach to removing certain items from total leverage exposure to address the negative externalities the SLR and eSLR requirements may have on market liquidity and low-risk assets. One such proposal included adjustments to the calibration of the eSLR and the leverage exposure calculation to exclude from the denominator of total leverage exposure cash on deposit with central banks, U.S. Treasury securities, and initial margin for centrally cleared derivatives. The Economic Growth, Regulatory Relief, and Consumer Protection Act of 2018 adopted part of the Treasury's recommendation by relaxing the leverage ratio for “custodial banks” by removing funds held at central banks from the leverage ratio's denominator. Furthermore, as FHFA did in the ERCF, there is precedent for bank regulators tailoring the leverage ratio to conform to an institution's unique circumstances. As an example, in 2015, the Federal Reserve reduced the eSLR requirement for GE Capital from 5 percent to 4 percent when it was designated a nonbank systemically important financial institution (SIFI) by the Financial Stability Oversight Council (FSOC).

The regulatory focus on reevaluating bank leverage ratio requirements has sharpened further during the COVID-19 pandemic. In March 2020, to stabilize dislocations in the market for U.S. Treasuries as a result of the pandemic, the Federal Reserve temporarily modified the SLR to exclude U.S. Treasury securities and central bank reserves from the leverage calculation. In March 2021, the Federal Reserve allowed this temporary relief to expire as the strains in the Treasury market resulting from COVID-19 had eased, but acknowledged it “may need to address the current design and calibration of the SLR over time to prevent strains from developing that could both constrain economic growth and undermine financial stability.” After allowing the temporary relief to expire, the leverage ratio became the binding capital constraint for JPMorgan Chase & Co., the largest GSIB. The Federal Reserve also stated that “to ensure that the SLR—which was established in 2014 as an additional capital requirement—remains effective in an environment of higher reserves, the Board will soon be inviting public comment on several potential SLR modifications.” Further, members of the Federal Reserve's Board of Governors recently confirmed that the Board is looking to make changes to the leverage framework.

The current circumstances in which tier 1 leverage capital requirements are binding for both Fannie Mae and Freddie Mac may lead to perverse incentives that have the Enterprises take on more risk than is prudent. By treating all risk similarly, a binding leverage ratio driven by the PLBA may incentivize risk-taking because the capital requirement would be the same for high-risk and low-risk loans. In addition, the Enterprises would have no capital incentive to transfer risk to achieve a risk-based capital requirement lower than their leverage requirement. However, when risk-based capital requirements are higher than leverage capital requirements, CRT represents a viable way to both lower risk at the Enterprises and to shrink the gap between capital requirements and available capital, promoting safety and soundness. These were pressing issues to commenters when FHFA re-proposed its Enterprise capital rule in 2020.

Prior to finalizing the ERCF, FHFA received a significant number of public comments on FHFA's proposed PLBA. Some commenters recommended a leverage buffer smaller than was proposed (both with and without corresponding recommendations for the leverage requirement). Most commenters focused on the size of the combined leverage requirement and PLBA as a single 4 percent leverage ratio. Most of those commenters recommended a combined leverage ratio smaller than 4 percent. Some suggested that 4 percent overstates potential risk in the Enterprises' books because FHFA's ERCF calibration was based on historical losses without adjusting for prevailing portfolio composition. That is, given that the Enterprises are no longer permitted to acquire many of the loans that precipitated the 2008 financial crisis, such as Alt-A loans and option ARMs, a leverage ratio corresponding to the Enterprises' current acquisition profile should not be calibrated to losses involving such loans. Relatedly, commenters suggested that concerns the Enterprises may again loosen underwriting standards have been addressed in several ways, including through post-crisis statutory and regulatory changes such as the Qualified Mortgage and Ability-to-Repay rule, which would require a statutory change and/or a notice of proposed rulemaking followed by a period of public comment in order to modify. In addition, commenters argued that these concerns were further addressed through post-crisis improvements in risk management and improved loss-mitigation capabilities, incorporation of automated tools into the underwriting process to verify the accuracy of data and detect loan manufacturing defects, tightened counterparty risk management, and improvements in fraud prevention.

Commenters also suggested that the Enterprises' recent Dodd-Frank Act Stress Tests (DFAST) results do not support a 4 percent leverage ratio. Commenters' analysis at the time indicated that 4 percent leverage would be between four and thirteen times DFAST losses, depending on which scenario was being compared. Commenters suggested this multiple was excessive. In addition, some commenters viewed the PLBA as being duplicative of other ERCF adjustments and buffers that also were designed to mitigate model and related risk. Finally, as stated above, many commenters stated that a binding leverage ratio would be a disincentive for CRT and encourage the Enterprises to take on more risk.

B. CRT

Background

The Enterprises' core businesses reflect the acquisition of mortgages from financial institutions and the bundling of those mortgages into collateral for MBS. The Enterprises sell to investors part of the cash flows that stem from the mortgages underlying the MBS. The Enterprises guarantee the principal and interest payments to investors and collect a guarantee fee from their sellers.

Mortgage exposures typically carry both interest rate and credit risk. In general, the Enterprises transfer mortgage interest rate risk and retain and manage mortgage credit risk. The interest rate risk on securitized mortgages is transferred to investors through MBS sales. The Enterprises' principal and interest guarantee helps to create a liquid and efficient MBS market. It also limits the credit risk assumed by MBS investors, except for an investor's counterparty exposure to the Enterprises. Credit risk can be broadly separated into expected losses and unexpected losses, as determined by a credit model. The Enterprises rely on guarantee fees to cover expected losses and, absent CRT, equity capital to cover unexpected losses.

In its role as conservator, FHFA established a goal of reducing taxpayer risk exposure to the credit guarantees extended by the Enterprises. To accomplish this objective, FHFA used its conservatorship strategic plans and scorecards to encourage the Enterprises to transfer credit risk to the private sector. In 2012, FHFA's Strategic Plan for Enterprise Conservatorships proposed the use of loss sharing agreements to reduce the credit risk incurred by the Enterprises. The 2013 Conservatorship Scorecard required each Enterprise to “demonstrate the viability of multiple types of [credit] risk transfer transactions” on single-family loans. The Enterprises first implemented their CRT programs that same year and have since transferred to private investors a substantial amount of the credit risk of new acquisitions the Enterprises assume for loans in targeted loan categories. The programs have become a core part of the Enterprises' single-family credit guarantee business and include or have included CRTs via capital markets issuances (both corporate debt and bankruptcy remote trust structures), insurance/reinsurance transactions, senior/subordinate transactions, and a variety of lender collateralized recourse transactions.

