Wallison is far too Kind to Fannie and Freddie

By William K. Black

It is easy to understand why Commissioner Wallison’s lengthy dissent to the report of the Financial Crisis Inquiry Commission has received such poor reviews.  The first page of his dissent [p. 443] insults Congress, President Obama, former Congressman Rahm, the Democratic Commissioners, the Commission’s staff, and Democratic Congressmen in 1977, particularly Representative Frank.  His dissent is partisan and unprofessional.  It is also long.

Despite those defects, however, Wallison has the virtue of emphasizing the key fact about the crisis that is most often misunderstood.  The conventional economic wisdom starts with the effort to explain a mystery – how could such a relatively small number of subprime loans have caused the Great Recession?  Wallison’s dissent stresses that there were in fact enormous numbers of nonprime loans.  The data on nonprime loans demonstrate several other points

  • Most of the nonprime loans were fraudulent “liar’s” loans
  • The industry knew that such loans caused staggering losses
  • Fannie, Freddie, large commercial banks, and large investment banks frequently purchased large amounts of liar’s loans and CDOs backed by liar’s loans
  • The share of total loans composed of nonprime loans grew rapidly in 2004-2007 – despite urgent, stark warnings from the FBI and the mortgage banking industry’s own anti-fraud experts that mortgage fraud was “epidemic” and would cause an “economic crisis” (FBI September 2004) and that liar’s loans were “an open invitation to fraudsters” with a fraud incidence of 90 percent (MARI 2006).  “Eighth Periodic Mortgage Fraud Case Report to Mortgage Bankers Association” (April 2006).  The nonprime lenders (and purchasers) ignored the warnings – the rapid growth of nonprime loans continued until the secondary market in nonprime loans collapsed.
  • Fannie, Freddie, and other large, publicly-traded holders of nonprime loans and CDOs commonly did not disclose honestly these holdings
  • The holders of nonprime loans and CDOs reduced their general provisions for loan losses as their portfolios of nonprime loans increased

Wallison ignores these facts, however, because they refute his overall theory of the crisis. Wallison chastises the Democratic members of the Commission for not giving more attention to the work of Edward Pinto, a lawyer who is Wallison’s colleague at AEI.  

One glaring example will illustrate the Commission’s lack of objectivity.  In March 2010, Edward Pinto, a resident fellow at the American Enterprise Institute (AEI) who had served as chief credit officer at Fannie Mae, provided to the Commission staff a 70-page, fully sourced memorandum on the number of subprime and other high risk mortgages in the financial system immediately before the financial crisis. In that memorandum, Pinto recorded that he had found over 25 million such mortgages (his later work showed that there were approximately 27 million). Since there are about 55 million mortgages in the U.S., Pinto’s research indicated that, as the financial crisis began, half of all U.S. mortgages were of inferior quality and liable to default when housing prices were no longer rising [p. 448].

The Commission report criticizes Pinto’s classifications.  It finds that delinquency rates on loans held by Fannie and Freddie that Pinto argues are equivalent in risk to nonprime loans are far lower than for nonprime loans held by non-GSEs [p. 219].  As the Commission concluded, this suggests that Pinto classified too many of the loans in Fannie and Freddie’s portfolio as nonprime.  The other Republican Commissioners do not cite Pinto’s work.  Pinto’s claim that nonprime loans were enormous would falsify the other Republican Commissioners’ principal argument in their dissent.  The Commission staff reviewed and responded substantively to Pinto’s work in the report – it is Wallison’s Republican colleagues who ignored Pinto.

