Nonbank Mortgage Servicers: Proposing a Better Path to Reduce Their Risk to Financial Stability

Introduction

The Financial Stability Oversight Council (FSOC)¹  published its “Report on Nonbank Mortgage Servicing 2024”² (the Report) earlier this year. This nearly-50-page document concluded that nonbank mortgage servicers (NMSs) collectively are a potential threat to financial stability. However, in my view, the Report’s recommendations are not well-constructed, particularly because they fail to address head-on that threat’s actual root cause: two long-established mortgage securitization standards that have evolved, over time, to create large and destabilizing financial burdens on NMSs.³ 

In this article, I argue that these two standards are no longer essential given the evolution of the markets and mortgage industry over the decades since they were established. Thus, to effectively remove the NMS industry as a source of potential systemic instability, it would be best to similarly eliminate, or at least very materially reduce, the destabilizing financial burdens imposed on NMSs by those standards.  Doing so would help transform the NMS industry into a more straightforward – and stable – financial transaction processor (FTP) industry,⁴ significantly lowering its potential to cause systemic instability. 

After first providing an overview and analysis of the Report and its recommendations, I will discuss the two problematic standards in some detail. Finally, I will present a list of various pragmatic actions from which policymakers can select several they view as most well-suited for implementation.  The selected actions, taken together, should reduce the NMS industry’s collective threat to financial stability to the point where it is no longer systemic.

Overview of the FSOC Report

The FSOC was created in 2010 as a federal interagency committee primarily to identify threats to financial stability and recommend actions to avoid them.  Its focus on NMSs reflects the much larger role that such firms now play compared to the relatively recent past.  During the global financial crisis (GFC), NMSs held a cyclically low 4 percent market share compared to bank⁵ servicers. By 2022, the NMS industry had grown to command 54 percent of the servicing market, with expectations for further share growth. 

The Report observes that the vast majority of NMS firms are monolines, meaning they focus on a single line of business.  Because of this, such firms share among themselves very similar economic and financial characteristics, leading to a high correlation of when and how they might be impacted by market stress.  The FSOC thus concludes that, while any individual NMS firm is not a source of systemic instability, the collective NMS industry is.   

Boiling down the Report’s extensive discussion of risks related to NMSs, the FSOC finds that there are structural vulnerabilities that contribute to the risk of systemic instability.  In my words, those concerns fall into three categories of weaknesses that will likely manifest themselves during a period of mortgage market stress: 

Inability to meet liquidity requirements.

NMSs may struggle to access sufficient liquidity to meet ballooning cash needs during periods of stress.  A notable example was during COVID-19 when Congress mandated unprecedented mortgage payment forbearance.  This potential source of instability is exacerbated by NMSs’ reliance on wholesale funding sources that are also sensitive to market fluctuations.

Inability to maintain “safe and sound” operations.⁶  

The regulations covering NMSs do not subject them to strong enough prudential regulatory requirements to ensure they can operate in a safe and sound manner even during times of significantly adverse economic and market circumstances.  The Report notes, by contrast, that the regulations that apply to banks are designed to do just that.

Inability to continue mortgage servicing operations. 

If an individual NMS cannot meet its financial obligations, it may well have to enter bankruptcy. In such cases, there is likely to be an immediate halt in its operations, thus disrupting its servicing activities.  Given the high correlation between NMSs in their financial structure, this could well happen simultaneously to a large number of such companies, producing major mortgage market disarray during a stress event.

This all adds up to a convincing case that NMSs collectively pose a remote but nevertheless real risk of major financial instability. The result could be substantial losses to homeowners, the financial institutions that own most mortgage assets, and others, while also having the potential to more broadly destabilize the financial system.⁷ 

The Report also provides recommendations to address the resulting risk of instability. Those recommendations center heavily on issues related to Ginnie Mae,⁸ an emphasis discussed further below. The recommendations fall into two categories: 

Calls for congressional action, specifically:

1. to give Ginnie Mae expanded authority to provide funds to NMS firms during a period of liquidity stress;
2. to create an industry-financed fund to provide a type of debtor-in-possession financing to NMS firms that are in or near bankruptcy so that those firms can continue uninterrupted their servicing operations;⁹ and
3. to give certain new legal authorities to various mortgage-related regulators and institutions.

