By William K. Black
Bloomington, MN: August 8, 2015
I have often noted the existence of a primitive tribal taboo shared by virtually all economists against using the “f” word – “fraud.” I have found a new example that sums up many of the pathologies of economics and economists. It is an article entitled “Going for Broke: New Century Financial Corporation, 2004-2006.” Given that New Century was a classic accounting control fraud, the use of the long-discredited gambling metaphor (our “autopsies” of S&L failures refuted it in 1984) demonstrates the crippling power of the taboo. The three economists who authored the September 2010 article are Augustin Landier (Toulouse School of Economics) David Sraer (Princeton University) David Thesmar (HEC & CEPR) (collectively, “LST”).
The Office of the Comptroller of the Currency (OCC) has published a list of the “worst of the worst” – the ten worst lenders in the ten worst markets for nonprime mortgage foreclosures. The absolute worst lender on that list is New Century. That makes it the worst of the worst of the worst. It also makes LST the worst of the worst of the worst economists writing about the fraud epidemics that drove the financial crisis.
The LST article is remarkably and inexcusably awful. At the time it was written it was representative of the lack of quality of orthodox economic “scholarship” about elite financial frauds. That has changed. As Mian and Sufi explained in their recent study demonstrating pervasive fraud in liar’s loans:
We know from a large body of research that both non-agency securitized mortgages and low-doc mortgages were associated with a high incidence of fraud (e.g., Ben-David (2011); Jiang, Nelson, and Vytlacil (2014), Griffin and Maturana (2014); Piskorski, Seru, and Witkin (2015)).
The most recent studies by very conservative financial scholars that actually study elite financial frauds have confirmed what we have been explaining for decades. While LST did not have access to these two recent studies, one of the studies (Piskorski, Seru, and Witkinl) was done on the basis of a New Century data set that the authors shared with other scholars who requested access, including the Jiang study that LST cite. Piskorski, Seru, and Witkin and Mian and Sufi exemplify how economists can conduct a real “forensic” investigation of fraud. The press picked up both the Piskorski, Seru, and Witkin and the Mian and studies’ findings. Here is the key passage from a story in which Piskorski was interviewed.
The authors specifically studied two types of mortgage misrepresentations: mortgages taken for primary residences that were not, in fact, primary residences, and mortgages taken for one property, while the second mortgage on the same property was concealed.
They compared loan-level mortgage data with credit reports and used New Century, a now-bankrupt subprime lender, as a case study. By comparing the loans with the credit profiles of their borrowers, they found that the loans New Century sold did not accurately reflect what the bank knew in their credit files.
While the data cannot establish how blatant the fraud was or how it was motivated, other research and insider accounts show that fraud was endemic among mortgage lenders in the years leading up the housing crash. And it wasn’t a case of a single bad company — the authors studied major companies and found similar results for all originators across the board.
“We didn’t find out this is the only the problem of a few bad apples. That would be kind of good news,” Piskorski said. “This is a really pervasive problem, and every single institution in the data misrepresented mortgages.”
This wasn’t a few cases of fraud, either. According to the study, 27 percent of loans obtained by non-owner occupants misreported their true purpose, and 15 percent of loans with second liens incorrectly reported that such loans were not present.
LST got New Century wrong despite an extensive, publicly available record that refuted their claims. LST ignored that record, failed to conduct any forensic examination of New Century, failed to conduct an effective hypothesis testing, and failed to alert their readers to any of these facts. They produced a travesty that exposes many of the pathologies of theoclassical economics.
LST have no excuse because their citations include George Akerlof and Paul Romer’s classic 1993 article – “Looting: The Economic Underworld of Bankruptcy for Profit.” LST, however, simply assumed that New Century could not be a control fraud. Here is the sole basis they provide for this assumption (with its accompanying footnote). LST use the acronym “NC” to refer to New Century.
Before going in detail over NC’s strategy following the 2004 monetary shock, it is interesting to note that the managerial team at NC had significant ownership stakes in the company. In 2001, Robert Cole, Brad Morrice and Edward Gotschall owned, according to EXECUCOMP, namely 15% of the company. With a market capitalization of around $277 Millions, this represented a $42 Million stake for the founding team. In 2005, because of multiple equity offerings, the ownership stake of the three founders went down to 7%. However, thanks to a striking increase in market capitalization (from $277M to $2B), their dollar stake did actually go up from $42M to $147M. As a consequence, we remark that during the entire period our sample covers, the top executives of New Century had significant incentives to maximize shareholder value. This evidence brings strong support to the risk-shifting view we develop in the remaining of the paper relative to explanations that would be based on “looting”.20
20 Besides, we have checked the insider filings on the SEC’s EDGAR website. Between 2003 and the default of NC, the founder-managers have sold less stocks than they were granted, in particular through stock-option exercise.
“This evidence” (A) is not “evidence” and (B) provides zero logical support for their assumption that it would demonstrate that the NC was not an accounting control fraud. It is not “evidence” because the LST authors assumed their conclusion – they assumed that NC’s “market capitalization” was not the product of accounting fraud. The reality, which LST studiously ignored in order to mislead the reader and policy makers, was that NC’s reported “profits” that drove the “striking increase in market capitalization” were fictional. They were produced by the accounting control fraud “recipe” for a lender. LST know this to be true from a raft of publicly available information from multiple sources. LST, however, provide none of that information to their readers. The information destroys their assumption that bankers who own stock in a company will never loot the company (an assumption also disproved by the relevant literature and history). The LST authors fail to engage Akerlof and Romer, the criminologists, the financial regulators, the secondary market participants, investigative, and judicial findings that had falsified their claim about stock ownership decades earlier. Indeed, Akerlof and Romer, the National Commission on Financial Institution Reform, Recovery and Enforcement, and the criminologists confirmed that we (the financial regulators) were correct that dominant (often 100%) ownership of an S&L by its controlling officers was a factor that greatly increased the risk of looting. LST hide all these findings from the reader. There are many reasons why a fraudulent CEO faces restraints in stock sales or simply waits too long to sell.
The first question is whether NC followed the accounting control fraud “recipe” for a lender. The answer, indisputably, is “yes.” (I provide details below.) The recipe for a fraudulent lender (purchaser) of loans has four “ingredients.”
- Grow like crazy by
- Making (or buying) really crappy loans at a premium nominal yield while
- Employing extreme leverage and
- Providing only grossly inadequate allowances for loan and lease losses (ALLL)
In NC’s case, in addition to the (deliberately and pathetically inadequate) ALLL on loans it held in portfolio (allowances that its controlling officers reduced as fraud, defaults, and losses surged in its massively expanding liar’s loans) the most relevant loss allowances were for repurchases of loans that it (A) fraudulently originated and (B) fraudulently sold to the secondary market through fraudulent “reps and warranties.”
Lenders led by honest bankers, of course, would never follow this recipe because it produces the classic three “sure things” – (1) the bank is guaranteed to report record (albeit fictional) profits in the near term, (2) the controlling officers will promptly be made wealthy through modern executive compensation, and (3) the firm will suffer catastrophic losses.
There is a literature on how to distinguish between accounting control fraud, incompetence, and “gambling for resurrection.” The beginning of that literature is cited by Akerlof and Romer.