I have serious concerns about Pinto’s work attempting to quantify nonprime loans and I believe his work on the Community Reinvestment Act (CRA) is very poor – but I also think he raises a critical question about the number of nonprime loans.  Pinto’s attempt to quantify the number of liar’s and subprime loans and the amount of those loans held by Fannie and Freddie, was an impossible task given the data available to him.  The data available to him – to all of us – are exceptionally poor.  
First, there never was an official definition of any of the three major classes of mortgage loans – prime, subprime, and “alt a” (aka stated income, NINJA, or liar’s loan).  Second, the categories are not exclusive, i.e., large numbers of liar’s loans were also subprime.  Third, the classification of loans by the FDIC and the OTS in its data base is false.  The data base classifies loans based on FICO score and treats subprime and liar’s loans as mutually exclusive categories.  Fourth, the firms holding nonprime loans had powerful incentives to misclassify the assets as “prime” loans.  Fannie, Freddie, and Lehman all called their liar’s loans “prime” loans in their financial reports.  Fifth, the information provided with nonprime loans the loans was frequently false, e.g., the borrower’s income and the value of the home were often inflated.  Sixth, FICO scores are inherently unreliable as a means of underwriting home mortgage lending.  A borrower could “rent” a straw’s FICO score or “improve” his FICO score.  A FICO score does not demonstrate that a borrower is capable of repayment.  Seventh, loan quality can vary greatly within a category.  A subprime loan with effective credit enhancements (admittedly, uncommon) was far less risky than a subprime loan with a simultaneous second lien loan purportedly secured by the same home. 
It appears that Pinto fell into some of these methodological traps.  He seems to have assumed that subprime loans could not also be liar’s loans.  Credit Suisse’s survey of 2006 originations found that “roughly 50% of all subprime borrowers in the past two years have provided limited documentation regarding their incomes.”  “Mortgage Liquidity du Jour: Underestimated no More.” (March 12, 2007).  Credit Suisse reported that “stated income” (aka: liar’s loans) constituted 49% of new mortgage originations in 2006.  The 49% figure is inconsistent with Pinto’s estimate that nonprime loans constitute 49% of total mortgages outstanding, but it represents an enormous expansion of nonprime lending that hyper-inflated the real estate bubble.  It also represents a staggering amount of incidence of mortgage fraud, since mortgage fraud was common in subprime loans and endemic in liar’s loans (see my testimony before the Commission and subsequent columns on that subject).   
Credit Suisse explained that alt-a loans, the most common form of liar’s loans, did not simply become vastly more common, but also became far more likely to default because of the nature of their loan terms.
While credit risk in this segment is often downplayed given the better credit profile of Alt-A borrowers relative to subprime borrowers (i.e. better FICO scores, lower CLTVs), we believe that the significant growth in this segment resulting from its exposure to exotic mortgages leaves the Alt-A mortgage market particularly susceptible….
[S]tated income loans represented a staggering 81% of total Alt-A purchase originations in 2006, up significantly from 64% just two years earlier. These loans are also sheepishly referred to as “liar loans” by many in the industry due to the propensity for borrowers to exaggerate their income on loan applications. In addition, the combined loan to value on Alt-A purchase originations was 88% in 2006, with 55% of homebuyers taking out simultaneous seconds (piggybacks) at the time of purchase. Investors and second home buyers represented approximately 22% of Alt-A purchase originations last year, which is the largest non-owner occupied share among the various segments of the mortgage market. Adding to the risk is the fact that 1-year hybrid ARMs represented approximately 28% of Alt-A purchase originations in 2006, setting the stage for considerable reset risk. The average loan size of Alt-A mortgages backing MBS in 2006 was roughly $287,700, while the average FICO score of an Alt-A borrower last year was 717.

Note the average loan size and FICO score for Alt-A borrowers – these were typically not loans to working class homeowners with known credit defects.  Lenders and their agents (principally loan brokers) routinely inflated the borrowers’ “stated income” – making it even less likely that the loans would be considered to be made to below median-income borrowers

The same Credit Suisse report explained why so many liar’s loans were being made and what effect the loans were having on the size of the particular housing bubbles.