Calls for regulatory action, specifically:

1. that the state agencies which are primarily responsible for the prudential regulation of NMSs hold them to higher requirements, more akin to those that apply to banks;
2. that Ginnie Mae develop recovery planning requirements with its servicers to reduce the impact of potential operational disruptions; and 
3. that regulators improve coordination among themselves. 

I find these recommendations lacking for four main reasons.  First, the implementation timeline is rather long, as key recommendations are significantly dependent upon Congress passing legislation – and it must be assumed that such passage will require, at a minimum, at least several years.¹⁰ 

Second, the recommendations’ emphasis on Ginnie Mae is misplaced. Ginnie Mae, along with Fannie Mae and Freddie Mac (together known as “the agencies” in the mortgage industry), is responsible for more than 90 percent of all mortgage securitizations. However, Ginnie Mae is the smallest of the three, only representing about 27 percent of the market share held collectively by them.  This means the far bigger influence comes from Fannie Mae and Freddie Mac, the government-sponsored enterprises (GSEs) that are currently under the control¹¹ of the Federal Housing Finance Agency (FHFA).  Compounding this misplaced focus is the fact that changes at Ginnie Mae often require legislation, making their implementation uncertain and slow, while changes at the GSEs can usually be done relatively quickly as they only require FHFA approval. 

Third, key Report recommendations significantly look to create new programs – like giving Ginnie Mae greater authority to provide funds to NMSs – that would be called on to help cushion problems but only after those problems have manifested themselves.  This is clearly inferior to preventing the problems from arising in the first place.

And fourth, the report calls for “improved regulation,” which clearly seems to mean that NMS firms should be treated more like banks. This includes imposing higher capital and liquidity requirements, as well as implementing resolution planning¹² and other requirements. This approach seems misguided – like putting the proverbial round peg into a square hole – because NMSs differ greatly from banks in both their economic role and financial structure.¹³

Looked at as a whole, the fundamental approach of the Report is to accept the NMS industry as it is, warts and all, and then recommend various programs to hopefully ameliorate the situation.   However, the recommendations are significantly centered around two new programs designed to spring into action only after a future stress event is underway – and it is not at all a sure thing that will work as hoped.  The alternative approach described below is to instead directly reduce the impact of the two long-standing mortgage industry securitization practices that create the potential for NMS-related financial instability in the first place. 

Why the two industry securitization standards are problematic

Standard #1:  NMSs are obligated to make cash payments – called ‘servicer advances’ – that are both unpredictable and also potentially large enough to threaten their solvency.  Mortgage loan servicing, as already mentioned, is generally defined and talked about as if it were a specialized FTP industry. It was, therefore, a surprise to many at the outset of COVID-19 that the mortgage servicing industry asked for federal help of up to $100 billion of liquidity – i.e., cash.¹⁴ Clearly, something more than solely transaction processing was taking place!

The Report describes how this standard works in the case of mortgages securitized through Ginnie Mae.  For such mortgage-backed securities (MBS), the servicers are required to advance any missing (i.e., when borrowers fail to make required payments) scheduled interest and principal payments to those MBS investors. If borrowers do not first catch up with the missing payments, the servicers will eventually get reimbursed instead by the government for the outlays.  However, it may take several years for that to happen as such reimbursement only occurs when there is final resolution of the underlying individual distressed loans.¹⁵ Thus, Ginnie Mae MBS bondholders are, as promised, receiving their principal and interest payments on time without interruption, but only because the servicers are fronting the cash to do so.¹⁶

That’s a highly unusual and potentially very large burden to put on an FTP firm. This amounts to NMSs being obligated to potentially lend large amounts of money to the federal government for an extended period to pay for the latter’s credit insurance promise to MBS investors.  However, the Report does not focus separately on each of the three agencies, despite it being quite instructive to do so. 