In contrast to popular accounts, economists’ work is typically weak on details because the incentives economists emphasize cannot explain much of the behavior that took place. The typical economic analysis is based on moral hazard, excessive risk-taking, and the absence of risk sensitivity in the premiums charged for deposit insurance. This strategy has many colorful descriptions: “heads I win, tails I break even”; “gambling on resurrection”; and “fourth-quarter football”; to name just a few. Using an analogy with options pricing, economists developed a nice theoretical analysis of such excessive risk-takings trategies.4 The problem with this explanation for events of the 1980s is that someone who is gambling that his thrift might actually make a profit would never operate the way many thrifts did, with total disregard for even the most basic principles of lending: maintaining reasonable documentation about loans, protecting against external fraud and abuse, verifying information on loan applications, even bothering to have borrowers fill out loan applications.5 Examinations of the operation of many such thrifts show that the owners acted as if future losses were somebody else’s problem. They were right (Akerlof & Romer 1993: 5).
4. See Merton (1978).
5. Black (1993b) forcefully makes this point.
The research paper I authored in 1993 for the National Commission on Financial Institution Reform, Recovery and Enforcement that Akerlof and Romer cited was not the genesis of this analysis. Our “autopsies” of every S&L failure in 1984 disclosed the fraud “recipe” and the fact that it was essential to gut underwriting in a manner that no honest lender would ever do in order to produce the first two “ingredients” of the recipe. We used that insight beginning in 1984 to identify the accounting control frauds while they were still reporting record (albeit fictional and fraudulent) profits (like NC’s reported “profits” that LST implicitly assume must be real). Indeed, to Charles Keating’s horror, we targeted the S&Ls reporting the highest profits as our examination and enforcement priorities, while economists praised the frauds on the basis of their record reported (albeit fictional and fraudulent) profits. Three decades, and tens of thousands of frauds later, economists shilling for the CEOs leading these frauds still accept reported income and capital as if it were obviously accurate.
In 1995, I supplemented that “point” about how one can distinguish accounting control frauds from (hypothetical) honest “gamblers for resurrection” even more “forcefully” in an article I co-authored with two top white-collar criminologists – Kitty Calavita and Henry Pontell. “The Savings and Loan Debacle of the 1980s: White-Collar Crime or Risky Business” Law and Policy Volume 17, Issue 1, pages 23–55, January 1995. LST ignore the relevant literature. As financial regulators we found, in all the S&L failures, not a single case of “gambling for resurrection” that accorded with the theoclassical economic model on which LST rely. This is not surprising, as Akerlof and Romer explained in the next paragraph of their article.
Some economists’ accounts acknowledge that something besides excessive risk-taking might have been taking place during the 1980s. Edward Kane’s comparison of the behavior at savings and loans (S&Ls) to a Ponzi scheme comes close to capturing some of the points that we emphasize. Nevertheless, many economists still seem not to under-stand that a combination of circumstances in the 1980s made it very easy to loot a financial institution with little risk of prosecution. Once this is clear, it becomes obvious that high-risk strategies that would pay off only in some states of the world were only for the timid. Why abuse the system to pursue a gamble that might pay off when you can exploit a sure thing with little risk of prosecution? (Akerlof & Romer 1993: 4-5).
In the current crisis, the “risk of prosecution” of banksters leading accounting control fraud has become nonexistent.
New Century Was a Classic Accounting Control Fraud
At the time the LST authors wrote the version of the paper I found on the web there was a wealth of information available about NC’s frauds and the fact that it followed the fraud “recipe.” Two of the sources were the bankruptcy examiners massive report on NC and securities litigation against NC.
Warning: I went really wonky on this article because I am so outraged at LSTs’ conduct in not revealing these facts to their readers. I want to give the reader an understanding of how detailed the forensic evidence of NC’s accounting control fraud is and how pervasive, large, and blatant NC’s controlling officers’ frauds were. My secondary purpose in providing such extensive, arcane quotations is to give my readers a real world sample of the kind of understanding of accounting fraud schemes that is essential to be an effective financial regulator, investigator, and prosecutor – or economist or white-collar criminologist. None of these roles is for the faint at heart or those unwilling to make a huge investment of time to understand how fraud schemes actually work.
The first ingredient in the fraud recipe is extreme growth.
Indeed, an investigation conducted into New Century’s business practices by the court-appointed Bankruptcy Examiner assigned to investigate the causes of New Century’s collapse concluded that “New Century had a brazen obsession with increasing loan originations, without due regard to the risks associated with that business strategy.”
That finding establishes the first ingredient and (if you know the industry) the second. The LST authors concede this point, referring to NC’s “amazing growth.” The SEC’s securities fraud complaint against NC’s three controlling officers was also available to the LST authors, but none of its relevant findings from the SEC investigation were disclosed to the reader. The SEC’s investigation confirmed the “phenomenal” growth of NC loans (and the effect on reported earnings).
From 2001 to 2005, New Century’s loan production increased more than nine-fold, and in 2005 it originated over $50 billion in mortgage loans in that year alone and, at its peak, had over 7,000 13 employees. New Century’s earnings per share grew commensurate with its phenomenal growth. The Company’s reported earnings per share grew from $0.13 in 1996 to $7.17 in 2005. In a May 4, 2006 news release, under New Century’s trademarked byline, “A New Shade of Blue ChipTM,” Morrice publicly reported that the Company’s wholesale business “ranked as the #1 non-prime wholesale lender and #4 wholesale lender in the overall mortgage market in 2005.”
The fact that NC’s controlling officers bragged about being the largest non-prime wholesale lender in the world also demonstrates the second ingredient. The SEC complaint explains that making poor credit quality loans was NC’s strategic plan.
According to its periodic filings, New Century focused on “lending to individuals whose borrowing needs are generally not fulfilled by traditional financing institutions because they do not satisfy the credit, documentation or other underwriting standards prescribed by conventional mortgage lenders and loan buyers.”
Even LST concede, albeit in one of history’s great understatements, that “Over the full period, there is some evidence that NC’s lending standards have worsened….” But they ignore the fact that to speak of NC’s “lending standards” is to discuss an oxymoron, for NC’s lending standards were not followed. They were simply a key element in the fraud scheme – the false representation that NC actually (1) had prudent underwriting standards and (2) actually followed them. This deception was essential to NC’s ability to sell tens of billions of dollars in fraudulently originated mortgages through fraudulent “reps and warranties” to secondary market purchasers. The fraudulent “reps and warranties” were that NC actually had, and followed, prudent underwriting standards. One of the publicly available securities law suits alleges the following about NC’s underwriting practices – and backs the allegations up with compelling evidence. (“CW” refers to “confidential witness” – whistleblowers from NC.)
New Century systematically disregarded its own underwriting guidelines and wholly disregarded a borrower’s ability to pay when originating loans….
- As noted in the New Century’s Bankruptcy Examiner’s investigation (the “Bankruptcy Examiner’s Report”), which included a review of a large number of documents and 110 interviews of 85 fact witnesses, numerous members of New Century’s board of directors and senior management stated that the predominant standard for loan quality was whether the loan could be sold in the secondary market to investors like Deutsche Bank, not—as stated in the ACE 2006-SL1 RMBS prospectus supplement—whether a borrower could meet the obligations under the terms of a loan. Indeed, according to the Bankruptcy Examiner’s Report, New Century’s Chief Credit Officer said that in 2004 New Century had “no standard for loan quality.”
- As reported by the Bankruptcy Examiner, reckless origination and underwriting of New Century loans was rampant. For example:
- Certain senior managers at New Century in 2004 were told by a New Century employee that when underwriting stated income loans, “we are unable to actually determine the borrowers’ ability to afford a loan.”
- In early 2006, one senior manager at New Century described the performance of a certain loan product as “horrendous.”
- In 2004, the number and severity of the exceptions to underwriting standards employed by New Century to originate greater volume was 67 described by one Senior Officer as the “number one issue” facing New Century.
- By 2004, New Century Senior Management became aware of spiking increases in Early Payment Default (“EPD”) rates—where a borrower fails to make even the first several payments on a loan—suggesting that the loan should never have been originated in the first place. In every month following March 2006, the EPD rate exceeded 10%, reaching to as high as 14.95% by year end.