[M]any of the states that had the greatest share of Alt-A mortgages in 2005 have also served as the primary growth engines for the major homebuilders in recent years. We estimate that Nevada, California, Arizona, Florida and Virginia had the greatest share of Alt-A originations in 2006. These five states are also the top five EBIT [Earnings Before Interest and Tax] generators for our homebuilding universe, representing roughly 75% of total operating profit in 2005. In a survey of our private homebuilders, our contacts confirmed that the Alt-A market is a significant portion of their overall business, representing 18% of home sales, on average, in 2006. In addition, our builder contacts specifically operating in Nevada (30% Alt-A share), California (28%) Florida (27%), and Arizona (20%) confirm that those states have an above average concentration of Alt-A loans of the overall mortgage pie, in-line with our state-by-state estimates. A few builders out west indicated that Alt-A represents up to 90% of their overall business. Suffice to say, any credit tightening in this segment of the market will likely have a negative impact on homebuilder profits.
Pinto was correct to try to estimate the total number of nonprime loans originated by year and who held the loans.  (His failure to look intensively at the purchases and holdings of nonprime mortgages – and their timing – by U.S. investment banks and foreign parties (neither of which was subject to the CRA) was analytically unsound – an unfortunate result of his holy war against Fannie and Freddie and the CRA.)  Pinto could not conduct a real investigation of Fannie and Freddie.  He looked at their financial statements and accompanying disclosures and publicly available information about Fannie and Freddie. 
The Federal Housing Finance Agency (FHFA), however, could find out the truth about Fannie and Freddie’s portfolio.  The FHFA, and the nation, have an urgent need to find the true condition of Fannie and Freddie and what caused their catastrophic failures.  This would be true even if they had no ability to “put” the fraudulent loans back to the sellers.  The fact that Fannie and Freddie have the ability to put the fraudulent loans back to the sellers means that conducting the factual investigations should be Fannie and Freddie’s dominant priority.
On January 12, 2010, Eliot Spitzer, Frank Partnoy, and I wrote a short open letter to the Commission “10 Questions the Financial Crisis Commission Must Ask.”  
We stressed that AIG, Fannie, and Freddie (each of which the public (in)effectively owns) were the treasure trove essential to the success of the Commission’s investigation.
The FCIC has not used subpoena authority or voluntary requests for information to obtain the background information essential in order to hold a real investigative hearing. In particular, it has not obtained AIG (and Fannie and Freddie’s) emails and other critical internal documents such as their financial models, internal accounting records, and loss reserve data that are readily available and vital to understand what caused the crisis. Any aircraft crash investigator knows how critical it is to find the “black box” that records the information that is typically essential to finding the cause. In the financial context, these AIG, Fannie & Freddie emails and internal accounting and risk records are the “black box” that any competent investigator would demand to review.

Fannie and Freddie’s “internal accounting records”, “financial models”, and “loss reserve data”, collectively, are precisely what the Commission needed to conduct a reliable study of Fannie and Freddie’s actual nonprime holdings.  The fact that Fannie and Freddie’s senior officers have not conducted such a study tells us that they need to be replaced.  The fact that the FHFA’s senior leaders did not require that Fannie and Freddie’s leaders to provide such a study tells us that FHFA’s senior leaders needed to be replaced.  If the FHFA did not trust Fannie and Freddie’s leaders to conduct the study then the FHFA should have replaced the leaders and conducted their own study.  Similarly, FCIC should have required Fannie, Freddie, and/or the FHFA to provide reliable data on their nonprime loans. 
So what did Pinto say about liar’s loans – the loans that according to the data were increasingly used to finance home buyers and speculators and hyper-inflate the bubble?  He testified before House on December 9, 2008 [page references are from a copy of his testimony on AEI’s website]:

[T]he Alt-A or “liar” loan is generally not classified as subprime, because the FICO score of the borrower was generally above 660, but this loan was the favorite of the real estate speculator, and are currently defaulting at rates approaching those of subprime loans [p. 2].