The 2014 Strategic Plan for the Conservatorships of Fannie Mae and Freddie Mac emphasized the desirability of greater use of CRT in the future. Additionally, the 2014 and 2015 Conservatorship Scorecards set more ambitious CRT performance goals for each Enterprise. Since that time, the Conservatorship Scorecards have included various goals to ensure the continued use of CRT as a means of reducing risk exposure to taxpayers. For example, the 2016 through 2019 Conservatorship Scorecards established an objective for the Enterprises to transfer a meaningful portion of credit risk on at least 90 percent of the unpaid principal balance (UPB) of their acquired single-family mortgage loans targeted for credit risk transfer. Targeted loans include fixed-rate, non-HARP loans with terms over 20 years and loan-to-value (LTV) ratios above 60 percent. Such loans represent a substantial amount of the credit risk associated with all new loan acquisitions.

From the beginning of the Enterprises' single-family CRT programs in 2013 through the end of 2020, Fannie Mae and Freddie Mac have transferred a portion of credit risk on approximately $4.1 trillion of UPB, with a combined risk-in-force (RIF) of about $137 billion, or 3.3 percent of UPB.

The Enterprises' CRT programs have evolved over time in response to changing macroeconomic conditions, loan acquisition risk profiles, and views of expected and unexpected losses. However, across the different types of CRT vehicles, the basic transaction is the same: An Enterprise pays private market participants to assume credit risk in a severe stress scenario on mortgages the Enterprise guarantees, where the severe stress scenario is generally comparable to the 2008 global financial crisis. Further, to ensure alignment of interests with investors, the Enterprises retain at least 5 percent of the risk exposure sold in their CRT transactions. This is referred to as vertical risk retention.

The Enterprises have developed their various CRT products in order to meet certain program goals established by FHFA in 2012. Among these goals is that CRT transactions should be economically sensible, repeatable, scalable, and structured to not disrupt the efficient operation of the “To Be Announced” (TBA) market (which provides the market with benefits including allowing borrowers to lock in rates in advance of closing). The widespread use of TBA trading has contributed significantly to the liquidity and efficiency of the secondary market for single-class MBS. A misconception is that “economically sensible” implies low-cost on an absolute basis. However, the costs of CRT should be evaluated relative to the cost of equity capital needed to self-insure the risk. To be economically sensible, an Enterprise should consider executing CRT transactions when the cost to the Enterprise for transferring the credit risk does not meaningfully exceed the cost to the Enterprise of self-insuring the credit risk being transferred. Market conditions in addition to a transaction's cost and structure ultimately determine a CRT's relative profitability, but if CRT premium payments are low relative to the capital reduction provided by the CRT, then the Enterprise has the opportunity to execute economically sensible CRT transactions, and CRT may provide taxpayer protection at a lower cost than equity capital.

A further goal was to develop different types of products to provide for the broadest possible access to investors with the expectation that at least some of those investors would remain in the market through all phases of a housing price cycle. Since the inception of the programs in 2013, the types of single-family CRT transactions have included structured capital markets issuances known as Structured Agency Credit Risk (STACR) for Freddie Mac and Connecticut Avenue Securities (CAS) for Fannie Mae, insurance/reinsurance transactions known as Agency Credit Insurance Structure (ACIS) for Freddie Mac and Credit Insurance Risk Transfer (CIRT) for Fannie Mae, front-end lender risk sharing transactions, and senior/subordinate transactions.

Most of the RIF has come from capital markets issuances (STACR and CAS). These securities were initially issued as direct debt obligations of each Enterprise; however, in 2018, both Enterprises transitioned their capital markets CRT issuances to a Trust structure with the notes being issued by a bankruptcy remote trust created for each individual CAS or STACR transaction. The proceeds from the sale of the notes are deposited into the bankruptcy remote trust and there is no direct counterparty exposure to the Enterprises for investors. By implementing the Trust structure, the Enterprises are now able to benefit from insurance accounting treatment for their capital markets CRT transactions. Insurance accounting treatment aligns the timing of the recognition of credit losses with CRT loss recoveries. Under the previous corporate debt structure, there was a significant timing mismatch between the recognition of losses and recoveries as the CRT benefit could not be recognized until the underlying delinquent mortgage loan had progressed through the often-lengthy disposition process.

In addition, both Fannie Mae and Freddie Mac now engage in CRT offerings under which the securities are issued by a third-party bankruptcy-remote trust that also qualifies as a Real Estate Mortgage Investment Conduit (REMIC). The transition of the capital markets CRT programs to the REMIC Trust structure was a collaborative, long-term effort between Fannie Mae, Freddie Mac, and FHFA. The REMIC Trust structure, like the trust structure described above, eliminates accounting mismatches associated with prior direct debt issuance transactions and limits investor exposure to Enterprise counterparty risk. Additionally, the REMIC structure is often more attractive to domestic Real Estate Investment Trusts (REITs) and foreign investors.

After exceptionally strong issuance volume between 2013 and the first quarter of 2020, neither Enterprise entered into new CRT transactions in the second quarter of 2020 due to the adverse market conditions stemming from the COVID-19 pandemic. However, Freddie Mac returned to the CRT capital markets and insurance/reinsurance market during the third quarter of 2020, executing nine transactions in the second half of the year. In contrast, and despite improved market conditions, Fannie Mae continued to pause issuance of new CRT transactions to evaluate the costs and benefits of CRT, including the capital relief provided by the transactions and the market conditions, as well as their overall capital requirements, risk appetite, and business plan. Overall, while down from its peak in 2019, total CRT volume in 2020 remained strong and exceeded 2018 volume despite the extreme and unforeseen difficulties arising from the COVID-19 pandemic. In 2021, both Enterprises are considering potential changes to their CRT programs to optimize risk transfer and capital relief under the ERCF.

Multifamily CRT

Even before the formalization of the single-family CRT programs, risk transfer to the private sector had long been an integral part of the multifamily business models at the Enterprises. Freddie Mac has traditionally focused on senior/subordinate structures via capital market transactions largely through its K-Deal platform. Fannie Mae has traditionally focused on pro-rata risk sharing directly with lenders through its Delegated Underwriting and Servicing (DUS) program. As the single-family CRT programs evolved and grew, the Enterprises worked to expand their existing multifamily risk transfer models to include structures similar to those of the single-family businesses.

