The Heavy Lift to Implement GSE Reform-Recap-Release: Part 2: Charter-defect Reforms

Introduction

The most discussed topic in housing finance policy in the last few months has been whether and when the second Trump administration (Trump II) might end the conservatorships of Fannie Mae and Freddie Mac (F&F), the two large government-sponsored enterprises (GSEs).  This four-part series addresses the “how” of such a conservatorship exit, i.e., what it will take to actually implement it.  Importantly, this series is based on the widely held view that congressional legislation to reform the GSEs will not occur in the foreseeable future and that any significant changes to the GSE structure will thus need to be implemented via “administrative means,” as defined directly below.

In 2016-2017, a policy consensus developed around the idea of keeping F&F rather than replacing them.[1]  However, this agreement hinged on the condition that they would only exit conservatorship after addressing “the five or six big things” generally believed to have gone wrong with their business model, which was designed by Congress when it established their charters.[2] The concept of GSE reform by administrative means was then developed as a way to implement those fixes, but without waiting for Congress to revise those charters. Instead, the required improvements would be implemented by a very limited number of new or revised Federal Housing Finance Agency (FHFA) regulations related to the safety and soundness of the GSEs or through modifications to the Preferred Stock Purchase Agreement (PSPA).[3]

In fact, according to the thinking associated with reform via administrative means, releasing F&F from conservatorship without fixing those fundamental defects in their charters would be considered policy malpractice, as the two GSEs could easily return to the problematic behaviors they displayed prior to 2008, including those that landed them in conservatorship in the first place.[4]

Part 2 discusses what those five or six big charter defects are, and what it will take to successfully remedy them as part of conservatorship exit. 

 

The “big three” charter-defect reforms

There was, from the very beginning of the proposed administrative path for reform, an extremely strong policy consensus on what three of those charter flaws were.  The “big three,” in order of most to least difficult to address, were:

• that the required minimum regulatory capital was too low and inadequately sensitive to risk;[5]
• that the implied guarantee was a bad system of government support. It created undesirable ambiguity for MBS investors during times of market stress and also failed to provide compensation to taxpayers for the risk they were taking on.[6]
• that the ability to carry an unlimited discretionary investment portfolio funded by implied-guarantee – and therefore subsidized – borrowings was too easily abused. This resulted in taxpayer subsidies to F&F being misused[7] and was the source of significant safety and soundness problems.[8]

The first three workstreams described below aim to successfully mitigate these defects through changes in FHFA regulations and PSPA amendments, i.e., via administrative means. 

Additionally, the fourth workstream discusses two potential charter defects that policymakers might consider to finalize the list of the “five or six big things.”

Workstream 1 – Regulatory capital:  Develop a high-quality minimum regulatory capital requirement, defined as one that (1) allows F&F to “remain viable as going concerns after a severe economic downturn,”[9]  and (2) is risk-sensitive at a detailed enough level to support F&F making economically accurate decisions about the balance between risk and reward in their transactions.[10]

The FHFA adopted a regulatory capital requirement in 2020, known as the Enterprise Regulatory Capital Framework (ERCF), which applies to F&F. It has been amended in several modest ways since then. Unfortunately, in my view, it does not adequately meet the two requirements for a high-quality regulatory capital requirement, as was apparent in the mostly poor response it received when proposed. Specifically:

• The level of capital required by the ERCF – currently about $330 billion – seems to be far in excess of what is justified by the actual losses the GSEs would incur after a severe economic downturn.[11] 
• At its core, the ERCF commendably calculates strongly risk-sensitive capital requirements. Unfortunately, it then overlays large, risk-insensitive buffers and other features to calculate a total capital required. This means that, in aggregate, the ERCF is not risk-sensitive enough to support economically accurate risk-versus-reward transaction-level decision-making.

As a result, the ERCF capital requirement is not only unnecessarily complex and onerous but could potentially incentivize distorted behavior by F&F. Thus, it behooves the FHFA to take a fresh look at developing a better regulatory capital requirement as it contemplates conservatorship exit.