Pinto knew the loans were fraudulent, for he called them “liar” loans.  He knew that they were “the favorite of the real estate speculator.”  These were not loans to borrowers with below median incomes.  Loans to speculators don’t qualify for affordable housing goals.  Inflating the borrower’s income is the last thing lenders would do if the goal of the loan was to qualify for affordable housing goals.  Pinto also knew what fraudulent loans inherently cause – catastrophic losses.  Pinto also testified that Fannie and Freddie purchased huge amounts of liar’s loans – but deceptively classified the great bulk of them as “prime” loans – which would be insane if the purpose of purchasing the loans was to help Fannie and Freddie meet affordable housing goals.  Pinto testified that Fannie and Freddie purported to justify this deception by simply adopting the seller of the loans misclassification of the loan as “prime” [p. 3].   But that would be insane if the lenders were making the loans to qualify for affordable housing treatment.   Pinto notes that Freddie knew from prior loss experience that making large amounts of nonprime loans would cause severe losses [p. 3]. 

Pinto estimated that Fannie and Freddie held “34% of all the subprime loans and 60% of all Alt-A loans outstanding” [p. 7].  Pinto seems to have treated subprime loans as non-liar’s loans, but that is clearly incorrect.  I cited Credit Suisse’s finding that by 2005 and 2006, half of all subprime loans were also stated income (liar’s loans).  The presence of such large amounts of Alt-A loans is one of the demonstrations that Pinto, Wallison, and the Republican Commissioners’ “Primer” are flat out wrong to claim that it was affordable housing goals that drove Fannie and Freddie’s CEOs’ decisions to purchase loans they knew would cause the firms to fail.  That claim doesn’t pass any logic test.  One of its unobvious flaws is that no one was making Fannie and Freddie buy liar’s loans.  For the reasons I’ve explained, and Pinto admits, Fannie and Freddie actions with respect to liar’s loans were the opposite of what they would have been if they were trying to demonstrate that the loans were made for affordable housing purposes.  This is the best, indeed the only, evidence Pinto cites to show a link between liar’s loans and the HUD goals:

“The Alt-A business makes a contribution to our HUD goals.” Internal Freddie Mac email from Mike May to Dick Syron, dated October 6, 2004.  FMACOO13694

Yes, some Alt-A loans doubtless did count toward the HUD goals.  But massive amounts did not.  According to Pinto’s numbers, Fannie and Freddie’s CEOs deliberately purchased extraordinary amounts of Alt-A loans that they knew would not qualify for affordable housing goals and would cause massive losses that would destroy Fannie and Freddie.  Pinto’s theory is that absent the HUD goals Fannie and Freddie would not have purchased liar’s loans.  His data refute his theory.   Moreover, Fannie and Freddie acted to minimize the number of liar’s loans that would qualify by (1) buying loans with grossly inflated “stated income” and (2) misclassifying the loans as prime.  Pinto’s grand conspiracy theory is that Fannie and Freddie created the HUD goals to protect itself from President Bush.  Pinto claims that Fannie and Freddie sought to emphasize at all times their critical role in aiding affordable housing.  But why did they misclassify their loans so that they would appear to make dramatically fewer (Pinto says only one-quarter the reality) nonprime loans to less wealthy Americans if their brilliant political strategy was to do the opposite?  Pinto’s data falsify his, and Wallison’s, claims view that the housing goals warped Fannie and Freddie into the Great Satans. 

I have emphasized that we, the West Region of OTS, used our normal supervisory powers to kill an earlier wave of liar’s loans being made by California S&Ls in 1990-1991.  Pinto adds to this point by noting that:  “In the early-1990s Fannie and Freddie publicly announced they were no longer buying low doc/no doc loans because they were too risky” [p. 9].  This confirms the point we’ve long made – it didn’t require any genius on our part to kill liar’s loans.  Bankers have known for hundreds of years that making large liar’s loans creates intense “adverse selection” and guarantees catastrophic losses.  That is further proof that the Commission report got one of its central points correct – this crisis could have been stopped.  We, and Fannie and Freddie, proved that in the early 1990s by preventing exactly this crisis – the beginnings of an epidemic of liar’s loans

Pinto then testified about an even more complicated conspiracy theory

By the early part of this decade, the GSEs realized that the private sector was beating them in terms of share and, default risk notwithstanding, these subprime and Alt-A loans were to affordable housing “goal rich” to ignore [p. 11].