Fannie Mae issued its first multifamily reinsurance transaction in 2016, the Multifamily Credit Insurance Risk Transfer (MCIRT), which was based on the framework of the existing single-family reinsurance (CIRT) transactions, where the Enterprise purchases insurance coverage underwritten by a group of insurers/reinsurers. Fannie Mae uses MCIRT to transfer credit risk on multifamily loan acquisitions with up to $30 million in UPB. Since the first transaction in 2016, Fannie Mae's MCIRT has become programmatic with a total of eight transactions executed. These transactions provide combined RIF of $1.9 billion on a total of $81 billion (as measured at time of deal inception) of Fannie Mae's multifamily loan acquisitions.

In 2018, Freddie Mac introduced its Multifamily Credit Insurance Pool (MCIP) program to transfer additional credit risk on its multifamily loan acquisitions to the reinsurance market. In the MCIP structure, as in Fannie Mae's MCIRT program, Freddie Mac purchases insurance coverage underwritten by a group of insurers/reinsurers that generally provide first loss and/or mezzanine loss credit protection. These transactions are also similar in structure to the single-family ACIS transactions.

In 2019, Fannie Mae expanded its multifamily CRT program by executing its first Multifamily Connecticut Avenue Securities (MCAS) CRT transaction which is based on the framework for Fannie Mae's existing single-family CAS execution. Fannie Mae uses MCAS to transfer credit risk on multifamily loans with UPBs greater than $30 million. However, this new product allowed Fannie Mae to reach a multifamily CRT investor base outside of the reinsurance industry. Fannie Mae has executed a total of two MCAS transactions which provide combined RIF of $0.9 billion on a total of $29 billion (as measured at time of deal inception) of Fannie Mae's multifamily loan acquisitions.

Freddie Mac's multifamily capital markets CRT program began with the issuance of three fixed-rate Multifamily Structured Credit Risk (MSCR) notes in 2016 and 2017 (as a separate offering from the K-deal program). These legacy MSCR notes use a fixed severity structure like early single-family CRTs and are unsecured and unguaranteed corporate debt obligations that transfer to third parties a portion of the credit risk of the multifamily loans underlying certain consolidated other securitizations and other mortgage-related guarantees. SCR Notes are synthetic instruments whose cash flows are driven by the performance of a pool of multifamily reference obligations, instead of actual collateral tied to a trust in a typical securitization such as K-Deals. In 2021, Freddie Mac's MSCR program transitioned to an actual loss/Trust structure, and coupon payments are now floating rate, indexed to the Secured Overnight Financing Rate (SOFR). These features align with the current single-family STACR CRT product.

CRT in the ERCF

The Enterprises manage mortgage credit risk through their underwriting systems, guarantee fee revenues, and CRT programs. The ERCF reflects the Enterprises' management of mortgage credit risk by allowing the Enterprises to reduce their credit risk-weighted assets for eligible CRT. However, the ERCF's treatment of CRT includes various components that limit the amount of capital relief provided by CRTs to ensure that all exposures retained by an Enterprise are meaningfully capitalized. Dollar-for-dollar capital relief should not be expected given that CRT transactions introduce counterparty and structural risk, and CRT has not yet been tested through a full economic cycle.

Under the ERCF, an Enterprise determines the capital treatment for eligible CRT by assigning risk weights to retained CRT exposures. The rule includes: (i) Operational criteria to mitigate the risk that the terms or structure of the CRT would not be effective in transferring credit risk; (ii) a tranche-specific prudential risk weight floor of 10 percent; and (iii) adjustments to reflect loss sharing effectiveness, loss-timing effectiveness, and a dynamic overall effectiveness adjustment meant to capture the differences between CRT and regulatory capital.

The operational criteria, risk weight floor, and effectiveness adjustments limit capital relief from CRT. The operational criteria act as a gateway by setting minimum criteria for potential CRT credit risk capital relief. The 10 percent risk weight floor adds minimum capital requirements to all retained CRT exposures, no matter how remote the credit risk. The effectiveness adjustments reduce the risk-weighted assets of transferred CRT tranches, thereby reducing the capital relief afforded by the CRT. Of these three elements included in the ERCF's CRT treatment, the risk weight floor drives the majority of the reduction in credit risk capital relief due to the relative size of the low-risk CRT exposures the Enterprises generally retain. For example, the stylized CRT transaction in FHFA's 2020 re-proposed capital rule showed capital relief of 38 percent due to the CRT. However, absent the risk weight floor on retained exposures, capital relief would have been approximately 66 percent.

Rationale for Revisiting the ERCF's CRT Treatment

CRT is an effective mechanism for distributing credit risk across a broad mix of investors and has become an integral part of the Enterprises' business models. FHFA is proposing amendments to the ERCF that would revise the CRT securitization framework for several reasons.

First, if an Enterprise retained every tranche of a CRT, its post-CRT credit risk capital requirement for the CRT exposures would be higher than its pre-CRT credit risk capital requirements for the underlying mortgage exposures due to the structural and modeling risk of the CRT itself. The capital relief afforded by the ERCF CRT securitization framework more than offsets this so-called securitization penalty, but within the securitization framework, potential capital relief is limited by adjustments that reflect various ways a CRT might be less than fully effective at transferring risk. Increasing the capital relief for CRT by reducing these effectiveness adjustments could improve the safety and soundness of each Enterprise by encouraging the transfer of risk so that each Enterprise can fulfill its statutory mission to provide stability and ongoing assistance to the secondary mortgage market across the economic cycle.

Second, FHFA believes that part of the process to responsibly end the conservatorships of the Enterprises includes the transfer of a portion of the Enterprises' credit risk to private markets. Such activity allows the Enterprises to maintain their core businesses, fulfill their statutory missions, and grow organically while simultaneously shedding risk that could otherwise prevent them from accomplishing these goals. It is possible that in the absence of risk transfer, required capital may increase faster than retained earnings and the Enterprises may therefore grow farther from achieving capital adequacy and exiting their conservatorships. To the extent that the earnings expenses of CRT are smaller than the capital relief provided by CRT, executing CRT would help alleviate this issue.

Third, a revised risk-based capital treatment for CRT could facilitate regulatory capital planning in furtherance of the safety and soundness of the Enterprises and their countercyclical mission. The Enterprises' CRT programs, which FHFA has in the past required to cover 90 percent of the UPB of target loans (generally those with an LTV greater than 60 percent and a loan term greater than 20 years), help facilitate the continued acquisition of higher risk loans throughout the economic cycle due to capital relief afforded to risk transfer. In addition, as adopted, the ERCF's CRT framework does little to complement the single-family countercyclical adjustment. Revised CRT incentives could, for example, help to align the issuance of CRT with changes in the countercyclical adjustment.