I note that, during the earlier years of conservatorship (circa 2016-17), a capital requirement –called the Conservatorship Capital Framework (CCF) – was developed, which had extensive risk-sensitivity to incentivize proper transactional decision-making by F&F during conservatorship.[12] It was later proposed, but never finalized, as an official regulatory capital requirement.  During that process, commentators criticized just one major aspect of it: the lack of a countercyclical buffer.[13] Otherwise, it was generally well regarded. I estimate, based on F&F’s current aggregate balance sheet and the addition of a countercyclical buffer, an enhanced CCF would very roughly require $200 to $225 billion of capital,[14] or about two-thirds of what the ERCF calls for.  Such an enhanced CCF could well serve as a starting place for developing a high-quality capital requirement replacement, potentially expediting a process that could otherwise stretch for more than a year.[15] 

Of course, any change in the capital requirement, whether up or down, will be controversial since it directly impacts G-fees and, consequently, mortgage rates. For example, small-government conservatives, who have long advocated for higher G-fees to help shrink F&F’s collective footprint by losing market share to other sources of mortgage credit,[16] might be expected to strongly oppose any reduction in the minimum regulatory capital required.

 

Workstream 2 – Determining the nature of government support, if any, and how it will be paid for. 

The F&F business model, from their first securitizations to the present, relies on MBS investors viewing their securities as having nil credit risk. This perception is largely due to a government backstop that ensures F&F’s guarantees to the MBS investors against credit loss are always effective.[17]  Before conservatorship, this backstop took the form of an implied guarantee; simultaneously with F&F entering conservatorship, that backstop became formalized by the PSPA.[18   It was the assumption inside conservatorship when the concept of “exit by administrative means” was developed that the PSPA would remain in effect. This assumption was also made later in the first Trump administration’s Housing Reform Plan issued in 2019 in the case of reform by administrative means. 

Continuing the PSPA was viewed as good policy for three reasons. First, the historic implied guarantee was considered by many policymakers to be bad public policy. This is because it was seen as ambiguous and weak when needed most during times of market stress, and because it unfairly left taxpayers uncompensated for the risk they were actually taking on. Second, there were concerns that an attempt to go back to the implied guarantee could once again risk a loss of market confidence, especially during future periods of mortgage market stress. This change would also likely increase MBS interest rates almost immediately given the reduced quality of the government backstop to the guarantees of F&F, which would, in turn, push mortgage rates up. And third, the PSPA is needed to function as the legal vehicle for imposing many restrictions and obligations on F&F. This includes correcting for charter defects and, as will be discussed in Part 4 of this article series, ensuring the implementation of market-changing reforms made over the years to F&F’s operations.[19] 

It is worth noting that an alternative to the PSPA has long been proposed. This alternative recommends that the government should enact a full-faith-and-credit guarantee of the MBS issued by the GSEs.[20]  However, this would require GSE-related legislation viewed as unlikely to happen for many years to come. So, the only obviously available option that would not require legislation and would not return to the discredited implied guarantee of the past is to continue the PSPA along with any amendments to it that policymakers believe make sense.

Thus, this workstream needs to confirm whether the PSPA will continue and determine what cost, if any, will be paid by F&F to Treasury for this ongoing support.[21]  The concept of a fee is not new; the PSPA agreement from 2008 includes a provision for such a fee, known as the “periodic commitment fee,” although it was never implemented.  The decision to impose a fee and how to calculate it could be quite controversial, so it could take time for Trump II to reach a decision on the matter. If this fee for support is more than minimal – anything perhaps higher than 0.05 percent per annum on the mortgage assets outstanding – there will likely be a negative reaction by many interest groups who may fear that this cost will be passed down to borrowers in the form of higher mortgage rates.

 

Workstream 3 – Establish a new limit for F&F’s investment activities to replace the current one.

Prior to conservatorship, F&F heavily exploited the access afforded to capital markets by the implied guarantee to borrow on an unsecured basis and use the funds to invest, without limit, in mortgage-related assets. That portfolio peaked at over $1.5 trillion circa 2006 and became the source of most F&F profits. In the early 2000s, the Bush administration and the Federal Reserve tried to rein this in with a cap, which required legislation. Their joint push failed in Congress, demonstrating the impressive lobbying strength of F&F. 