Pinto’s conspiracy theory and English usage are convoluted, but after several readings I interpret his argument as follows: 

1.     The mortgage bankers, mortgage brokers, investment banks, and commercial bank affiliates (not subject to affordable housing goals and virtually unregulated at the federal level – collectively, the Shadow Banking System) dominated subprime and liar’s loans
2.     As the Shadow banking investment banks rapidly increased the primary and secondary market in nonprime loans, Fannie and Freddie lost market share
3.     Therefore, Fannie and Freddie convinced Congress to increase their affordable housing goals in order to increase their political power.  They sought to meet their affordable housing goals by purchasing large amounts of subprime and liar’s loans [p. 11].

Wallison’s theories about Fannie and Freddie causing the crisis have been inconsistent over time and are logically incoherent for many reasons.  I’ll make a broader response in future columns, but I’ll make only a few points here. 

  • The California S&Ls that began to do large amounts of liar’s loans in 1990-1991 did not do so because of the CRA or any other form of affordable housing goal.  They did it because they were accounting control frauds following the four-part recipe for creating stellar short-term reported income and maximizing their CEOs’ compensation.  By making liar’s loans to those who would often be unable to repay their loans, both S&Ls were able to grow rapidly by making loans at premium yields.  This, along with providing only trivial loss reserves and extreme leverage, produced a “sure thing” (Akerlof & Romer 1993) of very high reported profits in the short-term.
  • When Long Beach Savings and Guardian Savings did large numbers of liar’s loans we (OTS-West Region) did not praise them for CRA performance – we took enforcement actions against their senior managers.
  • Long Beach Savings and Guardian Savings’ CEOs responded by starting mortgage banking firms precisely because they would no longer be subject to OTS-West Region’s jurisdiction.  As mortgage bankers, they had no CRA or affordable housing obligations or guidelines, yet they increased enormously the number of nonprime loans they made.  Again, they were maximizing short-term reported accounting income.  Long Beach became Ameriquest – notorious for its nonprime lending abuses.
  • The Shadow Banking participants that made large amounts of subprime loans in the late 1990s were not subject to the CRA and made the loans for the same reason as Long Beach.
  • The Shadow Banking participants that started the secondary market in nonprime mortgage loans were not subject to the CRA and created the market to achieve high reported accounting income and executive compensation. 
  • The Shadow Bank system did not pose an economic threat to Fannie and Freddie.  Losing market share to a competitor that will fail – and liar’s loans guarantee that mortgage lenders will fail – is a good thing for an honest competitor. 
  • If Fannie and Freddie’s controlling officers were honest, the Pinto/Wallison conspiracy theory makes no sense, for it would be suicidal. 
  • If Fannie and Freddie were accounting control frauds, then their behavior in going heavily into nonprime was a “sure thing” that was highly profitable for its senior managers.   

Pinto’s testimony goes on to explain how Fannie and Freddie’s senior managers acted in a manner that is sane only if the firms were control frauds.  As early as 1999 [p. 11]:

“Freddie Mac has found that 65% of its fraud cases involve loans produced by third-party originators [For 1999 OHFEO reported that third-party originators, ie. brokers, had a 26% market share with the GSEs.]….  Independent mortgage brokers account for 32% of the fraud cases’ while banks are the remaining 3%. The majority of the fraud – 60% — comes from defective loans.”

Pinto finds other behavior by Fannie and Freddie’s senior management irrational – which it would be for an honest firm.

