An Inquest into the Subprime Crisis

, image by Paul Townsend

Coroners Inquests in Gloucestershire from The Gloucester Journal 1814

Juan Ospina and Harald Uhlig have posted Mortgage-Backed Securities and the Financial Crisis of 2008: A Post-Mortem to SSRN. Given that the market for private-label MBS pretty much died by 2008, the title is apt. The paper presents a challenge to many of the standard narratives that have developed to explain the causes of the subprime crisis and the broader financial crisis that followed. Other researchers in this area will surely take up the gauntlet thrown down by this paper. Hopefully, we will collectively come up with the right narrative to explain the whole mess. The paper opens,

Gradually, the deep financial crisis of 2008 is in the rearview mirror. With that, standard narratives have emerged, which will inform and influence policy choices and public perception in the future for a long time to come. For that reason, it is all the more important to examine these narratives with the distance of time and available data, as many of these narratives were created in the heat of the moment.

One such standard narrative has it that the financial meltdown of 2008 was caused by an overextension of mortgages to weak borrowers, repackaged and then sold to willing lenders drawn in by faulty risk ratings for these mortgage back securities. To many, mortgage backed securities and rating agencies became the key villains of that financial crisis. In particular, rating agencies were blamed for assigning the coveted AAA rating to many securities, which did not deserve it, particularly in the subprime segment of the market, and that these ratings then lead to substantial losses for institutional investors, who needed to invest in safe assets and who mistakenly put their trust in these misguided ratings.

In this paper, we re-examine this narrative. We seek to address two questions in particular. First, were these mortgage backed securities bad investments? Second, were the ratings wrong? We answer these questions, using a new and detailed data set on the universe of non-agency residential mortgage backed securities (RMBS), obtained by devoting considerable work to carefully assembling data from Bloomberg and other sources. This data set allows us to examine the actual repayment stream and losses on principal on these securities up to 2014, and thus with a considerable distance since the crisis events. In essence, we provide a post-mortem on a market that many believe to have died in 2008. We find that the conventional narrative needs substantial rewriting: the ratings and the losses were not nearly as bad as this narrative would lead one to believe.

Specifically, we calculate the ex-post realized losses as well as ex-post realized return on investing on par in these mortgage backed securities, under various assumptions of the losses for the remaining life time of the securities. We compare these realized returns to their ratings in 2008 and their promised loss distributions, according to tables available from the rating agencies. We shall investigate, whether ratings were a sufficient statistic (to the degree that a discretized rating can be) or whether they were, essentially, just “noise”, given information already available to market participants at the time of investing such as ratings of borrowers.

We establish seven facts. First, the bulk of these securities was rated AAA. Second, AAA securities did ok: on average, their total cumulated losses up to 2013 are 2.3 percent. Third, the subprime AAA-rated segment did particularly well. Fourth, later vintages did worse than earlier vintages, except for subprime AAA securities. Fifth, the bulk of the losses were concentrated on a small share of all securities. Sixth, the misrating for AAA securities was modest. Seventh, controlling for a home price bust, a home price boom was good for the repayment on these securities. (1-2)

De Facto Housing Finance Reform

photo by The Tire Zoo

David Finkelstein, Andreas Strzodka and James Vickery of the NY Fed have posted Credit Risk Transfer and De Facto GSE Reform. It opens,

Nearly a decade into the conservatorships of Fannie Mae and Freddie Mac, no legislation has yet been passed to reform the housing finance system and resolve the long-term future of these two government-sponsored enterprises (GSEs). The GSEs have, however, implemented significant changes to their operations and practices over this period, even in the absence of legislation. The goal of this paper is to summarize and evaluate one of the most important of these initiatives – the use of credit risk transfer (CRT) instruments to shift mortgage credit risk from the GSEs to the private sector.

Fannie Mae and Freddie Mac have significant mortgage credit risk exposure, largely because they provide a credit guarantee to investors on the agency mortgage-backed securities (MBS) they issue. Since the CRT programs began in 2013, Fannie Mae and Freddie Mac have transferred to the private sector a portion of the credit risk on approximately $1.8 trillion in single-family mortgages (as of December 2017; source: Fannie Mae, 2017, Freddie Mac, 2017). The GSEs have experimented with a range of different risk transfer instruments, including reinsurance, senior-subordinate securitizations, and transactions involving explicit lender risk sharing. The bulk of CRT, however, has occurred via the issuance of structured debt securities whose principal payments are tied to the credit performance of a reference pool of securitized mortgages. A period of elevated mortgage defaults and losses will  trigger automatic principal write-downs on these CRT bonds, partially offsetting credit losses experienced by the GSEs.