- Up until 2005 New Century used a DOS-based underwriting system which, according to a New Century manager interviewed by the Bankruptcy Examiner, enabled employees to “finagle anything.”
- In December 2009, the SEC filed civil fraud charges against New Century’s former CEO, CFO and controller, alleging that despite New Century’s representations as a prudent lender, it “soon became evident that its lending practices, far from being ‘responsible,’ were the recipe for financial disaster.” The executives settled the SEC’s civil fraud charged in July 2010 for approximately $1.5 million.
- Numerous former New Century employees interviewed by TIAA’s counsel confirmed the conclusions of the New Century Bankruptcy Examiner and the SEC’s complaint against New Century’s executives explaining that loans were not originated according to New Century’s stated underwriting guidelines, but were instead originated without regard to a borrower’s ability to repay the loan. For example, according to CW 26, a former New Century fraud investigator and senior loan underwriter who examined numerous New Century mortgage loans from January 1999 until April 2007, New Century “started to abandon prudent underwriting guidelines” at the end of 2003 in order to “push more loans through.” According to 68 CW 26, New Century essentially “stopped underwriting.”12 CW 29, a former New Century Vice President and Regional Manager, employed by New Century from September 1996 until May 2007 explained that New Century made very low quality and extremely risky loans, even for a sub-prime lender, noting that: “If you had a heartbeat, we would give you a loan.”
- CW 30, another former New Century underwriter and risk manager employed at New Century from December 2001 until April 2007, explained that exceptions to underwriting guidelines were endemic and it was “more about quantity than quality,” with the attitude being “get the volume on; get the volume on.” Indeed, CW 30 reported that nine out of ten loans that CW 30 recommended denying were nevertheless approved by management.13
- Facts such as these led the New Century Bankruptcy Examiner to conclude that statements in New Century’s SEC filings declaring that “regardless of document type, New Century designed its underwriting standards and quality assurance standards to make sure that loan quality was consistent and met its guidelines” were “not supportable.” Rather, the Bankruptcy Examiner concluded that “New Century did not produce ‘high quality’ loans or have ‘high origination standards.’” Moreover, claims asserted against New Century for making false 12 CW 27, a former New Century Vice President, Corporate Finance, agreed that New Century began to lower credit standards beginning in 2003. At that time, New Century changed its practice with respect to stated income loans, which became to be known in the industry as “liar’s loans.” CW 28, a former New Century senior training development manager employed by New Century from March 2003 until March 2006 explained that underwriters often allowed borrowers to resubmit a rejected full-documentation loan (which had been rejected because the borrower’s income was too low) as a “stated loan” with a new and higher income, which was then approved. CW 28 stated that this practice was “taboo” in the mortgage industry but routinely occurred and was a “running joke” at New Century. 13 CW 31, a former New Century Vice President, Regional Manager, employed by New Century from October 1999 until March 2007, stated that starting in 2003 and 2004, roughly half of New Century’s loans contained exceptions. CW 32, a former New Century underwriter employed by New Century from May 2005 to March 2006 in Itasca, Illinois and, previously, from 2000 until 2003 in Cincinnati, Ohio, explained that he could not recall the last loan that he looked at that did not have an exception. 69 or misleading statements of material fact regarding New Century’s purported prudent underwriting guidelines have already been sustained under Section 10(b) and Rule 10b-5 of the Securities Exchange Act of 1934, and Section 11 of the Securities Act of 1933. See In re New Century, No. CV 07-00931 DDP (JTLx), ECF No. 333, at 34 (C.D. Cal. Dec. 3, 2008) (“This Court likewise agrees…that Plaintiffs’ Complaint alleges sufficient facts that the statements were material misrepresentations of New Century’s loan quality and underwriting practices.”).
- Statements in the Offering Materials representing that New Century ensured proper appraisals when issuing loans to borrowers were also false and misleading when made. In order to increase loan origination volume, New Century routinely hired biased appraisers and used inflated appraisals as a matter of course to issue loans to borrowers who would not otherwise qualify for the mortgage.
- As described in the Bankruptcy Examiner’s Report, in New Century’s wholesale division—which accounted for the vast majority (approximately 85%) of New Century’s loan originations—the regional managers who had lending authority could override the internal appraisers’ decisions. Moreover, the regional managers’ compensation was not tied to loan quality, but was rather based on the volume of loans originated, providing incentive to inflate appraisal values in order to increase origination of New Century loans.
- As revealed in a 2005 internal New Century audit disclosed in the Bankruptcy Examiner’s Report, 18 of 77 (or 23%) of the loans reviewed at one Sacramento fulfillment center had “exceptions with either the appraisal conducted or the review of the appraisal submitted with broker-provided loans or the review appraisal conducted by New Century’s Appraisal Department.” The results of that audit were not an anomaly. According to the Bankruptcy Examiner, the results of New Century’s own loan quality audits of underwriting procedures, account manager review/approval, appraisals and funding “were dismal.” As reported by the 70 Bankruptcy Examiner, of nine branches audited by New Century in 2005, none were rated satisfactory, seven were rated unsatisfactory and two were rated as needs improvement.
- Former New Century employees interviewed by TIAA’s counsel confirmed that loans were routinely originated using improperly inflated appraisals. For example, CW 33, a senior vice president enterprise program manager for New Century in Irvine, California from July 2005 to April 2006, stated that he could “guarantee” that large appraisers used by New Century gave New Century “the benefit of the doubt,” i.e., provided an overly optimistic appraisal, in order to maintain New Century’s business. As but one example, CW 33 stated that an appraiser might photograph only one side of the house but not the side that was run down and falling apart in order to justify the inflated valuation. According to CW 33, these appraisal companies did not “want to piss off New Century” because they were compensated on volume.
- New Century’s statements regarding its “improved underwriting controls and appraisal review process” have already been held by one court to be false or misleading statements of material fact.14 Indeed, the poor quality of New Century loans were cited in a action brought by the Massachusetts Attorney General, which led to Morgan Stanley’s payment of $102 million to resolve charges that the Morgan Stanley funded loans to target lower-income borrowers and lure them into loans that they predictably could not afford to pay. See Assurance of Discontinuance, In re Morgan Stanley & Co. Inc., Civ. A. No. 10-2538 (Mass. Super. Ct. June 24, 2010) (the “Morgan Stanley Settlement”). 14 See In re New Century, No. CV 07-00931, at 33-34 (C.D. Cal. Dec. 3, 2008), ECF No. 333 (“The pleadings adequately support a finding that these statements were false when made.”).
All of this was publicly available to the LST authors through a simple web search.
NC’s controlling officers knew that making massive amounts of liar’s loans would inherently produce far greater losses – and proceeded to dramatically increase the number of liar’s loans it made and combine them with features that were certain to increase the default rate (and loss upon default) but which would also produce non-cash “income” for the bank and delay those defaults. The two primary features that had these characteristics were making zero down payment loans and “exploding rate” adjustable rate mortgage (ARM) loans. By “qualifying” borrowers at the “teaser” “ARM” rate the bankers also sought to (falsely) make it appear that the borrowers had adequate income to repay the loan. As I have explained on many occasions, liar’s loans are ideal “ammunition” for accounting control fraud because they remove the paper trail that we used to produce many of the S&L fraud convictions. LST ignore each of these points.
No one serious disputes NC’s massive leverage. LST admit that the controlling officers developed a strategy designed to create a “sharp increase” in leverage.