Adding to this bias in favor of mortgage broker and mortgage banker sourced business was the fact that Fannie and Freddie offered its best pricing to its largest (and riskiest) customers, (i.e. Countrywide, Indy Mac) while offering much worse pricing to customers, i.e. community banks, with proven track records of delivering high quality loans done the traditional way [p.12].”
Countrywide and IndyMac, of course, offered higher yielding loans for sale to Fannie and Freddie than did the community banks.  That maximized Fannie and Freddie’s reported (albeit fictional) income and their senior executives’ compensation.  The National Commission on Financial Institution Reform, Recovery and Enforcement (NCFIRRE) understood this dynamic because it understood the recipes for accounting fraud.  Lending to the uncreditworthy allows exceptional growth while charging a higher interest rate.  The combination maximizes accounting income.  As James Pierce, Executive Director of the National Commission on Financial Institution Reform, Recovery and Enforcement (NCFIRRE) explained:
Accounting abuses also provided the ultimate perverse incentive:  it paid to seek out bad loans because only those who had no intention of repaying would be willing to offer the high loan fees and interest required for the best looting.  It was rational for operators to drive their institutions ever deeper into insolvency as they looted them [NCFIRRE 1993, pp. 10-11].
(Parenthetically, the Commission report and dissents do not appear to cite any of NCFIRRE’s findings.  Wallison quotes the famous warning about those that fail to learn the lessons of the past, but doesn’t follow the advice.)
Pinto understands that Fannie and Freddie were engaged in accounting fraud.
[A]fter their accounting scandals in 2003 and 2004, they were afraid of new and stricter regulation. By ramping up their affordable housing lending, they showed their supporters in Congress that they could be major sources of affordable housing financing.
This was not a failure of the free market. It is a failure of Congress and the ill-conceived regulatory regime it implemented [p. 13].
Pinto and Wallison know that Fannie and Freddie engaged in accounting control fraud in the early 2000s.  (I was an expert witness for OFHEO in its enforcement action against Fannie’s former CEO, Franklin Raines.)  Pinto and Wallison know that the SEC charged that the reason they engaged in the accounting fraud was to enrich their senior officers.  They know that Fannie and Freddie were caught at the fraud and the fraud scheme they were using – very rapid growth of portfolio in order to take interest rate risk (with losses hidden by abusive hedge accounting) – was ended just before Fannie and Freddie decided to purchase far greater amounts of nonprime loans, particularly liar’s loans.  Pinto and Wallison know that OFHEO restricted Fannie and Freddie’s growth.  Fannie and Freddie’s controlling officers, were they to renew the accounting control fraud, would have to find a way to increase yield sharply without growing the portfolio rapidly.  The obvious answer was to purchase much higher yielding loans – nonprime loans – and provide only trivial allowances for loan losses.  Pinto and Wallison, however, cannot even conceive that Fannie and Freddie’s senior managers might renew their accounting fraud. 
As to Pinto’s claim that Fannie and Freddie do not represent a failure of the “free market,” it turns out that he answers that point nicely in one of his attached exhibits.  Pinto writes that Fannie and Freddie’s response to efforts to regulate them:
[W]as crony capitalism at its worst. The mere fact that Congress continued to remain opposed to real reform after both Fannie and Freddie experienced massive accounting scandals in the early part of this decade is proof positive. Fannie and Freddie had gotten so powerful that they felt that they should be able to dictate the terms of their own reform to Congress or block the reforms if they did not like them.
Amen.  When private corporations like Fannie and Freddie become enormous they do gain extraordinary political as well as economic power.  This is the American version of “crony capitalism at its worst.”  We need to get rid of the systemically dangerous institutions (SDIs) that loot with impunity and we need to prosecute the senior officers leading the accounting fraud, including the senior officers of Fannie and Freddie.  Why aren’t Pinto and Wallison calling for those prosecutions?  Pinto is correct, we don’t have “free enterprise” in broad sectors of our economy and the results have been horrific.
Wallison and Charles Calomiris (long time co-directors of AEI’s financial deregulation project) advanced this same self-bondage theory of Fannie and Freddie’s actions in a September 2008 paper entitled:  The Last Trillion-Dollar Commitment: The Destruction of Fannie Mae and Freddie Mac.
The central problem was their dependence on Congress for continued political support in the wake of accounting scandals in 2003 and 2004. To curry favor with Congress, they sought substantial increases in their support of affordable housing, primarily by investing in risky and substandard mortgages between 2005 and 2007.
There’s a much simpler explanation, one that doesn’t require ornate conspiracies or irrational actions by Fannie and Freddie’s CEOs – Fannie and Freddie renewed their accounting control fraud and enriched their senior officers.  After all, Wallison and Calomiris complain bitterly about the weak response to Fannie and Freddie’s accounting control frauds, decrying “GSE immunity to accounting scandal….”  They claim that the decisive break in Fannie and Freddie’s prior behavior of largely standing by while the Shadow Banking sector made over a trillion dollars in nonprime loans was reaction to the discovery of their accounting control fraud.