Our thesis is that the CRT initiative has improved the stability of the  housing finance system and advanced a number of important objectives of GSE reform. In particular the CRT programs have meaningfully reduced the exposure of the Federal government to mortgage credit risk without disrupting the liquidity or stability of secondary mortgage markets. In the process, the CRT programs have created a new financial market for pricing and trading mortgage credit risk, which has grown in size and liquidity over time. Given diminished private-label securitization activity in recent years, these CRT securities are one of the primary ways for private-sector capital market investors to gain exposure to residential mortgage credit risk.

An important reason for this success is that the credit risk transfer programs do not disrupt the operation of the agency MBS market or affect the risks facing agency MBS investors. Because agency MBS carry a GSE credit guarantee, agency MBS investors assume that they are exposed to interest rate risk and prepayment risk, but not credit risk. This reduces the set of parameters on which pass-through MBS pools differ from one another, improving the standardization of the securities underlying the liquid to-be-announced (TBA) market where agency MBS mainly trade. Even though the GSEs now use CRT structures to transfer credit risk to a variety of private sector investors, these arrangements do not affect agency MBS investors, since the agency MBS credit guarantee is still being provided only by the GSE. In other words, the GSE stands in between the agency MBS investors and private-sector CRT investors, acting in a role akin to a central counterparty.

Ensuring that Fannie Mae and Freddie Mac’s credit risk sharing efforts occur independently of the agency MBS market is important for both market functioning and financial stability. The agency MBS market, which remains one of the most liquid fixed income markets in the world, proved to be quite resilient during the 2007-2009 financial crisis, helping to support the supply of mortgage credit during that period. The agency market financed $2.89 trillion of mortgage originations during 2008 and 2009, experiencing little drop in secondary market trading volume during that period. In contrast, the non-agency MBS market, where MBS investors are exposed directly to credit risk, proved to be much less stable; Issuance in this market essentially froze in the second half of 2007, and has remained at low levels since that time.4 (1-2, citations and footnotes omitted)

One open question, of course, is whether the risk transfer has been properly priced. We won’t be able to fully answer that question until the next crisis tests these CRT securities. But in the meantime, we can contemplate the authors’ conclusion:

the CRT program represents a valuable step forward towards GSE
reform, as well as a basis for future reform. Many proposals have been put forward for long-term reform of mortgage market since the GSE conservatorships began in 2008. Although the details of these proposals vary, they generally share in common the goals of

(1) ensuring that mortgage credit risk is borne by the private sector (probably with some form of government backstop and/or tail insurance to insure catastrophic risk and stabilize the market during periods of stress), while

(2) maintaining the current securitization infrastructure as well as the standardization and liquidity of agency MBS markets.

The credit risk transfer program, now into its fifth year, represents an effective mechanism for achieving these twin goals. (21, footnote omitted)

How Tight Is The Credit Box?

Laurie Goodman of the Urban Institute’s Housing Finance Policy Center has posted a working paper, Quantifying the Tightness of Mortgage Credit and Assessing Policy Actions. The paper opens,

Mortgage credit has become very tight in the aftermath of the financial crisis. While experts generally agree that it is poor public policy to make loans to borrowers who cannot make their payments, failing to make mortgages to those who can make their payments has an opportunity cost, because historically homeownership has been the best way to build wealth. And, default is not binary: very few borrowers will default under all circumstances, and very few borrowers will never default. The decision where to draw the line—which mortgages to make—comes down to what probability of default we as a society are prepared to tolerate.

This paper first quantifies the tightness of mortgage credit in historical perspective. It then discusses one consequence of tight credit: fewer mortgage loans are being made. Then the paper evaluates the policy actions to loosen the credit box taken by the government-sponsored enterprises (GSEs) and their regulator, the Federal Housing Finance Agency (FHFA), as well as the policy actions taken by the Federal Housing Administration (FHA), arguing that the GSEs have been much more successful than the FHA. The paper concludes with the argument that if we don’t solve mortgage credit availability issues, we will have a much lower percentage of homeowners because a larger share of potential new homebuyers will likely be Hispanic or nonwhite—groups that have had lower incomes, less wealth, and lower credit scores than whites. Because homeownership has traditionally been the best way for households to build wealth, the inability of these new potential homeowners to buy could increase economic inequality between whites and nonwhites. (1)

Goodman has been making the case for some time that the credit box is too tight. I would have liked to see a broader discussion in the paper of policies that could further loosen credit. What, for instance, could the Consumer Financial Protection Bureau do to encourage more lending? Should it be offering more of a safe harbor for lenders who are willing to make non-Qualified Mortgage loans? The private-label mortgage-backed securities sector has remained close to dead since the financial crisis.  Are there ways to bring some life — responsible life — back to that sector? Why aren’t portfolio lenders stepping into that space? What would they need to do so?