• Ginnie Mae and its related mortgage insurers,¹⁷ as it is colloquially said in the industry, “have no balance sheet” – i.e., the general ability to borrow to put on assets.  This is because they lack the congressional authorization to do so.  Thus, Ginnie Mae does not have the funds to generally make good on the timely payment of principal and interest to MBS investors, even though that is a key component of the government’s guarantee.  Hence, as described above, the government outsources to servicers – which today means mostly to NMSs – the obligation to make those payments in their stead. However, NMSs have significantly much less financial capacity to do so in comparison to bank servicers, thus raising concerns about their ability to meet those obligations fully and on time.¹⁸
• Freddie Mac, as a GSE, definitely has a very large balance sheet along with a strong financial capacity to fund timely payment of interest and principal.  Thus, it has long required that servicers only make those payments for up to four months, taking on the obligation itself after that point. 
• Fannie Mae, the largest of the three agencies, did not similarly limit the obligation of servicers to four months for such payments. Instead, it historically allowed this obligation to continue until the underlying loans were resolved, just like Ginnie Mae.  This approach seemed inconsistent with Fannie Mae’s strong balance sheet as a GSE.  As a result, when COVID-19 hit and Congress passed the nationwide forbearance program, the FHFA ordered Fannie Mae to instead align with Freddie Mac’s four-month time frame.¹⁹

It is worth noting that the GSE’s four-month obligation remains material to the typical NMS; for servicers of MBS insured by Ginnie Mae, the unlimited obligation to make servicer advances continues to this day.  Thus, NMSs still face being destabilized by potentially large servicer advances, just somewhat less than prior to COVID-19.  Also, looking back, this episode highlighted how change can be implemented relatively easily through the GSEs as compared to Ginnie Mae, since no legislation was required and also since the two GSEs have large balance sheets at their disposal.   

Standard #2:  NMSs are required, in order to receive their primary²⁰ fee revenue, to invest heavily in an asset called ‘mortgage servicing rights’ (MSRs).  However, MSRs are unusually volatile in value and also illiquid, thus significantly eroding the financial stability of NMS companies. Given the evolution of the markets, this practice has proven highly problematic. 

In the early days of agency mortgage securitization in the 1970s, it was established that servicers would be compensated not with an FTP-style fee²¹ but with an economic interest in the loans securitized.²²  Specifically, servicers receive per annum a percentage of the principal amount of the mortgages outstanding, akin to an interest rate.  Currently, the most common rate for such compensation is 0.25 percent.²³  Yet that 0.25 percent is much higher than today’s actual servicing costs for a performing GSE-guaranteed loan portfolio, which is considered in the industry to be under 0.10 percent.²⁴  This means a contract to service mortgages has an economic value that is not just based on the profits made over time on actual servicing operations, but even more based on the “excess” amount of the servicing fee that is to be received per annum (the latter being economically equivalent to a type of bond).  Thus, the servicing contract is a hybrid:  by value, it is primarily a type of bond and only secondarily an operating services contract.  In response, starting in 1995, the accounting authorities began to require that these contracts be put on a servicer’s balance sheet as an asset (called the MSR asset) equal to the estimated market value of the profit flows over the life of the underlying mortgage portfolio. 

As already mentioned, this market value can be quite volatile.  For example, it can significantly change with variations in interest rates (higher rates increase its value) and with credit quality (impaired loans are substantially more expensive to service).  The impact of this volatility on NMS earnings is potentially quite substantial.  According to the Report (p.28), MSR assets have recently risen to comprise about 30 percent of total NMS assets. This is a major reason that NMS companies have a reputation for being financially risky, with comparatively unstable and hard-to-predict earnings.²⁵ In fact, the Report (p. 34) indicates that no large nonbank mortgage company qualifies as “investment grade.”

The MSR asset is even more problematic in that it is rather illiquid, as servicing transfers are known to be time-consuming and operationally complex to carry out.²⁶

The recent rapid growth of the sub-servicing marketplace shows the way

In the mortgage industry, sub-servicing involves a servicer outsourcing the actual operational functions to another firm,²⁷ while still retaining the MSR asset on its balance sheet along with the obligation to make servicer advances.   As described by the Report (p. 8), a sub-servicer is a pure FTP since the original servicer retains the two potentially destabilizing financial burdens described above. Sub-servicers generally get paid in a manner that actually reflects the underlying cost economics of an FTP business:  they receive a certain number of dollars for each performing loan per month, with increasing increments above that level for loans that are increasingly impaired and thus also increasingly more costly to service.