Instead, it seems likely that the event responsible for the GSEs’ change in direction and culture was the accounting scandal that each of them encountered in 2003 and 2004.
That makes sense.  OFHEO responded to those frauds not by cleaning house, but by the selective removal of a few of the most senior officers.  The corrupt cultures and the executive compensation systems that created the perverse incentives to engage in accounting control fraud remained in place.  What changed was that OFHEO added two operational constraints – it restricted the growth of the portfolio and it continued to look closely at interest rate risk and hedging.  Fannie and Freddie, therefore, could not continue to use their prior accounting scam – extreme growth, the deliberate exposure to serious interest rate risk, and abusive hedge accounting.  There was one obvious way left to dramatically inflate yield – purchase nonprime loans and CDOs with high nominal yields and provide only trivial allowances for loan losses.  Fannie and Freddie could seek much greater yields without substantial growth if they took the enough lower yield mortgages and MBS that they had been holding in portfolio and sold them.  They could quickly substitute higher yield nonprime mortgages and CDOs for the lower yield paper that they ran off.  The net effect would show only modest growth but a significant increase in yield.
Wallison and Calomiris quote an article paraphrasing James Lockhart, Fannie and Freddie’s senior regulator as testifying that his agency recognized that Fannie and Freddie were greatly increasing their credit risk, but “the companies increased their exposure to risks in 2006 and 2007 despite the regulator’s warnings.”  Note that the regulator was, regardless of HUD guidelines, discouraging Fannie and Freddie from making additional nonprime loans.  James Lockhart was President Bush’s friend from childhood (they met in an elite prep school).  He had the President’s confidence and support.  Because Fannie and Freddie had just been caught in acts of repeated, severe fraud he had exceptional regulatory leverage over Fannie and Freddie.  He had ample regulatory authority to order Fannie and Freddie to cease increasing their credit risk and to reduce it.  Lockhart declined to use that authority.  He also declined to bring the fraud allegations to hearing against Franklin Raines (Lockhart settled for such small sums that Raines walked away wealthy).  Lockhart shared the fundamental anti-regulatory philosophy of President Bush – that’s why President Bush appointed him.  Still, as weak as the agency was made by the anti-regulatory dogma, it was superior in at least warning about Fannie and Freddie’s credit risk at a time when Wallison and Greenspan were focused entirely on fighting the last war – interest rate risk.
Why, if honest, would Fannie and Freddie’s CEOs have them function in a manner that would maximize short-term (fictional) reported income (and compensation) but cause catastrophic losses in the longer-term?  As Pinto and Wallison emphasize, Fannie and Freddie had plentiful experience demonstrating that liar’s loans were suicidal.  Further, why did they rely so heavily on liar’s loans – and cover up three-quarters of their non-prime loans through deceptive accounting – if the purpose was to meet self-imposed HUD requirements?  Wallison and Calomiris explicitly charge that Fannie and Freddie’s senior managers followed a deliberate strategy of accounting deception in order to dramatically understate how many nonprime loans they were making.  If they believe that Fannie and Freddie’s controlling officers engaged in accounting deception, why can they not even conceive that those managers would engage in a form of accounting deception that guaranteed that it would make them exceptionally wealthy within a year or two?  Wallison and Calomiris understand that the strategy of buying large amounts of nonprime loan created substantial (fictional) reported income from 2004-2008.
From the perspective of their 2008 collapse, this may seem to have been unwise, but in the context of the time, it was a shrewd decision. It provided the GSEs with the potential for continuing their growth and delivered enormous short-term profits. Those profits were transferred to stockholders in huge dividend payments over the past three years (Fannie and Freddie paid a combined $4.1 billion in dividends last year alone) and to managers in lucrative salaries and bonuses.
But perhaps I am too kind.  It may be that Wallison and Calomiris are so unaware of accounting control fraud that they do not understand that the “enormous short-term profits” were fictional.  They were the product of accounting fraud – a “sure thing.”  Had Fannie and Freddie established appropriate allowances for losses on loans this toxic they would have reported losses at the time they purchased the loans.  Wallison and Calomiris have described a classic accounting control fraud.  They have given a classic example of why Akerlof and Romer entitled their 1993 article “Looting: the Economic Underworld of Bankruptcy for Profit.” 
Although Fannie and Freddie were building huge exposures to subprime mortgages from 2005 to 2007, they adopted accounting practices that made it difficult to detect the size of those exposures.
Similarly, the New York Times reported.
 