The SEC complaint spends considerable time documenting the inadequacy of NC’s loss reserves and related accounting fraud, the resultant inflation of “profits” that were actually losses, and how this enriched the controlling officers that were looting NC. Again, all of this was publicly available to the LST authors, who ignored it. Why would the findings of SEC investigators about fraud and looting by NC’s controlling officers be relevant to an economic article that assumes without any meaningful analysis that NC’s controlling officers must not have engaged in fraud and looting? Plainly, the direct and overwhelming forensic evidence of numerous brazen frauds by NC’s officers that the SEC (and bankruptcy and civil plaintiffs’ investigations) documented was irrelevant because LST knew that those officers would never loot or engage in fraud because they owned stock in NC. LST were simply protecting their readers from the facts that proved that LST’s thesis was false.
38. The February 2006 Capital Markets Report, which Morrice, Dodge and Kenneally received on March 17,2006, contained a special report on loan losses and borrower age. The report confirmed that losses were greater for stated income loans than for full documentation loans and for second-lien loans than for first-lien loans. The report concluded by stating: “As New Century continues to fund a greater percentage of 80/20 stated documentation purchase [loans] for younger borrowers … you can expect losses to increase as well.”
39. After first payment defaults jumped from a low of 0.6% in March 2005 to a then historical high of 2% in April 2006, New Century conducted a study to determine the cause of the increase. The “First Payment Default Report,” which Morrice, Dodge and Kenneally received on July 26, 2006, noted that the increase in FPDs occurred after New Century had changed the “credit mix” of its loans, including higher overall loan-to-value ratio loans (i.e., more 80/20 loans) and an increase in stated income loans. The report further stated that this increase in FPDs had occurred even though the average FICO score of New Century’s borrowers had increased. The report concluded that in making loans, the most important factor was the “credit mix” of each loan and that the borrower’s FICO score was less important.
42. New Century, Morrice, and Dodge made numerous representations disclosures regarding New Century’s subprime business model and the characteristics of its loan production through its second quarter and third quarter Forms 10-Q, securities offerings (which incorporated by reference the periodic filings or referred to them as true and accurate), press releases, and earnings calls, but they failed to disclose known negative information that significantly altered the total mix of information available to investors regarding the Company and had, and could reasonably be expected to have, an unfavorable impact on net revenues and income from continuing operations.
44. Second, as discussed in paragraphs 31and 32 above, as part of its loan production presentation, New Century disclosed materially misleading LTV information on its loans. Specifically, New Century disclosed a “weighted average” LTV, which, in 2006, was between 80.9% and 81.4%. As discussed above, however, the “weighted average combined” LTV tracked in the Company’s internal Capital Markets Reports ranged from 86.6 % to 87.6 %. New Century’s public disclosure of the lower “weighted average” LTV, instead of the higher internally reported “weighted average combined” LTV was materially misleading, as it gave the false impression that its borrowers had, on average, put 18.6% to 19.1 % down, when, in fact, its borrowers had put, on average, only 12.4% to 13.4% down and more than 30% of its loans were 80/20 or 100% LTV loans with no down payment whatsoever.
51. New Century could be required to repurchase loans sold pursuant to repurchase agreements in two situations: (1) the representations and warranties about the loan were untrue (e.g., the represented value of the underlying property was overstated); or (2) the borrower defaulted on the loan by failing to make the first payment due after the loan was sold. New Century’s financial results were negatively affected by loan repurchases and the amount of its allowance for loan repurchase losses. In repurchasing a loan, New Century had to repay the loan purchaser the loan’s full unpaid principal balance, any missed interest payments, and any premium paid for the loan. Moreover, once it had repurchased a loan, New Century was left with a loan whose value was typically 80% of the repurchase price.
2. New Century’s Increasing Loan Repurchases In 2006
a) New Century’s Increasing FPDs And EPDs
52. In 2006, as shown in the chart below, New Century experienced an increasing rate of FPDs and EPDs, which could trigger New Century’s obligation to repurchase loans it previously had sold. These increasing FPDs and, EPDs were chronicled in the monthly Capital Markets Reports, which Morrice and Dodge received. [Chart omitted.]
53. As a result of the increasing rate of FPDs and EPDs, New Century’s loan repurchases in 2006 sharply increased as compared to the prior years, as shown in the chart below. [Chart omitted.]
54. In fact, New Century repurchased loans totaling $315.7 million during the first six months of 2006, which represented 95% of all loans repurchased in 2005. Dodge knew of these repurchases because she received a month!y “CFO Report” that included the amount of actual repurchases for the month.
55. New Century’s increasing repurchase activity negatively affected its net income and liquidity. On August 17, 2006, Dodge advised Morrice in an email that New Century started the “we started the quarter with $400mm in liquidity and we are down to less than $50mm today.” The attachment to the email attributed the decrease in liquidity to a variety of factors, including loan repurchases, and recommended that New Century raise money from securities offerings.
C) New Century’s Increasing Backlog Of Repurchase Requests
56. In addition to its actual repurchases, New Century had a substantial and rapidly growing backlog of pending repurchase claims – from at least $143 million at the end of2005, to $400 million at the end of the third quarter of 2006, to $545 million by the time the third quarter 2006 Form 10-Q was issued.
57. In the second quarter and early part of the third quarter of 2006, New Century studied its repurchase claims to better quantify them and to attempt to improve its internal controls for processing and reporting such claims. In the third quarter and early part of the fourth quarter of 2006, New Century produced and widely disseminated several internal reports that discussed the increase in outstanding repurchase claims.
58. First, according to the August and September 2006 “Repurchase Activity” reports, as of July 31, 2006, New Century had $154 million in pending repurchase claims. Morrice, Dodge and Kenneally received the August report on August 26, 2007 report; Morrice and Dodge received the September report on September 7, 2006.
59. Second, on September 7, 2006, Morrice and Dodge received another report, entitled “Outstanding Repurchase Summary Report,” showing total outstanding claims of$281.9 million as of that date. The email accompanying the report stated that New Century “clearly got [its] teeth kicked in with regard to purchase requests in Aug[ust] and thus far in September.” Third, a report entitled “Inventory Management” reported that as of September 8, 2006, repurchase claims were trending up and pending repurchase claims totaled $382 million. Dodge received drafts of this report on September 27 and 28, 2006; both Morrice and Dodge received final versions of this report on at least three different occasions, on September 28, October 5, and October 12, 2006.
61. Finally, beginning in mid-October 2006, New Century began. Internally distributing a weekly report that was originally titled “Storm Watch” (later renamed “Key Indicators”) that was designed to summarize its key operating metrics, including repurchase claims. The “Storm Watch” reports, which were distributed to Morrice, Dodge, Kenneally and others, chronicled the outstanding repurchase claims, which grew from $143 million at the end of2005, to $400 million at the end of third quarter of 2006, to $545 million as of October 26, 2006 (before the third quarter 2006 Form 10-Q was issued). D) New Century’s False And Misleading Disclosures Regarding Loan Repurchases
62. New Century made substantial disclosures regarding New Century’s loan repurchases, but failed to disclose known increases in its loan repurchase obligations that would have significantly altered the total mix of information available to investors regarding the Company and had, and could reasonably be expected to have, an unfavorable impact on net revenues and income from continuing operations.
63. New Century disclosed in its 2006 Forms 10-Q that it could be required to repurchase loans sold pursuant to repurchase agreements in two situations: (1) where its representations and warranties about the loan were untrue; or (2) where there was an FPD or EPD. New Century also disclosed that it typically could sell or finance repurchased loans only at a significant discount to the unpaid principal balance, and significant repurchase activity could ~arm its cash flow, results of operations, financial condition, and business prospects. New Century further disclosed in its Q3 2006 Form 10-Q that it had $150.9 million in repurchases for the quat1er and $469.3 million in repurchases year to date, that repurchases had increased as a result of higher EPDs, that it expected the trend in increased repurchases to continue in the near term, but that it was refining its underwriting standards to mitigate against the trend.