Fourth, prior to finalizing the ERCF, FHFA received a significant number of comments on FHFA's proposed approach to CRT. Many commenters expressed the view that CRT is an effective means by which to transfer risk to private markets, protect taxpayers, and stabilize the Enterprises and the housing finance market more generally. Consequently, most of these commenters suggested that the proposed treatment of CRTs was too punitive and would imprudently discourage CRTs. Many commenters criticized the 10 percent risk weight floor and the overall effectiveness adjustment, arguing that FHFA's proposed policy choices would unduly decrease the capital relief provided by CRT and reduce the Enterprises' incentives to engage in CRT. FHFA nevertheless adopted the risk weight floor as proposed, citing a belief that 10 percent represents an appropriate capitalization for the credit risk in these retained risks and a favorable comparison to the U.S. bank regulatory framework. To account for the fact that CRT does not provide the same loss-absorbing capacity as equity financing and to reduce the extent to which the proposed 10 percent adjustment may lead to more regulatory capital than is necessary to ensure safety and soundness, FHFA adopted a modified overall effectiveness adjustment that starts at 10 percent and decreases with an exposure's credit risk.

FHFA also received comments on the interaction of CRTs and the leverage ratio requirement. Several commenters expressed concern about the potential adverse impact of a binding leverage requirement on CRTs. Specifically, commenters indicated that a binding leverage requirement would provide no incentive for the Enterprises to lower their risk-based capital requirements and therefore would disincentivize CRTs, which could lead the Enterprises to reduce or halt their CRT programs and increase the risks held in portfolio.

III. Proposed Requirements

A. PLBA

The proposed rule would amend the ERCF by replacing the fixed PLBA equal to 1.5 percent of an Enterprise's adjusted total assets with a dynamic PLBA equal to 50 percent of the Enterprise's stability capital buffer as calculated in accordance with 12 CFR 1240.400.

The Enterprise-specific stability capital buffer was designed to mitigate risk to national housing finance markets by requiring a larger Enterprise to maintain a larger cushion of high-quality capital to reduce the likelihood of a large Enterprise's failure and preclude the potential impact a failure would have on the national housing finance markets. Such a buffer creates incentives for each Enterprise to reduce its housing finance market stability risk by curbing its market share and growth in ordinary times, preserving room for a larger role during a period of financial stress, and may offset the funding advantage that an Enterprise might have on account of being perceived as “too big to fail.” The stability capital buffer is based on a market share approach, where each Enterprise's stability capital buffer is directly related to its relative share of total residential mortgage debt outstanding that exceeds a threshold of 5 percent market share. The stability capital buffer, expressed as a percent of adjusted total assets, increases by 5 basis points for each percentage point of market share exceeding that threshold.

The proposed rule would replace the fixed 1.5 percent PLBA with a dynamic leverage buffer determined annually and tied to the stability capital buffer. The stability capital buffer is an effective proxy for the U.S. banking framework's GSIB capital surcharge and the Basel higher loss-absorbency risk-based requirement as it is designed to address the predominant threat an Enterprise poses to national housing markets—its size. Thus, in a manner similar to the U.S. banking regulators' proposal to set the eSLR buffer to one-half of the GSIB surcharge, an Enterprise's PLBA would equal one-half of its stability capital buffer under the proposed rule. Under the amended rule, as shown in the figure below and as of March 31, 2021, Fannie Mae's PLBA would decrease from approximately $62 billion, or 1.5 percent of the prior quarter's adjusted total assets, to approximately $23 billion, or 0.53 percent of adjusted total assets. Freddie Mac's PLBA would similarly decrease from $46 billion, or 1.5 percent of the prior quarter's adjusted total assets, to approximately $11 billion, or 0.35 percent of adjusted total assets.

There are several benefits of the proposed approach. First, decreasing the PLBA to the point where risk-based capital is the binding capital constraint at the Enterprises would promote safety and soundness by lessening the likelihood that an Enterprise has an incentive to take on more risk in a capital optimization strategy. Setting the PLBA to 50 percent of the stability capital buffer would not guarantee that leverage capital is never binding, but it would restore leverage capital to a position of a credible backstop rather than the binding capital constraint for the foreseeable future. This would allow the other aspects of the ERCF, namely the risk-based capital requirements, including the single-family countercyclical adjustment, to work as intended. For example, the single-family countercyclical adjustment works by increasing risk-based capital requirements to largely offset capital benefits driven by house price appreciation. This effective tool alleviates concerns that risk-based capital will artificially decline with increasing property values, thereby lessening the need for a consistently binding leverage capital framework. An unduly high leverage requirement dampens the functionality of the single-family countercyclical adjustment.

The ERCF does not currently contain an exposure-level method to mitigate the pro-cyclicality of the credit risk capital requirements for multifamily mortgage exposures. FHFA has, in two notices of proposed rulemaking, indicated it would like to implement such an adjustment, and has twice sought recommendations for potential approaches. Although FHFA has received numerous suggestions for a multifamily countercyclical adjustment, most have relied on proprietary data or indices to some extent. FHFA is again expressing its desire to include a multifamily countercyclical adjustment in the ERCF that is not reliant on proprietary information and is seeking input on how that adjustment should be constructed.

Question 1: What approach that relies only on non-proprietary data or indices should FHFA consider to mitigate the pro-cyclicality of the credit risk capital requirements for multifamily mortgage exposures?

Second, the proposed rule's PLBA will encourage the Enterprises to transfer risk rather than to buy and hold risk. Leverage capital requirements and buffers treat each dollar of exposure equally and incentivize risk-taking to the point where risk-based capital equals leverage capital. At the Enterprises, seasoned portfolios generally require less capital than new acquisitions because risk determinants such as the loan-to-value ratio typically improve as mortgage loans age. Therefore, higher leverage requirements incentivize an Enterprise to acquire riskier, higher-yielding exposures and then to hold that risk so that risk-based capital on the book approximates leverage capital on the book. A lower PLBA directly encourages a risk transfer strategy by lowering the long-run risk-based capital target for an Enterprise's book. Buying and holding risky assets would likely no longer be optimal from a capital perspective if the risk-based capital on an Enterprise's seasoned portfolio exceeded leverage capital.