Charles W. Calomiris, a finance professor at Columbia, testified that nobody saw the crisis coming because the two mortgage giants “adopted accounting practices that masked their subprime and Alt-A lending,”
Wallison needs to man up and learn to call powerful business leaders frauds.  Wallison’s lengthy dissent does not use the words “liar’s” loan, crime or criminal.  He uses the word “fraud” once, and the way he uses it is revealing.
Predatory lending. The Commission’s report also blames predatory lending for the large build-up of subprime and other high risk mortgages in the financial system. This might be a plausible explanation if there were evidence that predatory lending was so widespread as to have produced the volume of high risk loans that were actually originated. In predatory lending, unscrupulous lenders take advantage of unwitting borrowers. This undoubtedly occurred, but it also appears that many people who received high risk loans were predatory borrowers, or engaged in mortgage fraud, because they took advantage of low mortgage underwriting standards to benefit from mortgages they knew they could not pay unless rising housing prices enabled them to sell or refinance. The Commission was never able to shed any light on the extent to which predatory lending occurred. Substantial portions of the Commission majority’s report describe abusive activities by some lenders and mortgage brokers, but without giving any indication of how many such loans were originated [p. 447].
Only borrowers warrant the “f” word.  I have written extensively as to why it was overwhelmingly lenders and their agents that put the lies in “liar’s” loans. 
The “CYA” efforts get intense after a crisis.  As a lagniappe, I leave you with this gem I came across researching this column.  At the same conference that Wallison conducted to showcase Pinto’s work, Jay Brinkman the Mortgage Bankers Association’s (MBA) Chief Economists presented (his slides are available on AEI’s site).  Brinkman captures the essence of the MBA.  It’s his last slide, with the title “Credit Failures of the GSEs.”
Individual lenders cannot drive credit decisions and credit pricing.  That was and is the role of the GSEs, but they failed….
When I sat on a bank credit committee we believed that making credit decisions was the core of lending and that the adequacy of the price relative to the risk was always a critical consideration as to whether we should approve the loan.  “They failed.”  It’ll look good as a business motto over the entrance to the MBA headquarters – the new one after they strategically defaulted on the old one (shortly after they opined that any homeowner that strategically defaulted was a moral degenerate). 

3 responses to “Wallison is far too Kind to Fannie and Freddie