64. These disclosures were misleading because New Century omitted known material information regarding its loan repurchases, including the substantially increasing early default rates (which, as reflected in the above chart at paragraph 52, greatly exceeded New Century’s reported FPD rates) and growing backlog of repurchase claims ($281.9 million at September 7, 2006; 400 million at September 30, 2006; and $545 million at the time the third quarter 2006 Form 10-Q was filed). Despite having repeatedly received information regarding these matters and substantially participating in preparing the Forms 10 Q, Morrice and Dodge failed to take any action to provide for proper disclosure of this negative information in New Century’s second and third quarter 2006 Forms 10-Q.
65. New Century also disclosed that provisions for estimated repurchase losses were charged to (or reduced) gain on sale of loans and credited to (or increased) the repurchase reserve and that actual repurchase losses (or charge-offs) reduced the repurchase reserve, i.e., charge-offs reflected actual repurchase losses and, by extrapolation, its actual repurchase activity. In the Critical Accounting Policies portion of its Forms 10-Q New Century presented the activity in its repurchase reserve, which showed that Company had only a small amount of charge-offs for loan repurchases:
B. New Century Improperly Ceased Accounting For Loss Severity
84. New Century’s loan repurchase rate and inventory of repurchased loans began to substantially increase in 2006, which required a greater reserve for inventory severity in accordance with SFAS 140, [Paragraphs] 11 & 55. Rather than increase the repurchase reserve and recognize a larger expense to fund the reserve, New Century changed its accounting methodology, first by eliminating inventory severity on repurchased loans in the second quarter of 2006 and then by eliminating future loss severity on current loan sales estimated to be repurchased in the future in the third quarter of 2006. These two changes that New Century implemented in the face of rising repurchases violated SFAS 140, [Paragraphs] 11 & 55 and had the effect of overstating the value of its repurchased loans; understating its repurchase reserve expense; and overstating its financial results.
85. By eliminating inventory severity, New Century understated its repurchase reserve expense by $81.871 million in the second quarter of 2006 and $3.917 million in the third quarter of2006. By eliminating future loss severity, New Century further understated its repurchase reserve expense by $42.095 million in the third quarter of 2006.
86. Kenneally, who was responsible for making the methodology changes, knew, or was reckless in not knowing, that neither of these changes were in accordance with GAAP and materially affected the Company’s financial statements. Under SFAS 65, 4, New Century was required to record loans held for sale at lower of cost or market (“LOCOM”) and to take as a current expense the amount by which the cost of loans held for sale exceeded their market value. Under SFAS 65,9 New Century was to determine LOCOM by the type of loan (residential or commercial), either on an aggregate or individual loan basis for each type of loan. Because New Century held, more performing loans that sold at a premium than non-performing loans (including repurchased loans) that sold at a discount, New Century’s aggregate unrealized gains on the performing loans offset the aggregate unrealized losses on non-performing loans. As a result, New Century never recorded a LOCOM expense.36
87. Before the filing of both the second and third quarter 2006 Forms 10 Q, Dodge learned that New Century had changed its methodology for calculating the repurchase reserve. In late July 2006, Kenneally told Dodge that he had changed New Century’s methodology for calculating the repurchase reserve. Kenneally also told Dodge the methodology change was to eliminate inventory severity and resulted in a $23 million decrease in the repurchase reserve expense. Prior to New Century filing its third quarter 2006 Form 10-Q, Kenneally reminded Dodge that he had previously advised her that New Century had eliminated inventory severity from the repurchase reserve calculation in the second quarter of 2006.
88. In order to comply with GAAP, SFAS 154, New Century’s was required to disclose its changes to its measurement of its repurchase reserves and repurchased loans, including a description of both the nature of and reason for the changes, an explanation of why the newly adopted accounting principle was preferable, and the method of applying the change, including the effect of the change on net income. In addition, Regulation S-X also required that interim financial statements shall include disclosures of “significant changes since the end of the most recently completed fiscal year in such items as: accounting principles and practices; estimates inherent in the preparation of financial statements….” 17 C.F.R. § 210.10-01(a)(5). Both Dodge and Kenneally knew, or were reckless in not knowing of these accounting changes. Despite their knowledge of these accounting changes, Dodge and Kenneally failed to take any action to ensure that New Century made these required disclosures. Indeed, in a disclosure checklist provided to other New Century officers, Kenneally indicated that there had been no accounting changes in the second quarter of 2006 by writing “N/A” for not applicable next to a question about accounting changes. Moreover, despite the Board of Directors’ specific questions about the adequacy of the repurchase reserve, Dodge and Kenneally never advised them of these accounting changes.
89. As a result of Dodge’s and Kenneally’s failure to disclose these significant accounting changes, or the material impact those changes had on the Company’s reported financial results, New Century’s second and third quarter Forms 10-Q materially overstated the Company’s financial results.
C. New Century Improperly Failed To Account For The Backlog Of Repurchase Claims
90. New Century failed, as required by GAAP, to estimate the loss it would incur from the backlog of repurchase claims, as required by SFAS 5, 8. Both Dodge and Kenneally knew, or were reckless in not knowing, ofNew Century’s backlog of repurchase claims. Indeed, Dodge and Kenneally were repeatedly advised of the unprecedented backlog of unprocessed repurchase claims prior to New Century filing its third quarter 2006 Form 10-Q, but they failed to account for the contingent losses it represented. By failing to account for the backlog of repurchase claims, New Century materially understated its repurchase reserve expense by $62.481 million in the third quarter of2006. D. New Century Improperly Excluded Interest Recapture
91. In estimating the repurchase reserve, New Century also violated SFAS 140, 11 by not providing for interest recapture prior to the third quarter of 2006. Dodge and Kenneally knew, or were reckless in not knowing, that New Century failed to provide for interest recapture prior to the third quarter of 2006. By failing to account for interest recapture, New Century understated its repurchase reserve expense by $2.122 million in the second quarter of2006.
E. Defendants’ Circumvention and Failure to Implement Internal Controls
92. Morrice, Dodge, and Kenneally knowingly failed to implement appropriate internal controls to track repurchase requests and their disposition. Each of them knew that New Century had no standardized procedure f9r receiving and processing repurchase requests during the second and third quarters of 2006. 1 Dodge and Kenneally also failed to implement any system of internal ac£ounting 2 controls relating to changes in accounting principles, and the disclosure thereof.3
F. New Century’s Material Overstatement Of Its Financial Results 4
93. As a result of its improper repurchase reserve accounting described above, New Century materially overstated its financial results, as shown on the 6 table below. As also shown on the table below, with correct accounting, New 7. Century would have reported financial results far below analysts’ estimates. When 8 New Century announced these overstated financial results in its earnings releases, 9 its stock trading volume (but not price) rose significantly.
Collectively, these frauds not only created massive fake profits, but very real and massive compensation to NC’s controlling officers. The SEC complaint explains:
X. DEFENDANTS BENEFITED FROM THE FRAUD
100. During 2006 and 2007, the Defendants received the following salaries, cash bonuses, and incentive compensation based on employment contracts and pursuant to a 2004 performance incentive plan:
The chart showing the compensation will not reproduce well, but the total compensation for 2006-2007 for the three senior officers that the SEC sued was:
The LST Authors’ “Gambling for Resurrection” Assertions
LST know about the NC bankruptcy examiners’ report, for they cite it in their fourth footnote – but for an utterly trivial point while ignoring all of its devastating content. LST present zero direct evidence that the NC controlling officers gambled for resurrection. Their approach, instead, is a model that inherently cannot evaluate whether the controlling officers were looting, which makes the model useless for its purported purpose. Indeed, the authors never formally state any hypothesis test and never attempt to test a hypothesis of control fraud (“looting”) because they know their model is incapable of testing that hypothesis.