Third, a leverage framework with a dynamic PLBA that grows and shrinks as an Enterprise grows and shrinks, respectively, would function as a better backstop to a risk-based capital framework that includes a systemic risk component such as the stability capital buffer. In the 2020 ERCF notice of proposed rulemaking, FHFA argued that a larger Enterprise's default would pose a greater threat to the national housing finance markets than a smaller Enterprise's default. As a result, a probability of default that might be acceptable for a smaller Enterprise could be unacceptably high for a larger Enterprise, necessitating the need for an Enterprise-specific stability capital buffer based on size. For similar reasons, a smaller leverage buffer may not be appropriate for a larger institution, and a larger leverage buffer may not be appropriate for a smaller institution. Therefore, a leverage buffer that adjusts with the stability capital buffer would help resolve this type of inconsistency and allow the leverage capital framework to better serve as a credible backstop to the risk-based capital framework.

Fourth, a dynamic PLBA that is tied to the stability capital buffer would further align the ERCF with Basel III standards. Internationally, GSIBs are required to hold a leverage buffer equal to 50 percent of their higher loss-absorbency risk-based requirements—a measure akin to the GSIB surcharge in the U.S. banking framework. FHFA believes that tailoring an Enterprise's leverage ratio to its business activities and risk profile, to the extent that these characteristics are related to an Enterprise's share of the residential mortgage market, will allow for leverage to remain a credible backstop to risk-based capital without discouraging the Enterprise from participating in low-risk activities.

Question 2: Is the proposed PLBA appropriately formulated? What adjustments, if any, would you recommend?

Question 3: Is the PLBA necessary for the ERCF's leverage framework to be considered a credible backstop to the risk-based capital requirements and PCCBA?

Question 4: In light of the proposed changes to the PLBA and the CRT securitization framework, is the prudential risk weight floor of 20 percent on single-family and multifamily mortgage exposures appropriately calibrated? What adjustments, if any, would you recommend?

B. CRT

CRT Risk Weight Floor

The proposed rule would replace the prudential floor of 10 percent on the risk weight assigned to any retained CRT exposure with a prudential floor of 5 percent on the risk weight assigned to any retained CRT exposure.

The prudential risk weight floor plays an important role in the ERCF securitization framework. The risk weight floor is designed to mitigate certain risks and limitations associated with underlying historical data and models, including that crisis-era losses at the Enterprises were mitigated by federal government support that may not be repeated during the next crisis and that potential material risks are not assigned a risk-based capital requirement. In addition, banking agencies believe requiring more capital on a transaction-wide basis than would be required if the underlying assets had not been securitized is important in reducing the likelihood of regulatory capital arbitrage through securitizations. CRT may pose similar structural risks that merit a departure from capital neutrality. Therefore, the ERCF's risk weight floor helps mitigate the model risk associated with the calibration of the credit risk capital requirements of the underlying exposures and the model risk posed by the calibration of the adjustments for loss-timing and counterparty risks.

In sizing the 10 percent prudential risk weight floor, FHFA sought to promote consistency with the U.S. banking framework and strike an appropriate balance between permitting CRT while also mitigating the safety and soundness, mission, and housing stability risk that might be posed by some CRT. FHFA continues to believe that an Enterprise retains credit risk to the extent it retains CRT exposures and that such risk should be appropriately capitalized. There is the risk that the structuring of some CRT is driven by regulatory arbitrage, with an Enterprise focused on CRT structures that obtain capital relief that is disproportionate to the modeled credit risk actually transferred. There is also the risk that a CRT will not perform as expected in transferring credit risk to third parties, perhaps because a court will not enforce the contractual terms of the CRT structure as expected. Because CRT tranches, even senior CRT tranches, are not risk-free, each Enterprise should maintain regulatory capital to absorb losses on those retained CRT exposures. However, FHFA believes that the current CRT risk weight floor may not achieve the proper balance between permitting CRT and safety and soundness.

As currently calibrated, the 10 percent floor on the risk weight assigned to a retained CRT exposure unduly decreases the capital relief provided by CRT and reduces an Enterprise's incentives to engage in CRT. This occurs in part because the aggregate credit risk capital required for a retained CRT exposure is often greater than the aggregate credit risk capital required for the underlying exposures, especially when the credit risk capital requirements on the underlying whole loans and guarantees are low or the CRT is seasoned. Decreasing the CRT risk weight floor to 5 percent would directly lessen this disincentive while still ensuring that all retained exposures are treated as being not risk-free.

In addition, the 10 percent risk weight floor discourages CRT through its duplicative nature. Per the ERCF's operational criteria for CRT, FHFA must approve each transaction as being effective in transferring the credit risk of one or more mortgage exposures to another party, taking into account any counterparty, recourse, or other risk to the Enterprise and any capital, liquidity, or other requirements applicable to counterparties. This regulatory approval process mitigates the safety and soundness risk posed by CRT structures and contractual terms, lessening the need for a tranche level risk weight floor as high as 10 percent. Moreover, the Enterprises are able to further lessen the need for a punitive CRT risk weight floor with their ability to mitigate unknown risks through their underwriting standards and servicing and loss mitigation programs. The standards and programs are flexible, rigorous, and constantly evolving, helping minimize losses through the entire life cycle of a mortgage loan.

FHFA continues to believe that CRT can play an important role in ensuring that each Enterprise operates in a safe and sound manner and is positioned to fulfill its statutory mission across the economic cycle. FHFA also continues to believe that an Enterprise does retain some credit risk on its CRT and that the risk should be appropriately capitalized. FHFA believes that a 5 percent CRT risk weight floor will enhance the safety and soundness of the Enterprises by increasing the incentives to undertake risk transfer activities while continuing to capitalize retained CRT tranches against structure, model, unforeseen, and other risks. Furthermore, lowering the tranche level risk weight floor should reduce the extent to which the CRT effectiveness adjustments may require more regulatory capital for retained CRT exposures than is necessary to ensure safety and soundness, and help ensure that FHFA does not unduly discourage CRT on mortgage exposures with risk profiles similar to those of recent acquisitions by the Enterprises.

Question 5: Is the 5 percent prudential floor on the risk weight for a retained CRT exposure appropriately calibrated? What adjustment, if any, would you recommend?

Overall Effectiveness Adjustment

The proposed rule would remove the requirement that an Enterprise must apply an overall effectiveness adjustment to its retained CRT exposures in accordance with the ERCF's securitization framework in 12 CFR 1240.44(f) and (i).

FHFA included an overall effectiveness adjustment in the CRT securitization framework largely in response to comments received on FHFA's 2018 notice of proposed rulemaking on Enterprise capital. Commenters argued that CRT has less loss-absorbing capacity than an equivalent amount of equity financing due to the upfront and ongoing costs of CRT, and that while CRT coverage is only on a specified pool, equity financing can cross-cover risks throughout the balance sheet.