Sub-servicing was often originally undertaken by banks that looked to outsource operational functions to obtain lower costs.²⁸  But, surprisingly, the Report (p. 9) describes how today, of the large market share of servicing held by NMSs, about half is actually held specifically by a type of NMS that it calls a nonbank “passive MSR investor” (e.g., a financial firm such as a real estate investment trust, private equity fund, etc.) which in turn contracts out the actual operational servicing to a sub-servicer. 

This is an extremely significant development in my view.  In essence, the marketplace is responding to the same problem identified in the introduction of this article:  the inappropriateness of having a processing function intermixed with two potentially large and destabilizing financial obligations.  The rise of passive MSR investors to account for such a large share of the NMS market means that these functions are very often no longer mixed together but instead separated out and assigned to specialized firms:  on one side to FTP firms using the sub-servicer business model, and on the other to financial intermediaries better equipped to take on the potentially destabilizing financial obligations as part of their investing in the bond-like excess servicing cashflows.  To me, this shows the way to a different approach to reduce NMS-related financial instability.

A list of pragmatic GSE de-risking solutions

To recap, the core argument of this article is that reforming two key standards in the mortgage securitization industry is the best approach to eliminate NMSs as a source of potential instability. These industry standards currently impose significant financial obligations on what otherwise would be a straightforward FTP business model.  I list below five actions, each of which will reduce the risk of the industry causing a systemic problem, from which policymakers can choose one or more.  If the actions are chosen to collectively have enough impact, the risk of instability should be reduced such that the NMS industry no longer poses a systemic concern.  These solutions focus primarily on two key areas: (1) adopting the sub-servicing business model in lieu of today’s servicing one as much as possible and (2) implementing change via the two GSEs. (A focus on MBS securitized through the smaller Ginnie Mae can follow later, as legislation would likely be required.) Most of the suggested measures can be implemented on a voluntary basis,²⁹ and in the span of just a few years might well very materially reduce the systemic risk identified in the Report.

Servicer advances to become promptly reimbursed.  The objective is to move this function fully to the sub-servicer business model, which means that servicer advances would be promptly reimbursed, typically within one month. This would eliminate the need for NMSs to ever front large amounts of interest and principal payments for the GSEs. 

1. The FHFA would announce that, in order to achieve some of the risk reduction called for in the Report, it is reducing the reimbursement timeline for advances by Fannie Mae and Freddie Mac from four months to the “prompt” standard used in the sub-servicing market. 

The GSEs become large-scale “passive MSR investors.”  NMSs can be substantially de-risked by transitioning further to a sub-servicer business model. In this model, NMS firms sell their MSR on GSE-guaranteed pools of mortgages back to the specific guaranteeing GSE³⁰ in exchange for cash and a reciprocal sub-servicing contract back.  This allows the servicer to continue performing the FTP function that is at the heart of mortgage servicing, leaving borrowers undisturbed.  Fannie Mae and Freddie Mac are extremely well positioned, likely better than almost any passive MSR investor, to take on the MSR asset onto their balance sheets, efficiently funding it and also managing its volatile value. 

1. Sellers of new mortgages to the GSEs can be given the opportunity to sell the related MSR for cash at a negotiated price and enter into a matching sub-servicing contract. Seller participation in this process would be strictly voluntary. 
2. The two GSEs can also apply this approach to current outstanding GSE-guaranteed MBS.  This just requires the GSEs to contact the servicers with an offer to buy the related MSRs. This would also be on a voluntary basis.  I note that if the GSEs offer higher prices, the transfer of risk from the NMSs to the stronger hands of the GSEs should happen more quickly.
3. This process can also apply to new mortgages on an involuntary, i.e., mandatory, basis. To implement this, pricing formulae for MSRs would need to be established – which would require vetting and approval by the FHFA to avoid the GSEs abusing their market power.  This task is made harder by the well-known difficulty of modeling MSR values.³¹

The GSEs replace the historic servicer compensation standard with today’s sub-servicer FTP-style approach.  This change would eliminate the MSR asset altogether, dramatically reducing the risks of NMS industry instability. 