LST started out wrong, guaranteeing that they would end wrong.
[A]n institution in distress is biased toward projects that pay-off in the state of the world where it escapes bankruptcy (the seminal model of risk-shifting is Stiglitz and Weiss, 1983). Such biased project choice can lead to the selection of negative NPV projects, while remaining compatible with ex-post shareholder value maximization.
Institutions” do not make these decisions. The person who controls the firm makes the decisions. As we found as regulators, investigators, prosecutors, and criminologists and as Akerlof and Romer agreed, the CEO’s optimal strategy (if he is willing to violate the law) is to pick the “sure thing” that makes him wealthy and keeps his job alive several years longer. That is particularly true when regulation and prosecution are weak – and they have never been weaker in the modern era. The idea that NC’s best high risk honest bet was making massive and sharply increasing amounts of liar’s loans is preposterous. Such loans were sure to crash as soon as the bubble stalled – and the risk premium was trivial relative to the massive increase in default risk inherent in failing to underwrite loans prudently (a point documented in a study of liar’s loans that LST cites. (LST cite: Jiang, Nelson and Vytlacil, 2010, “Liar’s loans? Effects of origination channel and information falsification on mortgage deliquency”, Working paper. The published version of the paper can be found at this link.) An honest gambler would under the LST’s authors’ own theories never have rationally chosen this strategy rather than a high risk/high yield – non-negative net present value (NPV) investment in a financial derivative. The LST authors admit that NC’s controlling officers greatly increased liar’s loans: “There is a sharp drop in the fraction of full documentation loans in late 2005.” This is consistent with looting and refutes the honest gambler thesis.
Recall Akerlof and Romer (paraphrasing me) responding to this exact line of argumentation:
[E]conomists developed a nice theoretical analysis of such excessive risk-taking strategies.4 The problem with this explanation for events of the 1980s is that someone who is gambling that his thrift might actually make a profit would never operate the way many thrifts did, with total disregard for even the most basic principles of lending: maintaining reasonable documentation about loans, protecting against external fraud and abuse, verifying information on loan applications, even bothering to have borrowers fill out loan applications.
NC’s controlling officers – just like their S&L counterparts – made loans that no honest gambler would make because they required the “total disregard for even the most basic principles of lending” which made huge losses inevitable. Akerlof and Romer specifically referenced what we now call “liar’s” loans – which is precisely the type of loan that NC’s controlling officers increased massively even after they knew that such loans were endemically fraudulent and were causing a surge in defaults even as its “amazing growth” in lending hyper-inflated the bubble.
We know why accounting control frauds made liar’s loans their “ammunition of choice.” Now ask the question that the LST authors never asked – would an accounting control fraud find it useful in 2003-2007 to add additional risks such as zero down payment, qualifying borrowers at a teaser ARM rate, and making exploding rate ARMs? The answer is an emphatic “yes” for the reasons I explained above. This means that the LST authors’ model is incapable of evaluating whether NC’s controlling officers were running a control fraud.
The LST authors miss the fact that the theory they purport to believe (gambling for resurrection) refutes their claim that NC’s controlling officers would never loot NC because they owned stock in the company. Recall that their claim was that NC had a high reported net worth, so NC’s officers would never loot it. Now they advance a model that requires that rather than having ample capital NC was in “distress.” That phrase, as the standard models uses it, means “massively insolvent.” The LST authors state that in such circumstances the CEO will cause the firm to select “negative NPV projects.” Under the theory LST purport to believe, that would be true only if the firm were (A) deeply insolvent and (B) there were no projects in which the firm could invest with non-negative NPV that – were they to succeed – would provide a large enough profit to restore the firm to solvency. Condition “B” would be exceptionally rare and true only of a firm with some combination of massive insolvency and minimal liquidity with which to fund the ultra-risky investment. I know of no real world examples that satisfy condition “B” for a corporation.
But LST suggest a change in the traditional gambling for resurrection theory that was falsified in the S&L debacle. They purport that corporations suffer from far more perverse incentives than the traditional “gambling” theory posits.
Note that such distortion in project selection does not rely on a very high probability of default (we used 3% in this example). Risk-shifting is a major friction even for companies which are far from being insolvent.
LST do not understand banking, risk, return, or the implications of their proposed revision to theory. Their theory assumes implicitly that NC’s controlling officers were simultaneously, deliberately seeking to dramatically increase NC’s beta and dramatically decrease its alpha. That makes no sense. “Beta” refers to systematic risk, while “alpha” is supposed to reflect the return earned (or lost) due to skill (or incompetence or fraud) in choosing particular investments.
In banking, alpha is maximized through superb underwriting and alpha becomes sharply – fatally – negative when underwriting is not done properly. By definition, real estate liar’s loans represent pathetic underwriting. The same is true of appraisal fraud, which was also endemic at NC. No honest lending officers would ever engage in or permit material appraisal fraud. But LST do not even discuss NC’s pervasive fraudulent origination of loans through liar’s loans and appraisal fraud or the fact that there is no fraud exorcist. If you originate a loan fraudulently you can only sell it to the secondary market through fraudulent reps and warranties. NC’s controlling officers’ endemic loan origination frauds inherently beget endemic fraudulent sales and inherently beget endemic accounting and securities frauds because if NC established proper loss reserves for its fraudulently originated and fraudulently sold loans it would have reported the truth – that it was insolvent and massively unprofitable. LST demonstrate no understanding of any of these facts.
The markets do not reward “risk” – they reward certain kinds of risks. If a bank does not underwrite properly – if it extorts appraisers to inflate appraised values and makes liar’s loans – its risk explodes – but the markets do not reward taking that gratuitous risk with a higher expected return. Similarly, no one rewards you if you walk into the street without looking or even listening for cars first. The negative alpha in making liar’s loans and inflating appraisals is enormous. No honest gambler would make liar’s loans or allow (much less extort) inflated appraisals. LST’s failure to see that demonstrates that they do not understand risk, return, or banking.
LST assume (and then propose a theory that rejects) efficient markets and capital assets pricing model (CAPM), Gaussian distributions – all the old discredited model failures. They appear to have learned nothing from the crisis. But the implications of their theory would be enormous. Virtually every corporation in the world has a probability of default well in excess of “3%.” If making investments with a negative present value (NPV) is the norm for corporations then corporations represent (as Adam Smith warned) a clear and present danger to the world economy and radical reform is urgently required. Among the lesser implications of their theory is that markets are nearly always inefficient (which makes their reliance on the efficient market hypothesis a tad suspect).
But does any of this have anything to do with reality? An alternative theory explains the same result – control fraud (“looting”). That theory explains why those that control banks cause them to make massive amounts of fraudulent liar’s loans, to extort appraisers to inflate appraisals, to increase those fraudulent loan origination practices even as defaults surge, to provide grotesquely inadequate loss reserves through accounting and securities fraud, and to compound that loan origination fraud by selling the loans through fraudulent reps and warranties. Each of these fraudulent actions is documented by the bankruptcy examiner, the SEC, and large numbers of whistleblowers – and by NC’s actual default and loss experience. LST ignore the implications of each of these points.
Instead, LST propose a gambling theory in which whatever loans a bank CEO has made previously cause him to double-down on the beta risk of that portfolio. Specifically, they claim that bankers who make fixed rate loans subject to interest rate risk will make new loans that compound that interest rate risk. While they are not explicit on this point, their logic also requires bankers to further increase the resultant interest rate risk by increasing leverage. This dynamic should cause risk and leverage to increase substantially over time and cause recurrent financial crises and greatly increased failure rates (unless bailed out as too big to fail). Indeed, it implies that becoming so large that the firm becomes a systemically dangerous institution (SDI) should become the dominant strategy.