However, commenters on the 2020 ERCF notice of proposed rulemaking argued that while these considerations are reasonable, in the context of the totality of the proposed CRT framework and a credible leverage ratio requirement as a backstop, the overall effectiveness adjustment is not needed and creates unnecessary disincentives for the Enterprises to engage in CRT. In addition, commenters stated that the CRT tranche risk weight floor covers the risk that a CRT will not perform as expected in transferring credit risk to third parties, which is similar to the risk that the overall effectiveness adjustment was designed to cover.

Unlike the counterparty and loss-timing effectiveness adjustments in the CRT securitization framework, the overall effectiveness adjustment does not target specific risks. For this reason, and given the opinions of commenters on the overall effectiveness adjustment, FHFA has determined that it is an appropriate place to make a refinement within the CRT securitization framework to further promote the use of CRT without increasing safety and soundness risks at the Enterprises. FHFA is proposing to remove the adjustment rather than to reduce it due to the lack of empirical evidence suggesting that a lower overall effectiveness adjustment is less duplicative than the adjustment in the ERCF final rule published on December 17, 2020.

Question 6: Is the removal of the overall effectiveness adjustment within the CRT securitization framework appropriate in light of the proposed rule's 5 percent prudential floor on the risk weight for retained CRT exposures?

Adjustments to CRT Capital Relief

The two proposed CRT modifications would increase the capital relief afforded an Enterprise for well-structured CRT on many common mortgage exposures, increasing incentives for the Enterprises to engage in CRT. For existing CRT, the two changes would increase capital relief compared to the current ERCF; however, the changes may not impact future CRT in exactly the same way. Each Enterprise has designed its existing CRT structures with attachment and detachment points, collateralization, and other terms based on the current ERCF and previous guidance. Each Enterprise will likely be able to structure the tranches and other aspects of its future CRT somewhat differently, taking into account modifications in any finalized rule amendments. Nonetheless, FHFA believes that the proposed rule's modifications would reduce the extent to which the CRT methodology may require more regulatory capital for retained CRT exposures than is necessary to ensure safety and soundness. FHFA also believes that these modifications would provide each Enterprise a mechanism for flexible and substantial capital relief through CRT, and CRT likely will remain a valuable tool for managing credit risk and that each Enterprise will base its CRT decisions on its own risk management assessments, not solely on the regulatory risk-based capital requirements.

The proposed rule would implement a modified ERCF CRT framework through which an Enterprise determines its credit risk-weighted assets for any eligible retained CRT exposures and any other credit risk that might be retained on its CRT. Under the proposed rule, an Enterprise would calculate credit risk-weighted assets for retained credit risk in a CRT using risk weights and exposure amounts for each CRT tranche. The exposure amounts of the retained CRT exposures for each tranche would be increased by adjustments to reflect counterparty credit risk and the length of CRT coverage ( i.e., remaining time until maturity). Unlike the current ERCF, the proposed framework would not include an overall effectiveness adjustment. Further, the proposed rule would also set a credit risk capital requirement floor for retained risk through a tranche-level risk weight floor of 5 percent rather than 10 percent.

The two proposed modifications to the CRT securitization framework could lead to a significant increase in capital relief. For Fannie Mae and Freddie Mac combined, capital relief from single-family CRT would increase by an estimated 45 percent, while capital relief from multifamily CRT would increase by an estimated 33 percent. Together, aggregate capital relief on the Enterprises' books of business would increase by an estimated 40 percent, where the increase is driven primarily by the change to the CRT tranche risk weight floor as evidenced by the example below. These modifications could help to ensure that the rule does not create undue disincentives to utilize CRTs.

Question 7: Is the proposed approach to determining the credit risk capital requirement for retained CRT exposures appropriately formulated? What adjustments, if any, would you recommend?

Question 8: Will the proposed amendments to the CRT securitization framework provide the Enterprises with sufficient incentives to engage in more CRT transactions without compromising safety and soundness?

CRT Example

To provide clarity on how the proposed modifications would alter the CRT risk weight calculations, we provide an example using the same stylized CRT that was used as an example in the ERCF notice of proposed rulemaking. Consider the following inputs from an illustrative CRT:

• $1,000 million in unpaid principal balance of performing 30-year fixed rate single-family mortgage exposures with original loan-to-values (OLTVs) greater than 60 percent and less than or equal to 80 percent;
• CRT coverage term of 10 years;
• Three tranches—B, M1, and AH—where tranche B attaches at 0% and detaches at 0.5%, tranche M1 attaches at 0.5% and detaches at 4.5%, and tranche AH attaches at 4.5% and detaches at 100%;
• Tranches B and AH are retained by the Enterprise, and ownership of tranche M1 is split between capital markets (60 percent), a reinsurer (35 percent), and the Enterprise (5 percent);
• The aggregate credit risk-weighted assets on the single-family mortgage exposures underlying the CRT are $343.8 million;
• Aggregate expected losses on the single-family mortgage exposures underlying the CRT of $2.5 million; and
• The reinsurer posts $2.8 million in collateral, has a counterparty financial strength rating of 3, and does not have a high level of mortgage concentration risk.

The Enterprises would first calculate risk weights for each tranche assuming full effectiveness of the CRT in transferring credit risk on the underlying mortgage exposures. In general, tranche risk weights are the highest for the riskiest, most junior tranches (such as tranche B), and lower for the more senior tranches (such as tranches M1 and AH). The proposed rule would lower risk weights on senior tranches compared to the current ERCF.

For the illustrative CRT, the overall risk weights for the proposed rule across tranches AH, M1, and B are 5%, 783%, and 1,250%, where 5% reflects the proposed minimum risk weight. By comparison, the overall risk weights under the ERCF across tranches AH, M1, and B are 10%, 785%, and 1,250%, where 10% reflects the minimum risk weight. The difference between the M1 risk weights, 783% for the proposed rule and 785% for the ERCF, reflects a weighted average risk weight calculation for M1 because M1's attachment and detachment points straddle stress loss. That is, the weighted-average risk weight would be the average of 1,250 percent, weighted by the portion of the tranche exposed to projected stress loss, and the minimum risk weight (5 percent for the proposed rule and 10 percent for ERCF) weighted by the portion of the tranche not exposed to projected stress loss.