1. Servicer compensation by the GSEs on all new production would mandatorily switch to the sub-servicing cost-based payment approach. 

Policymakers will of course need to extensively study and consult with the mortgage industry, mortgage securities investors and dealers, and other stakeholders on the proposed changes before implementing them.  Many of the alternatives listed above could also be developed further via pilot programs before a broad adoption.  Almost all will encounter opposition to a lesser or greater degree, as virtually any change in America’s complicated housing finance system will lead to some stakeholder group earning less than they currently do or feeling that their notion of how things should work is challenged.  Obviously, the involuntary menu items described above – specifically, the last two on the list – would generate the most pushback.³²

After the selected listed items have been successfully implemented through the GSEs, it would be time to consider what legislation or other changes would be needed for Ginnie Mae and its affiliated insurers to move in a similar direction.  But, as already stated, even if these Ginnie Mae-related changes never occurred, the substantial risk reduction for GSE-guaranteed MBS would likely eliminate the NMS industry as a source of systemic risk. 

Conclusion

The Report has rightly identified a valid concern: as NMSs have become the dominant provider of mortgage servicing, there is a remote but real risk they could become a source of financial system instability.  This article proposes a very different path to solve this problem than that recommended by the Report:  to instead substantially reduce, through changes implemented via Fannie Mae and Freddie Mac (and potentially later through Ginnie Mae), the impact of the two long-standing securitization industry standards that are the root cause of the potential systemic instability. By doing so, NMSs would operate more as pure mortgage FTPs, significantly reducing their ability to impact financial stability.

1 FSOC was created by the Dodd-Frank Act in response to certain regulatory failings revealed by the financial crisis of 2007-2009.

2 See https://home.treasury.gov/system/261/FSOC-2024-Nonbank-Mortgage-Servicing-Report.pdf.

3 These two standards also impact bank servicers, but are far more impactful for NMSs, as explained further below.

4 That a mortgage servicer is meant to be an FTP is demonstrated by a typical definition of what it does: “Servicing consists of collecting loan payments, remitting principal and interest payments to investors, managing escrow funds for the payment of mortgage-related expenses, such as taxes and insurance, performing loss mitigation activities on behalf of investors and otherwise administering [a] mortgage loan servicing portfolio.”  Source:  Annual Report 10-K for 2023 of Mr. Cooper Group, the largest nonbank mortgage company.

5 The term “banks,” for simplicity, is used to denote banks, subsidiaries of bank holding companies, and also other depositories such as savings & loans or credit unions.

6 “Safe and sound” is a standard banking regulatory phrase. It refers to the goal of ensuring that the regulated banks meet strong capital and liquidity requirements and that they undergo regular examinations. The objective is to make sure everything is operating well and operations can continue properly even in stressful environments. 

7 It is clear that the three “inabilities” were much reduced in potential impact when servicers were almost all banks. This is because banks possess greater liquidity strength, a diversified business mix, and higher capital and liquidity requirements.  Also, banks do not go through bankruptcy, but instead an FDIC-led resolution process that allows for continued operations. 

8 Ginnie Mae is the government department through which investors in certain mortgage-backed securities are insured against credit-related losses by the full faith and credit of the U.S. government.

9 This has already been criticized as a “taxpayer bailout,”  See Politico, July 25, 2024, https://www.politico.com/news/magazine/2024/07/25/government-taxpayer-bailout-00166835.

10 In fact, Congress may never get around to passing the required legislation at all, which significantly undermines the overall approach adopted by the Report to address the systemic risk. 

11 The FHFA is the regulator of the two GSEs.  In addition, since 2008, it is also their conservator, which gives it powers to directly control all their operations and decision-making. 

12 A resolution plan, colloquially known as a “living will,” describes a bank’s plan for winding down its affairs in an orderly manner in the event of its grave financial distress or failure. 

13 By contrast, the recommendation related to Ginnie Mae developing counterparty recovery planning is perfectly appropriate but certainly didn’t need FSOC’s involvement to develop.