When the Qualified Mortgage rule was being hashed out, there was a debate as to whether there should be any non-Qualified Mortgages available to borrowers.  Some argued that every borrower should get a Qualified Mortgage, which has so many consumer protection provisions built into it. I was of the opinion that there should be a market for non-QM although the CFPB would need to monitor that sector closely. I stand by that position. The credit box is too tight and non-QM could help to loosen it up.

Subprime v. Non-Prime

photo by TaxRebate.org.uk

The Kroll Bond Rating Agency has issued an RMBS Research report, Credit Evolution: Non-Prime Isn’t Yesterday’s Subprime. It opens,

Following the private label RMBS market’s peak in 2007 and the ensuing credit crisis, non-agency securitizations of newly originated collateral have focused almost exclusively on prime jumbo loans. This is not surprising given the poor performance of loosely underwritten residential mortgage loans that characterized certain vintages leading up to the crisis. While legacy prime, in absolute terms, performed better than Alt-A and subprime collateral, it was apparent that origination practices had a significant impact on subsequent loan performance across product types.

Many consumers were caught in the ensuing waves of defaults, which marred their borrowing records in a manner that has either barred them from accessing housing credit, or at best made it extremely challenging to obtain a home loan. Others that managed to meet their obligations have been unable to qualify for new loans in the post-crisis era due to tighter credit standards that have been influenced by regulation.

The private label securitization market has not met the needs of these consumers for a number of reasons, including, but not limited to, reputational concerns in the aftermath of the crisis, regulatory costs, investor appetite, and the time needed for borrowers to repair their credit. The tide appears to be turning quickly, however, and Kroll Bond Rating Agency (KBRA) has observed the re-emergence of more than a dozen non-prime mortgage origination programs that intend to use securitization as a funding source. Of these, KBRA is aware of at least four securitization sponsors that have accessed the PLS market across nine issuances, two of which include rated offerings.

Thus far, KBRA has observed that today’s non-prime programs are not a simple rebranding of pre-crisis subprime origination, nor do they signal a return to the documentation excesses associated with “liar loans”. While the asset class is meant to serve those with less pristine credit, and can even have characteristics reminiscent of legacy Alt-A, it is expansive, and underwriting practices have been heavily influenced by today’s consumer-focused regulatory environment and government-sponsored entity (GSE) origination guidelines. In evaluating these new non-prime programs, KBRA believes market participants should consider the following factors:

■ Loans originated under sound compliance with Ability-To-Repay (ATR) rules should outperform 2005-2007 vintage loans with similar credit parameters, including LTV and borrower FICO scores. The ATR rules have resulted in strengthened underwriting, which should bode well for originations across the MBS space. This is particularly true of non-prime loans, where differences in origination practices can have a greater influence on future loan performance.

■ Loans that fail to adhere to GSE guidelines regarding the seasoning of credit dispositions (e.g. bankruptcy, foreclosure, etc.) on a borrower’s credit history should be viewed as having increased credit risk relative to those with similar credit profiles that lack recent disposition activity. This relationship likely depends on, among other things, equity position, current FICO score, and the likelihood that any life events relating to the prior credit issue remain unresolved.

■ Alternative documentation programs need to viewed with skepticism as they relate to the ATR rules, particularly those that serve borrowers with sub-prime credit histories. Although many programs will meet technical requirements for income verification, it is also important to demonstrate good faith in determining a borrower’s ability-to-repay. Failure to do so may not only result in poor credit performance, but increased risk of assignee liability.

■ Investor programs underwritten with reliance on expected rental income and limited documentation may pose more risk relative to fully documented investor loans where the borrower’s income and debt profile are considered, all else equal. (1, footnotes omitted)

I think KBRS is documenting a positive trend: looser credit for those with less-than-prime credit is overdue. I also think that KBRS’ concerns about the development of the non-prime market should be heeded — ensuring that borrowers have the ability to repay their mortgages should be job No. 1 for originators (although it seems ridiculous that one would have to say that). We want a mortgage market that serves everyone who is capable of making their mortgage payments for the long term. These developments in the non-prime market are most welcome and a bit overdue.