Risk weights from the proposed rule:

Next, the Enterprise would calculate the adjusted exposure amount of its retained CRT exposures to reflect the effectiveness of the CRT in transferring credit risk on the underlying mortgage exposures. For the illustrative CRT, tranches AH and B are retained by the Enterprise, and do not need further adjustment. Risk associated with tranche M1 is transferred through a capital markets transaction and a loss sharing agreement. For the proposed rule, risk transfer on this tranche is subject to the following two effectiveness adjustments, which are reflected in the Enterprise's adjusted exposure amount: Loss sharing effectiveness adjustment (LSEA) and loss timing effectiveness adjustment (LTEA). The current ERCF includes an additional on-the-top overall effectiveness adjustment (OEA), which acts like a capital relief haircut.

Both the proposed rule and the current ERCF utilize the same methodology when accounting for the effectiveness of loss sharing on tranche M1. In particular, both methods adjust the Enterprise's exposure amount on tranche M1 to reflect the retention of some of the counterparty credit risk that was nominally transferred to the counterparty. To do so, the methods adjust effectiveness for: (i) Uncollateralized unexpected loss (UnCollatUL); and (ii) uncollateralized risk-in-force above stress loss (SRIF). The approaches differ in their capitalization of SRIF. The proposed rule would capitalize SRIF at a 5% risk weight and the current ERCF capitalizes SRIF at a 10% risk weight, where the difference reflects the different risk weight floors.

For the illustrative CRT, the counterparty haircut is 5.2% as per the ERCF's single-family CP haircuts, UnCollatUL is 42.5%, and SRIF is 37.5%. The proposed rule's LTEA on tranche M1 would be 96.5%, which when rounded, is the same figure for LTEA under the current ERCF.

LSEA from the proposed rule:

Both the proposed rule and the current ERCF utilize the same methodology when accounting for effectiveness from the timing of coverage by adjusting the Enterprise's exposure amount for tranche M1 to reflect the retention of some loss timing risk that was nominally transferred. The loss timing factor addresses the mismatch between lifetime losses on the 30-year fixed-rate single-family mortgage exposures underlying the CRT and the CRT's coverage. The loss timing factor for the illustrative CRT with 10 years of coverage and backed by 30-year fixed-rate single-family whole loans and guarantees with OLTVs greater than 60 percent and less than or equal to 80 percent is 88 percent for both the capital markets transaction and the loss sharing agreement. For the illustrative CRT, tranche M1's LTEA is 85.6% and is derived by scaling stress loss by the 88% loss timing factor.

LTEA from the proposed rule and the current ERCF:

Where

LTK _A, % = max ((2.75% + 0.25%) * 88%−0.25%, 0%) = 2.39%

The current ERCF includes a third adjustment, the OEA, that the proposed rule omits.

OEA from the current ERCF:

ERCF OEA % = 100% * (1.06667−4.1667 * K _A ) = 95.2%

The next steps convert the effectiveness adjustments into Enterprise exposures. In particular, the adjusted exposure amounts (AEAs) combine the effectiveness adjustments, aggregate UPB, tranche thickness, and an adjustment for expected losses (to tranche B in the example). For the illustrative CRT, the proposed rule would calculate AEAs as follows:

AEA % _,AH = EAE % _,AH * AggUPB _$ * ( D−A ) = $1,000 m * (100%−4.5%) = $955 m

AEA % _,M1 = EAE % _,M1 * AggUPB _$ * ( D−A ) = 19.7% * $1,000 m * (45%−0.5%) = $7.9 m

where the Enterprise's adjusted exposures (EAEs) for tranches A and B are 100% and

EAE %_,M1 = 100% − (60% * 85.6%) − (35% * 96.5% * 85.6%) = 19.7%.

The current ERCF calculates AEAs including the OEA, thus increasing the Enterprise's exposure on M1. For tranches AH and B, the current ERCF's AEAs are the same as those of the proposed rule because the Enterprise does not transfer risk on the AH and B tranches.

ERCF_AEA %_,M1 = ERCF_EAE %_,M1 * AggUPB _$ * (D − A) = 23.6% * $1,000 m * (4.5% − 0.5%) = $9.4 m

ERCF_EAE %_,M1 = 100% − (60% * 85.6% * 95.2%) − (35% * 96.5% * 85.6% * 95.2%) = 23.6%

Finally, the risk weights and exposures are combined to calculate risk-weighted assets. For the illustrative CRT, the proposed rule would calculate risk-weighted assets (RWA) as follows:

RWA _$,AH = AEA _$,AH * RW %_,AH = $955 m * 5% = $47.8 m

RWA = AEA _$,M1 * RW %_,M1 = $7.9 m * 783% = $61.8 m

RWA = AEA _$,B * RW %_,B = $2.5 m * 1250% = $31.3 m

with total RWAs on the retained CRT exposures at $140.8 million, a decline of $202.9 million from the aggregate credit risk-weighted assets on the underlying single-family mortgage exposures of $343.8 million.

By comparison, the current ERCF's total RWA are higher primarily due to its higher risk weight floor on the senior AH exposure:

ERCF_RWA _$,AH = ERCF_AEA _$,AH * ERCF_RW %_,AH = $955 m * 10% = $95.5 m

ERCF_RWA _$,M1 = ERCF_AEA _$,M1 * ERCF_RW %_,M1 = $9.4 m * 785% = $74.1 m

ERCF_RWA _$,B = ERCF_AEA _$,B * ERCF_RW %_,B = $2.5 m * 1250% = $31.3 m

with total RWAs on the retained CRT exposures at $200.8 million.

Overall, for this stylized CRT, the proposed rule's total RWA capital relief of $202.9 million is 42 percent higher than the $143.0 million in capital relief from the current ERCF.

C. ERCF Technical Corrections

The proposed rule would make technical corrections to the ERCF related to definitions, variable names, the single-family countercyclical adjustment, and CRT formulas that were not accurately reflected in the ERCF final rule published on December 17, 2020. These technical corrections would revise the ERCF for the following items:

• In § 1240.2, the definition of “Multifamily mortgage exposure” would be moved from its current location to a location that follows alphabetical order relative to the other definitions within the section. The definition of a multifamily mortgage exposure would not change.

• In § 1240.33, the definition of “Long-term HPI trend” would be updated to correct a typographical error that resulted in only the coefficient of the trendline formula, 0.66112295, being published. The corrected trendline formula would be 0.66112295_e ^{0.002619948*t)} . The Enterprises use the long-term HPI trend as the basis for calculating the single-family countercyclical adjustment. As published, the trendline would be a time-invariant horizontal line rather than a time-varying exponential function.