14 See letter from several trade associations, in particular Attachment 3.  https://www.housingwire.com/wp-content/uploads/2020/03/d315af_183302700622495987393da848602e83.pdf. 

15 See the Report, Page 32.  Note that there is also the issue of who pays property taxes, homeowner’s insurance, etc., to maintain a property when the homeowner is not doing so.  This also can create a great need for servicer advances, but is outside the scope of this paper.

16 Another way to view this is that the government has outsourced one aspect of its guarantee – that payment of scheduled interest and principal are received on time by investors – to the servicers. 

17 Loans securitized via Ginnie Mae are actually first approved and insured by the agency’s affiliated mortgage insurers, the most well-known of which is the Federal Housing Administration (FHA).  Both are units of the U.S. Department of Housing and Urban Development (HUD).  They work together in a complex manner which is not necessary to describe for purposes of this paper. 

18 NMSs, in comparison to banks, do not have as strong liquidity since, among other reasons, they lack a deposit base and do not have access to the Federal Reserve’s discount window.

19 NMSs were able to weather the extraordinary call for servicer advances during COVID-19 despite initial concerns about needing government emergency funding.  This was achieved through a combination of three key factors (1) Fannie Mae reduced its unlimited requirement down to four months, (2) lenders to NMSs expanded their available lines of credit, and (3) an unprecedented refinancing boom that soon occurred generated tremendous cashflow from loan origination (as most NMS firms are also originators).

20 Servicers also earn ancillary fees, largely from borrowers asking for additional services. 

21 An FTP-style fee is described further below. 

22 At the time of those initial securitizations in the early 1970s, “servicing” was mainly just bank back offices doing processing work, and it is doubtful that they even knew the related costs with any precision.  Today, with servicing mostly being done by third parties, costs are very well known. 

23 This is the rate that applies to GSE 30-year fixed-rate mortgages, by far the most common type of mortgage. 

24 Such costs could balloon if credit quality deteriorates, as the unit cost to service impaired loans rises dramatically as impairments get more severe. Thus, some of the excess fee is justified as compensation to take the risk of such deterioration.  In the GFC, apparently, some very distressed portfolios even produced negative margins.  

25 As a recent example of this, Mr. Cooper Group, the largest nonbank mortgage company, reported a net income of $80 million in the third quarter of 2024, down from $204 million in the second quarter. This decline was primarily due to a $126 million reduction in the market value of its MSR asset, even net of hedges.  See https://s203.q4cdn.com/464899268/doc_financials/2024/q3/3Q-24-Earnings-Presentation-FINAL-v2.pdf, page 11. 

26 Unsurprisingly, over the years, there have been discussions to reform the industry’s servicing compensation standard to eliminate its complexity, expense, and instability.  Even the FHFA, during 2010 and 2011, attempted to organize such a reform.  But they were all pursued on a voluntary basis, meaning any proposed reform would only be implemented if all the involved stakeholders agreed, which in practice translated into no stakeholder facing reduced profits or increased risks.   This hurdle has been so high that such reform efforts never reached agreement, so that problems surrounding the MSR asset persist to this day.

27 Servicers can customize the sub-servicing contract to keep certain operational functions for themselves, but most do not. 

28 Servicing operations are considered to have great economies of scale. Thus, unless a bank has a very large portfolio of mortgages, it could outsource to a sub-servicer and reduce its costs. 

29 “Voluntary” means that the transactions described occur in the open market, with both sides agreeing that the result is mutually beneficial on the terms negotiated.  In other words, no party is compelled to enter into the transactions. 

30 In other words, Fannie Mae would purchase the MSR on MBS they have guaranteed, and ditto for Freddie Mac.

31 It is well-established that MSR values are extremely model-dependent, and different firms often come up with significantly different values for the same portfolio. 

32 Pushback can come from many quarters.  The most noteworthy pushback during previous reform attempts came from MBS dealers and investors, who faced transition costs and likely lower profits and/or increased risk.   However, there is no way that the reforms needed to eliminate NMSs as a source of financial instability can be pursued on a voluntary basis, i.e., with no interest group being disadvantaged.  So, the policymakers looking to reduce NMS-related systemic risk will have to expect an outcome that leaves some groups unhappy.