Treasury Gives RMBS a Workout

The Treasury has undertaken a Credit Rating Agency Exercise. According to Michael Stegman, Treasury

recognized that the PLS market has been dormant since the financial crisis partly because of a “chicken-and-egg” phenomenon between rating agencies and originator-aggregators. Rating agencies will not rate mortgage pools without loan-level data, yet originator-aggregators will not originate pools of mortgage bonds without an idea of what it would take for the bond to receive a AAA rating.

Using our convening authority, Treasury invited six credit rating agencies to participate in an exercise over the last several months intended to provide market participants with greater transparency into their credit rating methodologies for residential mortgage loans.

By increasing clarity around loss expectations and required subordination levels for more diverse pools of collateral, the credit rating agencies can stimulate a constructive market dialogue around post-crisis underwriting and securitization practices and foster greater confidence in the credit rating process for private label mortgage-backed securities (MBS). The information obtained through this exercise may also give mortgage originators and aggregators greater insight into the potential economics of financing mortgage loans in the private label channel and the consequent implications for borrowing costs.

While this exercise is very technical, it contains some interesting nuggets for a broad range of readers. For instance,

The housing market, regulatory environment, and loan performance have evolved significantly from pre-crisis to present day. Credit rating agency models appear to account for these changes in varying ways. All credit rating agency models incorporate the performance of loans originated prior to, during, and after the crisis to the degree they believe best informs the nature of credit and prepayment risk reflected in the market. Credit rating agency model stress scenarios may be influenced by loans originated at the peak of the housing market, given the macroeconomic stress and home price declines they experienced. The credit rating agencies differ, however, in how their models adjust for the post-crisis regime of improved underwriting practices and operational controls. Some credit rating agencies capture these changes directly in their models, while other credit rating agencies rely on qualitative adjustments outside of their models. (10)

It is important for non-specialists to realize how much subjectivity can be built into rating agency models. Every model will make inferences based on past performance. The exercise highlights how different rating agencies address post-crisis loan performance in significantly different ways. Time will tell which ones got it right.

The Cost of Doing Nothing

Yesterday, I wrote about the Securities Industry and Financial Markets Association (SIFMA)’s FHFA comment letter. Today I write about SIFMA’s comment letter in response to Treasury’s request for input relating to the future of the private-label securities market. Like the FHFA comment letter, this one is written with the concerns of SIFMA’s members in mind, no others, but it identifies many of the structural problems that exist in the housing finance system today.

If I were to identify a theme of the comments, it would be that the federal government has not moved with sufficient speed to establish a well delineated infrastructure for the housing finance market. Some commentators identify benefits of a slow approach — time to get consensus, time to get rules right, time to for trial and error before committing for the long term. Few identify the costs of regulatory uncertainty — failure to get buy-in for capital-intensive ventures, atrophy of existing resources, limited investor interest.

Now, SIFMA’s members want a vibrant private-label MBS market to make money. But a vibrant private-label MBS market is also good for the overall health of the mortgage market as it spreads risk to private MBS investors and reduces the footprints of the gargantuan GSEs and the government’s own FHA. After all, most of us want the private sector taking a lot of the risk, not the taxpayer.

Notwithstanding the strengths of SIFMA’s comment letter to Treasury in critiquing the status quo, I will highlight a few passages from it that hit a false note. The first relates to the role that private-label securities (PLS) have played

in funding mortgage credit where loan size or other terms may differ from those available in the Agency markets, or where economics dictate that PLS execution is superior. The PLS market may also be more innovative and flexible than the Agency markets in adapting to economic conditions or consumer preferences, or to changing capital markets appetite. (3)

This innovation has obviously cut both ways in terms of introducing new products that can help expand access to credit as well as expand access to credit on abusive terms. The latter way seems to have predominated during the most recent boom in PLS MBS.