Very similar claims were made by economists about the S&Ls during the debacle – that they doubled-down on interest rate risk by growing massively and increasing their fixed rate mortgages. Those claims proved to be false. Traditional S&Ls grew moderately and made increasing numbers of mortgages with lower interest rate risk. Every one of those roughly 3,000 S&Ls was insolvent, on average by about 15 percent, yet none of them acted in accordance with the “gambling for resurrection” model. This was not because they did not gamble. Their gamble was that interest rates would fall and the market value of their fixed rate mortgages would rise. But they did not double-down in accordance with the predictions of LST’s proposed “gambling” model. The roughly 300 S&Ls that grew massively also did not double down on making fixed rate mortgages. Instead, they followed the fraud recipe and made loans that optimized their frauds.
LST ignore the fact that the NC was massively insolvent on a market value basis – and its controlling officers knew it. The LST theory further requires that NC’s creditors and shareholders not know that NC was massively insolvent. This scenario should have led the LST authors to an immediate problem with their “honest gambler” theory for NC was a publicly traded corporation and its controlling officers had a duty to disclose these facts. In other words, the LST theory is that the officers could not have engaged in control fraud – but must have engaged in securities fraud. Indeed, the LST theory implies that NC’s CEO would be acting on shareholders’ behalf by committing securities fraud because honest disclosures would have led to the immediate closure of the deeply insolvent firm as early as 2004 – and such a closure would wipe out the shareholders (and NC’s controlling officers were its largest shareholders) because of the shareholders’ exceptionally low priority in bankruptcy. The CEO’s duty, therefore, to the shareholders under LST’s logic was to commit massive securities fraud to delay NC’s bankruptcy and then use that time to engage in an ultra-risky investment that was so likely to fail to spectacularly that it had a negative expected value. Of course, those losses on that ultra-high risk gamble would be borne entirely by the creditors because of the limited liability of shareholders.
The strategy that LST considers optimal constitutes “waste” at law and some courts consider it to be a breach of fiduciary duties by the CEO of an insolvent firm (a position that the Office of Thrift Supervision embraced) to engage in waste. LST are not strong in law, criminology, regulation, fraud schemes, accounting, corporate governance, ethics, or economics.
But LST are particularly weak in logic. NC’s controlling officers likely knew no later than 2004 that their firm was doomed, but no one knows when the bubble will cease to inflate rapidly and the fraud scheme will collapse. Under LST’s own logic, looting would produce by far the highest expected payoff to the officers in such a situation – regardless of their stock holdings. I emphasize that LST’s own logic requires that NC’s controlling officers engage in securities and accounting fraud to hide its massive, growing insolvency – so LST obviously do not think that NC’s controlling officers were effectively circumscribed by morality.
LST provide a revealing, false tale about liar’s loans.
Because borrowers and mortgage brokers have strong incentives to lie about socio-demographics (in particular on income), we focus here on fully documented loans.
How about bank CEOs’ “strong incentives” to ensure mass lies about the borrower’s income? Who created and made possible those “strong incentives to lie?” Who expanded and maintained those “strong incentives” to lie even as the internal and external evidence and warnings about the endemically fraudulent nature of liar’s loans proliferated. Indeed, how much warning do you need as a banker to know that something you call a “liar’s loan” is a terrible idea? LST are exceptionally incurious about these points even though economists (falsely) pride themselves as the masters of analyzing “incentives.”
What LST have conveniently done is excluding liar’s loans from their analysis. Liar’s loans and appraisal fraud provide the “natural experiments” that analysts can use to determine whether the controlling officers are looting or “gambling for resurrection.” By refusing to “focus” on the obvious evidence of endemic fraudulent loan origination (which inevitably flowed through to fraudulent sales to the secondary market) and by failing to inquire why NC’s controlling officers crafted and maintained the perverse “strong incentives” that they knew were producing endemic fraud LST give up the best analytical tools to evaluate whether their gambling hypothesis is correct. LST were delighted to give up their best analytical tools because those tools refuted falsified their thesis and demonstrated looting (control fraud).
No one forced NC’s controlling officers to make liar’s loans. NC was not subject to the Community Reinvestment Act. (The CRA has no requirement to make liar’s loans.) NC’s owners chose to make liar’s loans even though we had established by 1991 that they were an optimal “ammunition” for accounting control fraud and inevitably produced severe losses– which is why we as regulators began to successfully drive them out of the S&L industry in that year. NC’s controlling officers knew these facts, and new that the liar’s loans they were making were blowing up and would cause catastrophic losses as soon as the bubble’s hyper-inflation slowed. As it slowed, LST admit that NC’s controlling officers caused a “sharp drop” in non-liar’s loans. (In plainer English, NC’s controlling officers caused the mortgage bank to rapidly expand liar’s loans and make them NC’s priority.)
Similarly, no one forced NC’s controlling officers to use loan brokers. Akerlof and Romer warned in 1993 that loan brokers had long had a terrible reputation. NC’s controlling officers knew that using loan brokers increased the already extreme incidence of fraud in liar’s loans – and responded by substantially increasing liar’s loan origination, largely through brokers.
But NC had plentiful fraud origination in liar’s loans its own staff made. That is because NC loan officers – like loan brokers – had perverse financial incentives structured by NC’s controlling officers. NC’s controlling officers knew that these incentives were perverse and were leading (think “revealed preferences”) to endemic fraud. They refused to end the perverse incentives.
Even the Bush administration financial regulators – as weak a group as ever assembled in modern America – recurrently warned the industry against making liar’s loans. NC’s controlling officers did the opposite – they significantly increased liar’s loans from 2000-2003 and then massively increased liar’s loans in 2005 and continued that increase until the frauds inevitably caused its catastrophic failure (see Appendix figure 2 of the LST article). In addition, NC’s controlling officers added “layered” risks certain to vastly magnify defaults and losses (but also, in conjunction with their accounting and securities fraud, to delay loss recognition and prolong their fraud schemes). Indeed, NC’s fraudulent loan origination became so large that it contributed materially to hyper-inflating and extending the real estate bubble. It also acted as a fraud “vector” – spreading its fraudulently originated mortgages throughout the financial system through its fraudulent sales of fraudulent mortgages to the secondary market. All this disappears in LST’s account, in which NC and its controlling officers are presented as if they were the innocent victims of “borrowers and mortgage brokers” who “have strong incentives to lie about [the borrower’s] income.” LST thought it was not worth mentioning that it was NC’s controlling officers who shaped those perverse incentives and maintained or even exacerbated them as they were repeatedly informed of the obvious – that their perverse incentives were producing widespread fraud, mounting defaults, grotesquely inadequate loss reserves, and grossly inflated “profits” that were actually losses so severe that NC was insolvent. LST also decided it was not worth informing their readers that NC’s controlling officers hid these facts from the public and investors through widespread, multiple, and brazen accounting and securities frauds – and adopted new forms of accounting fraud that reduced the already inadequate loss reserves in response to the surging frauds, defaults, and losses.
As NC’s controlling officers were informed of the rapidly rising defaults and loan repurchases the SEC investigation found that NC engaged in a series of accounting frauds. Those accounting frauds caused an insolvent and money-losing mortgage bank to report that it was profitable and had substantial capital. That led to very large compensation to the controlling officers and kept the insolvent firm open for years. By massively expanding NC’s fraudulent liar’s loans its controlling officers simultaneously inflated reported income and helped hyper-inflate the bubble, which kept them in power and increased their compensation as a reward for destroying and looting NC and placing hundreds of thousands of borrowers in positions in which they would suffer foreclosure. NC’s controlling officers did this so egregiously that they made NC the worst of the worst of the worst.