Unsurprisingly, with the advent of conservatorship, the newly instituted PSPA had a clause requiring F&F to reduce over time the level of these discretionary investment portfolios down to $250 billion each, which is about a two-thirds reduction from their peak. Since then, further changes to the PSPA and additional conservatorship directives from the FHFA have resulted in an even more significant reduction in the size of these portfolios. Today, they amount to about $100 billion each, down about 85 percent from the peak, and a significant share of the remaining investments is not discretionary but instead consists of assets such as defaulted and modified mortgages that are integral to the single-family guarantee business.[22] 

This workstream should develop a long-term replacement for the current limit, to be added to the post-conservatorship PSPA as a contract requirement. It should likely not be a set dollar amount in order to accommodate F&F balance sheet growth (due to house price inflation, if nothing else) over the coming decades.[23]  Instead, it could perhaps be as simple as a percentage of the balance sheet size of each GSE, such as 4 or 5 percent of assets.[24]

A proposal to modify the current limit should be relatively easy and quick to develop and implement. It does not seem likely that any such proposed limit, as long as it does not significantly allow growth in these portfolios above their current levels, would encounter significant criticism. By contrast, a failure to include a limit going forward – i.e., a return to the unlimited amount allowed pre-conservatorship – would be highly controversial.

 

Workstream 4 – Determine and develop any additional reforms to address significant defects in the charters. 

This workflow examines additional features of F&F’s charters that could be deemed defective as part of addressing the “five or six big things” required before an exit from conservatorship should be considered.  While many interest groups can potentially claim their concerns rise to this level of importance, I am only aware of two other features of how the GSEs operate that realistically might be considered by policymakers to be charter defects in the same way as the “big three” discussed above.   

Non-discrimination by size. Prior to conservatorship, F&F – following a standard business practice justifiable by the economy of scale of their serving a primary market lender – gave volume discounts on G-fees. Unsurprisingly, smaller lenders considered this unfair.  Their logic was that F&F were not just typical private sector companies, but alter egos for the U.S. government, designed to support the mortgage markets and promote homeownership.  Thus, F&F putting a thumb on the competitive scale to advantage larger lenders via G-fee volume discounts was, in their view, counter to good public policy. This argument won the day during conservatorship, as the FHFA made it a requirement, via a directive, that there be no volume discounts, and also made it into  Trump I’s Treasury Housing Reform Plan as a requirement.[25]  So, while volume discounts were not a cause of F&F’s financial distress prior to entering conservatorship, it is clear many policymakers regard it as a requirement that needs to be “fixed” before F&F exit conservatorship. 

This can easily be implemented via a contract clause in the PSPA, as it is already partially in place. Maintaining this “level G-fee” policy is unlikely to cause controversy; however, allowing it to lapse would result in strong opposition from groups representing smaller lenders. 

Limits on multifamily activities. Recognizing that F&F are a channel for government subsidy to housing, Congress wanted to prevent wealthier individuals from benefiting from that subsidy. This took the form of the conforming loan limit (CLL), i.e., the maximum principal amount of a single-family mortgage loan that F&F are allowed to purchase, according to their charters. Although this limit is fairly high – today it amounts to $806,500 generally, and up to $1,209,750 in identified “high-cost areas” – the principle is clear: the subsidy is designed to be limited.

When it comes to the multifamily business, there is surprisingly no such similar limit.  The equivalent to the CLL would be, for example, some defined limit on the amount of the multifamily mortgage being purchased by F&F, calculated on a per-apartment basis for the building being financed.  But no such limit exists, although the GSEs in conservatorship have generally shied away from luxury properties under the FHFA’s guidance. So, possibly one of the five or six big things to fix before F&F exit conservatorship could be establishing a limit on multifamily lending comparable to the CLL.[26]  However, as the average calculated loan per apartment is so much lower than for a single-family home, there has been almost no focus by policymakers on this topic to date. 