The second one relates to assignee liability. SIFMA states that

Investors are concerned with the prospect of assignee liability stemming from violations of the ability-to-repay rules contained in Title XIV of Dodd-Frank and embodied in the CFPB’s implementing regulations. SIFMA has raised concerns with assignee liability in many forms over the years based on the fact that mortgage investors are not at the closing table with the lender and borrower, and should not be held liable for defects of which they have no knowledge or ability to prevent. While efforts were made by policymakers to provide some level of certainty through the inclusion of safe-harbor provisions, no safe harbor is entirely safe, and it is important to note that none of these provisions have been tested in court. It will be in litigation where the market learns the exact boundaries of the protections provided by any safe harbor. This potential liability for investors is likely to reduce the availability of higher-priced QM loans and non-QM loans, all else equal, due to higher required yields to compensate for the increased risk. (5-6)

This focus on assignee liability seems to be a red herring, one that SIFMA has floated for years. The risk from assignee liability provisions is not limitless and it can be modeled. Moreover, the notion that investors should face no liability because they are not at the closing table is laughable — without them, there would be no closing table at all. They paid for it, even if they are not in the room when the closing takes place.

The last one relates to the threatened use of eminent domain by some local governments to take underwater mortgages and refinance them to reflect current property valuations:

Investors have significant concerns with, and continuing distrust of the policy environment because of a sense that rules have been and continue to be changed ex-post. The threat by certain municipalities to use eminent domain to seize performing mortgage loans has been a focus of MBS investors for the last two years and would introduce a significant new risk into investing in PLS. These municipalities propose to cherry-pick loans from PLS trusts and compensate holders at levels far below the actual value of the loans. SIFMA’s investor members view such activity as an illegal taking of trust assets, and successful implementation of these plans would severely damage investor confidence in investing in PLS. (6)

This is another red herring as far as I am concerned.  The use of eminent domain is not an ex post legal maneuver. Rather, it is an inherent power of government that precedes the founding of this country. I understand that MBS investors don’t like it, but it is not some kind of newfangled violation of the rule of law as many investor advocates have claimed.

Notwithstanding its flaws, I recommend this letter as a trenchant critique of the housing system we have today.

Housing Finance Abhors A Vacuum

The Securities Industry and Financial Markets Association (SIFMA) released their comment letter to the Federal Housing Finance Agency’s request for input relating to the role of the Fannie and Freddie guarantee fee (g-fee) in the housing finance market. While clearly reflecting the concerns of SIFMA’s members, the letter provides a thoughtful take on the complexities of the housing finance system. SIFMA writes,

Policymakers should not assume that increases in g-fees alone will lead to a significant increase in PLS issuance. Specific decisions on best execution for a given loan vary depending on the terms of the loan being originated. In some instances, a portfolio purchase may offer best execution, and in other instances the GSEs, private label MBS (PLS) or FHA may be optimal. Taken wholly in isolation, we do agree that increases in guarantee fees should cause originators to look toward other avenues to fund loans – in their portfolios, FHA, or in PLS. However, it is not so simple that an across the board increase in guarantee fees will result in a corresponding uptick in private-label securitization. To the extent GSE securitization becomes more expensive for issuers, PLS are one of a number of options, and not necessarily the most attractive in all instances. Today bank portfolios offer a more attractive funding alternative to the GSEs than PLS for most institutions. Of course, the appetite of banks for loans held in portfolio will vary with economic and regulatory conditions, and cannot always be assumed to comprise a certain percentage of the market.

There are also a number of reasons that increases to g-fees will not directly lead to increased PLS issuances that are not precisely quantifiable or directly related to cost. PLS issuers and investors face uncertainty as to the future shape of the mortgage market and questions related to compliance with the future regulatory regime. The re-regulation of the mortgage and securitization markets is not complete, and a number of consequential rulemakings are incomplete. These include but are not limited to risk retention and proposed revisions to the SEC’s Regulation AB. The final form of the definition of QRM and the rest of the risk retention rules will directly impact the economics of securitization. Regulation AB will impact the offering process, disclosure practices, and require fairly massive infrastructure adaptation at many RMBS issuers and sponsors. Of course, given that final rules are not available for any of these items, issuers and sponsors cannot begin this work. In this environment of uncertainty, it is difficult and indeed may be unwise for issuers or investors to expend resources to develop long-term issuing and investment platforms.

*     *     *

For these reasons, we do not believe FHFA or other policymakers should look at increases to GSE g-fees in a vacuum, and must consider them within the broader context of mortgage finance conditions. (6-7, footnotes omitted)

SIFMA is right to emphasize the regulatory uncertainty that its members face.  The federal government has not done enough to address this.  Housing finance, like nature, abhors a vacuum.  More on this tomorrow.