But LST never ask “why?” Instead, they shill for NC’s controlling officers and attempt to shift all responsibility for NC’s endemic fraudulent loan origination and fraudulent sales to the secondary market to others. In particular, the effort to blame – with zero facts – the borrowers as the principal source of mortgage fraud is preposterous and ignoring who created the perverse incentives that led to endemic fraud by NC’s loan brokers.
Even though LST know that NC engaged in the massive origination and sale to the secondary market of fraudulent loans through fraudulent reps and warranties, they repeatedly treat the inflated incomes as if they were real. LST also make this unsupportable statement that indicates their unacknowledged biases.
13 Appendix figure 2 reports the fraction of “full documentation loans”. There is a sharp drop in the fraction of full documentation loans in late 2005. This suggests that socio-demographics (in particular income, but not the fico score, which is directly obtained by the loan officer or the broker) should be treated with a grain of salt after the end of 2005. It turns out that the results we present here are not materially affected by the inclusion, or not, of the partial-doc and no-doc samples.
Put aside the idiocy of claiming that endemically fraudulent liar’s loans only became an issue affecting the reliability of NC’s (pervasively fraudulent) accounts “in late 2005” and the further idiocy of describing loans that MARI (the mortgage banking industry’s own anti-fraud specialists) warned had (A) a 90% fraud incidence with (B) roughly 60% of the frauds inflating the borrower’s income by at least 50% as being a matter that requiring “a grain of salt.” Ignore for the moment (as LST do) the logical implications of NC’s controlling officers adopting a strategic plan of massively increased originations (and subsequent fraudulent sales) of what it knew to be loans that were endemically fraudulent. LST cannot know whether their analysis would be “materially affected” by considering the endemic fraud in NC’s liar’s loans because they do not even attempt to study the magnitude of that fraud, the resultant defaults, the losses, and the pathetic level of the loss reserves that produced the NC’s fictional income and capital when it was in reality losing money and insolvent on an economic basis.
But now return to the implications of NC’s controlling officers’ decision to respond to the copious internal warnings of surging fraud and defaults in their fraudulent liar’s loans by massively increasing their fraudulent origination of such loans, massively increasing the fraudulent sales of these fraudulently originated loans, engaging in massive and blatant accounting and securities fraud to hide the extent of the frauds and losses, the absurd inadequacy of NC’s loss reserves, the failure of NC’s controlling officers to recognize losses, the resultant inflation of reported income and net worth and the fact that each of these forms of accounting and securities fraud helped NC’s controlling officers loot the mortgage bank. The implications, of course, are fatal to LST’s hypothesis, so there is no mystery about why economists refused to employ “revealed preferences” to analyze why NC’s controlling officers took each of these actions.
LST repeatedly make unsupported and unsupportable assumptions. Focus on the words “obvious” and “sought to” in this passage.
At the same time, interest rates are increasing sharply from 9 to 11%, or about 200bp. The right panel in figure 3 reproduces the average interest rate (at loan issuance). In doing this, they track the monetary policy implemented in the spring of 2004 by the Federal Reserve (see figure 5). LIBOR, which is used as the reference interest rate by New Century, increases by 400bp between mid 04 and mid 06. Comparing the two, it is obvious that NC has sought to absorb, through lower margins, about half of the increase in interest rate created by the Fed.
LST’s claim is not “obvious” and the claim about what NC’s controlling officers “sought to” achieve is exceptionally unlikely to be true.
Similarly, LST’s description of what it takes to “confirm” one of their implicit assumptions is instructive.
To obtain a picture of interest-only origination for the whole industry, we went through the 10k filings in of several known subprime mortgage issuers. We calculate the fraction of interest-only loans.14
In total originations over the 2002-2006 period: quite clearly, NC’s move toward deferred amortization is representative of the rest of the industry, confirming that there is no specific pathology in the firm we are studying (see appendix figure 3).
No, this does not “confirm” “that there is no specific pathology” at NC. The bankruptcy report and the SEC and plaintiff investigations that identified a host of whistleblowers at NC confirm that NC’s controlling officers created a culture at NC that was deeply, broadly, brazenly, and specifically pathological. What LST should be saying is that NC’s pathology – accounting control fraud – was common among nonprime lenders and purchasers of nonprime loans. What we are seeing, again, is economists’ primitive taboo about fraud. LST implicitly assume (by ignoring the relevant literature and history) that control fraud must be rare and that if a practice is common in an industry it must not be criminal. There is no logical or empirical basis for that implicit assumption. As Piskorski explained: “This is a really pervasive problem, and every single institution in the data misrepresented mortgages.” That commonality of fraud demonstrates an exceptionally criminogenic environment and the presence of a “Gresham’s” dynamic. That environment and dynamic refute rather than support LST’s thesis.
LST make, unintentionally, clear their extreme bias in favor of their thesis.
We have here documented a very striking change in NC’s product mix and consumer base. Hereafter, we will interpret it as evidence of risk shifting by a distressed institution, and provide further evidence consistent with this. But before doing this, let us look at the exact predictions of a canonical model of financial distress. Understanding these predictions will guide our empirical strategy.
Yes, they did interpret it as supporting their thesis and at no point considered and provided a reasoned basis for rejecting alternative explanations “inconsistent” with their hypothesis. LST do not even attempt to explain why NC’s controlling officers continued to cause it to make sharply increasing amounts of fraudulently originated liar’s loans and then sell those fraudulent loans through fraudulent reps and warranties.
Every piece of “evidence” that LST advance that supposedly supports their hypothesis is consistent with and far better explained by NC’s controlling officers following the fraud recipe. But LST ignore this fact because they have already assumed the validity of their hypothesis.
LST also do not understand the word “forensic.”
This paper has provided forensic evidence on the risk-shifting behavior of a large mortgage originator. The sharp rise in interest rate in 2004 destroyed a large fraction of New Century’s net present value. In reaction, New Century drastically modified its business model. It introduced a new, more price sensitive product: the interest-only loan. It changed its customer base, selling this new product to more credit-worthy, wealthier households, whose repayment decisions are also more sensitive to real estate prices. Finally, it changed the geography of its operations – selling more and more of these new loans in cities with real estate prices correlated with its legacy assets. This new business strategy is consistent with that of a financially distressed company that starts taking long bets on its own survival.
LST provide no forensic evidence on anything. They present the standard economic model drivel with econometric work that cannot test the rival hypotheses. There is, however, a huge amount of “forensic” evidence on NC – and that evidence establishes that it was a classic accounting control fraud. LST know about this evidence but they don’t inform their readers and they do not even attempt to refute the forensic evidence. Further, as LST admit, the customers NC increasingly lent to were not “more credit-worthy” but instead less credit-worthy. Enormous, increasing numbers of these customers were not “wealthier” – the bankers simply created a system that would lead to the endemic and dramatic inflation of the borrowers’ incomes so that NC could more easily sell the loans to the secondary market through fraudulent reps and warranties. LST have done a disservice to everyone through a shockingly misleading article. It was bad enough as the Queen of England observed that conventional economists’ ideologies shaped a criminogenic environment that produced the financial crisis and the Great Recession. But it obscene that economists even after they helped produce that calamity are shilling for the controlling officers of a bank that was the worst of the worst of the worst frauds. It is embarrassing that they do so by making the facially, historically, and theoretically preposterous claim that officers who own stock in a corporation would never loot the corporation. And it is shameful that they did so by never informing the reader of the copious forensic evidence that New Century was a classic accounting control fraud.