Instead, the focus has been on limiting the value of all the multifamily mortgages purchased in a year. The FHFA, as conservator, first put such a limit in place over a decade ago, with it varying over time based on the size of the market; it also has had other criteria or exclusions added in line with the views of whoever was director of the FHFA at the time. The logic behind the cap was to prevent F&F from using its government subsidy to unfairly compete with private sector multifamily lenders, a very different reasoning than what is behind the CLL for single-family housing. In fact, at one point, the market share of F&F in multifamily lending was threatening to exceed 50 percent, which the FHFA considered a highly undesirable potential outcome at the time. 

This then raises the question of how best to design a limit that can adapt to the changing size of the market. Other criteria or exclusions, such as affordable housing percentages, could also be considered. The appropriate wording, which could be quite complex, would then need to be added to the PSPA to clearly reflect that limit.   

A limit in general, and one that is considered “tight” in particular, would likely be disliked by builders, landlords, and other groups. It could, however, be viewed positively by banks and insurance companies who are also in the business of financing multifamily properties, as it would result in less competition from F&F.

Part 2 has listed six arguable defects in F&F’s charters that need to be addressed, along with possible ways to do so: three that are almost universally associated with F&F’s collapse in 2008, and three others that could be designed to prevent future abuse of the subsidies enjoyed by F&F. In aggregate, the chosen fixes will require Trump II policymakers to make several challenging policy and political choices .  Additionally, any necessary changes to regulations needed to implement some of these fixes must undergo a formal and lengthy process, which extends their implementation time. 

The next installment of this series, Part 3, will focus on recapitalization. It will explore how F&F can raise enough capital to be eligible to exit conservatorship, focusing on how unprecedentedly large the amounts of capital needed are and also how it is likely to take years to reach full recapitalization. Part 3 will also explore what it will then take for Treasury, which will end up as each of F&F’s majority shareholder upon their conservatorship exit, to later sell off its equity ownership position, which is needed for full re-privatization.

 

Footnotes

[1] From 2009 to then, the policy consensus, supported by Obama administration policy, was to wind down F&F and replace them in some fashion, with Congress in charge of determining what exactly would be the replacement.

[2] F&F’s “charters” are the name for the legislation establishing the two companies, which also lists their mission, things they must do, things they can’t do, and other operational requirements.

[3] This is the legal agreement between each of F&F and Treasury, where Treasury provides them with the government-quality credit support needed to backstop the guarantee against credit losses they give to investors in their mortgage-backed securities (MBS). 

[4]Treasury Secretary Mnuchin, the first to publicly announce the alternative of an exit by administrative means, indicated that, because this approach was less permanent than doing reform via legislation – after all, a new administration could change those regulations or the PSPA – he considered an exit by administrative means a good place to start. He hoped Congress would follow up later with legislation to enshrine the “fixes.” 

[5] The charters of F&F had a capital requirement for just 0.45 percent of mortgages securitized and guaranteed (plus additional amounts for other, smaller asset classes). This was very low even in an era of generally low capital requirements applied to financial institutions. It was also insensitive to the riskiness of the underlying loans, thus encouraging high-risk lending.

[6] See Treasury’s Housing Reform Plan, September 2019, page 1, which notes that “…the existing Government support of the secondary market should be explicitly defined, tailored and paid-for…” to remedy the weaknesses of the pre-conservatorship implied guarantee. https://home.treasury.gov/system/136/Treasury-Housing-Finance-Reform-Plan.pdf

[7] Such subsidies were meant to support homeownership by making mortgages cheaper and more readily available. Instead, they became heavily used to support the discretionary investing activities of the GSEs, which meant the benefit of the subsidies primarily went to increasing shareholder profits and management compensation.

[8] The large discretionary investment portfolios introduced large liquidity and earnings risks onto the balance sheet of F&F, which became evident in 2007 and 2008. In fact, during the months in 2008 that led to conservatorship, F&F’s investment securities portfolios were a very major source of liquidity and earnings stresses, rivaling those caused by their core mortgage purchase activities.

[9] See Treasury’s Housing Reform Plan, page 1, for this definition. The focus on losses incurred in a severe economic downturn was a major conceptual change that arose from the great financial crisis (GFC).

[10] A financial institution, to deliver a proper return to its shareholders that reflects the risks the company is taking, needs to acquire and hold assets that earn more than their cost of capital on a risk-adjusted basis. This requires that the institution internally allocate its capital to transactions based upon their riskiness, with more capital allocated to riskier transactions.  If the amount of capital required by the regulatory capital system is not a reasonably accurate measure of that riskiness, it will thus introduce distorted decision-making. The most well-known example of this occurred when the regulatory capital required was calculated solely as a simple percentage of assets, without adjusting for the risk associated with those assets. As a result, riskier assets seemed attractive due to their calculated higher return because the regulatory capital was too low, leading to an inflated return-on-capital calculation. This caused banks and other lenders to excessively take on riskier loans. 

[11] The FHFA, consistent with Federal Reserve specifications, requires F&F to run stress tests annually. In recent years, the results of those tests have shown small or near-zero losses.

[12] As a matter of information, the CCF forms the risk-sensitive core of the ERCF, before the addition of those large and risk-insensitive buffers and other non-economic features. 

[13] A countercyclical buffer is an amount of capital needed to replace losses that might occur in the future during an economic downturn. While one could try to raise capital at the time of those losses, market conditions might be very hostile; thus, it is really appropriate to raise the money beforehand during good markets instead. A countercyclical buffer does just that. 

[14] There are many subtly different definitions of “capital” in financial institution regulatory history.  For simplicity, in this article, I am assuming “capital” equals the net worth of F&F.

[15] A regulatory change must go through the steps outlined by the Administrative Procedures Act, which includes a mandatory public comment period. The CCF and ERCF histories indicate the process could take most of a year or even longer. 

[16] The other relevant sources of mortgage credit would primarily be the Federal Housing Administration and bank balance sheets. 

[17 As a result, “agency MBS” investors – the ones who buy F&F’s securitizations, as well as those issued via Ginnie Mae – regard MBS as alternatives to Treasuries but with extra complexity related to the interest rate risk associated with how much, if at all, borrowers prepay their mortgages in advance of final maturity. Such investors, known as “rates investors,” do not look to take credit risk, either on the underlying mortgages or the corporate credit of F&F. 

[18] The PSPA is designed to give the market confidence that F&F will make good on these guarantees. It does this by promising, in a situation where a GSE would incur losses large enough to drive its net worth below zero, to inject new equity to bring its net worth back up to zero. In technical financial terms, this is called a “net worth keepwell.”  The PSPA successfully restored market confidence in F&F in 2008. The promise to do so is not unlimited, however. Currently, there is a limit as of year-end 2024 of $254.1 billion ($140.2 billion for Freddie Mac, $113.9 billion for Fannie Mae) on the amount that can be injected.

[19] The only other available vehicle to require the continuation of those reforms is to have a regulation declared by the FHFA. However, such a regulation must have a defensible nexus to safety and soundness, which only certain issues do.

[20] For example, Treasury’s Housing Reform plan proposed this if exit were to occur via legislative action. The Mortgage Bankers’ Association recently also announced support for this alternative. 

[21] Treasury conducted a study about how much should be paid for ongoing Treasury support during the Obama administration, but the results have never been made public. 

[22] In fact, the balance of such assets would be far higher if not for programs, again as directed by the FHFA, to sell off modified and non-performing loans.

[23] It should also be a percentage to reflect that F&F are not the same size, and over time their size relative to each other may change significantly. Today, Fannie Mae is about one-third larger than Freddie Mac. 

[24] By comparison, circa 2006, it accounted for roughly one-third of mortgage assets.

[25] This also has taken the form of F&F being required to operate their “cash window,” a mechanism used by smaller lenders to sell smaller-sized packages of mortgages to F&F. See page 3 of the Housing Reform Plan. 

[26] Given how the CLL varies by geographic area to reflect different cost of living and housing prices found throughout the country, a multifamily (MF) limit could similarly include such a sensitivity to cost/prices instead of being uniform nationwide.