• In § 1240.33, the definition of OLTV for single-family mortgage exposures would be amended to include the parenthetical (original loan-to-value) after the acronym to provide additional clarity as to the meaning of OLTV. Single-family OLTV would continue to be based on the lesser of the appraised value and the sale price of the property securing the single-family mortgage.

• In § 1240.37, the second paragraph (d)(3)(iii) would be redesignated as paragraph (d)(3)(iv) to correct a typographical error.

• In § 1240.43(b)(1), the term “KG” would be replaced with “ K _G ” to correct a typographical error.

• In § 1240.44,

○ In paragraph (b)(9)(i)(C), the term “( LTFUPB% )” would be replaced with the term “( LTFUPB % )” to correct a typographical error;

○ In paragraph (b)(9)(i)(D), the term “LTF%” would be replaced with the term “ LTF % ” to correct a typographical error;

○ In paragraph (b)(9)(ii), the term “ LTF%” would be replaced with the term “ LTF % ” to correct a typographical error;

○ In paragraph (b)(9)(ii)(B), the term “( CRTF15% )” would be replaced with the term “( CRTF15 % )” to correct a typographical error;

○ In paragraph (b)(9)(ii)(C), the term “( CRT80NotF15% )” would be replaced with the term “( CRT80NotF15 % )” to correct a typographical error.

○ In paragraph (b)(9)(ii)(E)( 2 )( i ), the equation would be revised to correct a typographical error. The revised equation would be:

LTF % = ( CRTLT 15 * CRTF 15 % ) + ( CRTLT 80 Not 15 * CRT 80 NotF 15 % ) + ( CRTLTGT 80 Not 15 * (1− CRT 80 NotF 15 % − CRTF 15 % ));

○ In paragraph (b)(9)(ii)(E)( 2 )( iii ), the term “ LTF% ” would be replaced with the term “ LTF % , ” to correct a typographical error;

○ In paragraph (c) introductory text, the term “ RW%” would be replaced with the term “ RW % ” to correct a typographical error;

○ In paragraph (c)(1), the term “ AggEL%” would be replaced with the term “ AggEL % ” to correct a typographical error;

○ In paragraph (g), the first three equations would be combined into one equation to correct a typographical error that erroneously split the equation into three distinct parts. The revised equation would be:

IV. Paperwork Reduction Act

The Paperwork Reduction Act (PRA) (44 U.S.C. 3501 et seq. ) requires that regulations involving the collection of information receive clearance from the Office of Management and Budget (OMB). The proposed rule contains no such collection of information requiring OMB approval under the PRA. Therefore, no information has been submitted to OMB for review.

V. Regulatory Flexibility Act

The Regulatory Flexibility Act (5 U.S.C. 601 et seq. ) requires that a regulation that has a significant economic impact on a substantial number of small entities, small businesses, or small organizations must include an initial regulatory flexibility analysis describing the regulation's impact on small entities. FHFA need not undertake such an analysis if the agency has certified that the regulation will not have a significant economic impact on a substantial number of small entities. 5 U.S.C. 605(b). FHFA has considered the impact of the proposed rule under the Regulatory Flexibility Act. The of FHFA certifies that the proposed rule, if adopted as a final rule, would not have a significant economic impact on a substantial number of small entities because the proposed rule is applicable only to the Enterprises, which are not small entities for purposes of the Regulatory Flexibility Act.

Proposed Rule

List of Subjects for 12 CFR Part 1240

• Capital
• Credit
• Enterprise
• Investments
• Reporting and recordkeeping requirements

Authority and Issuance

For the reasons stated in the Preamble, under the authority of 12 U.S.C. 4511, 4513, 4513b, 4514, 4515-17, 4526, 4611-4612, 4631-36, FHFA proposes to amend part 1240 of title 12 of the Code of Federal Regulation as follows:

Chapter XII—Federal Housing Finance Agency

Subchapter C—Enterprises

PART 1240—CAPITAL ADEQUACY OF ENTERPRISES

1. The authority citation for part 1240 is revised to read as follows:

Authority: 12 U.S.C. 4511, 4513, 4513b, 4514, 4515, 4517, 4526, 4611-4612, 4631-36.

2. Amend § 1240.2 by removing the definition of “Multifamily mortgage exposure” and adding the definition of “Multifamily mortgage exposure” in alphabetical order to read as follows:

3. Amend § 1240.11 by revising paragraph (a)(6) to read as follows:

4. Amend § 1240.33(a) by:

a. In the definition of “Long-term HPI trend”, removing “0.66112295” and adding “0.66112295 e ^{0.002619948*t)} ” in its place; and

b. Revising the definition of “OLTV”.

The revision reads as follows:

5. Amend § 1240.37 by redesignating the second paragraph (d)(3)(iii) as paragraph (d)(3)(iv).

6. Amend § 1240.43 in paragraph (b)(1) by removing the term “KG” and adding the term “ K _G ” in its place.

7. Amend § 1240.44 by:

a. In paragraph (b)(9)(i)(C), removing the term “( LTFUPBE% )” and adding the term “( LTFUPB % )” in its place;

b. In paragraph (b)(9)(i)(D) introductory text, removing the term “LTF%” and adding the term “ LTF % ” in its place;

c. In paragraph (b)(9)(ii) introductory text, removing the term “ LTF%” and adding the term “ LTF % ” in its place;

d. In paragraph (b)(9)(ii)(B), removing the term “( CRTF15% )” and adding the term “( CRTF15 % )” in its place;

e. In paragraph (b)(9)(ii)(C), removing the term “( CRT80NotF15% )” and adding the term “( CRT80NotF15 % )” in its place;

f. Revising the equation in paragraph (b)(9)(ii)(E)( 2 )( i );

g. In paragraph (b)(9)(ii)(E)( 2 )( iii ) introductory text, removing the term “ LTF% ” and adding the term “ LTF % _, ” in its place;

h. In paragraph (c) introductory text:

i. Removing the term “ RW%” and adding the term “ RW % ” in its place; and

ii. Removing “10 percent” and adding the term “5 percent” in its place;

i. In paragraph (c)(1), removing the term “ AggEL%” and adding the term “ AggEL % ” in its place;

j. In paragraphs (c)(2) and (c)(3)(ii), removing the term “10 percent” and adding the term “5 percent” in its place;

k. Revising the first equation in paragraph (d);

l. In paragraph (e), removing the term “10 percent” and adding the term “5 percent” in its place;

m. Revising paragraph (f)(2)(i);

n. In paragraph (g), revising the first three equations;

o. Revising the first equation in paragraph (h); and

p. Removing and reserving paragraph (i).

The revisions read as follows: