FILED: NEW YORK COUNTY CLERK 01/29/2013 INDEX NO /2012 NYSCEF DOC. NO. 6 RECEIVED NYSCEF: 01/29/2013

Transcription

1 FILED: NEW YORK COUNTY CLERK 01/29/2013 INDEX NO /2012 NYSCEF DOC. NO. 6 RECEIVED NYSCEF: 01/29/2013 SUPREME COURT OF THE STATE OF NEW YORK COUNTY OF NEW YORK HOME EQUITY MORTGAGE TRUST SERIES , HOME EQUITY MORTGAGE TRUST SERIES , and HOME EQUITY MORTGAGE TRUST SERIES , Index No.: COMPLAINT -against- Plaintiffs, DLJ MORTGAGE CAPITAL, INC. and SELECT PORTFOLIO SERVICING, INC., Defendants. Plaintiffs Home Equity Mortgage Trust Series (the HEMT Trust ), Home Equity Mortgage Trust Series (the HEMT Trust ), and Home Equity Mortgage Trust Series (the HEMT Trust ) (collectively, the Trusts ), acting by U.S. Bank National Association, solely in its capacity as Trustee (hereinafter the Trustee ) of the Trusts, which is in turn acting at the direction of certain holders of Certificates issued by the Trusts (the Directing Certificateholders ), for their Complaint against Defendants DLJ Mortgage Capital Inc. ( DLJ ) and Select Portfolio Servicing, Inc. ( SPS ) (together, Defendants ) hereby allege as follows: NATURE OF THE ACTION 1. In this action, Plaintiffs seek redress from two affiliated entities, DLJ and SPS, for their breaches of contracts governing the mortgage-backed securities issued by each of the Trusts. These breaches have contributed to losses which now stand at more than $730 million. As set forth in detail herein, DLJ and SPS have not only failed to cure these breaches but have obstructed Plaintiffs efforts to ascertain their extent by restricting access to documents that would very likely 1

2 further substantiate and establish more breaches. This action is brought to require DLJ and SPS to honor their contractual obligations by repurchasing all breaching loans from the Trusts after affording the Plaintiffs reasonable access to key evidence. 2. Upon the issuance of the mortgage-backed securities, DLJ made a series of representations with respect to the thousands of mortgage loans it sold to the Trusts. It promised to buy the loans back if those representations proved false. Confronted with comprehensive, forensic evidence of hundreds of breaches of those representations, DLJ has thus far failed to repurchase a single mortgage loan among those identified in the Complaint. 3. This Complaint relates to three distinct securitization trusts. These were created in 2006, when DLJ purchased from mortgage originators, and then securitized, approximately 23,900 residential mortgage loans (9,729 for the HEMT Trust; 5,334 for the HEMT Trust; and 8,846 for the HEMT Trust) (collectively, the Mortgage Loans ). For each of the Trusts, securitization of these loans was accomplished through two contracts: an Assignment Agreement by which DLJ transferred the Mortgage Loans from itself to a special purpose securitization entity known as a Depositor, and the Pooling and Servicing Agreement (the PSA or PSAs, as applicable) (copies of which are attached hereto as Exhibits A, B and C), under which the Depositor in turn transferred the Mortgage Loans to the respective Trusts. Collateralized by the Mortgage Loans, the mortgage backed securities, also known as Certificates, were issued on February 1, 2006 (for the HEMT Trust), June 1, 2006 (for the HEMT Trust) and August 1, 2006 (for the HEMT Trust). 4. The credit quality of the Certificates is inextricably linked to the credit quality of the Mortgage Loans because cash flow generated by payments of principal and interest on the underlying Mortgage Loans is the source of revenue for payments due to the purchasers of the 2

3 Certificates. DLJ made representations and warranties about the fundamental credit-quality characteristics of the Mortgage Loans in the PSA for each of the Trusts. Among other things, DLJ represented that the Mortgage Loans complied with all the terms, conditions and requirements of the mortgage originators underwriting standards, and that specific information attached to the PSAs that described the characteristics of those loans was accurate and true. 5. To date, the Trusts have collectively suffered losses of more than $730 million, a figure that continues to escalate. Accordingly, in late 2011 and into 2012, a series of reviews of samples of the Mortgage Loans was undertaken. These reviews included a close examination of the Mortgage Loans origination files (the Origination Files ) related to 806 Mortgage Loans (the Re-underwriting Review ), many of which had defaulted or were delinquent. Of these 806 Mortgage Loans, 513 were from the HEMT Trust, 249 from the HEMT Trust, and 44 from the HEMT Trust. 6. The Re-underwriting Review established that, contrary to the facts DLJ represented to be true, many of the loans were defective in one or more of the following categories: (a) (b) (c) (d) Misstatement of Income and Employment. Many Origination Files misstated the borrowers income, misstated the borrower s employment, or listed incomes that were facially unreasonable for the borrowers stated occupations. Misstatement of Debts. Many Origination Files misstated the borrowers existing debt obligations, despite clear indications that the borrowers had more debts than were listed in the borrowers applications. Misstatement of Debt-to-Income Ratios. A borrower s debt-to-income ratio (or DTI ratio) compares the borrower s monthly debt obligations to the borrower s monthly income. The higher the DTI ratio, the greater the likelihood of a default. For many Mortgage Loans, the actual DTI ratios were far higher than the DTI ratios represented by DLJ. Misstatement of Owner-Occupancy Status. Borrowers who live in mortgaged properties are much less likely to default than borrowers who do not. If a property is not owner-occupied, then the risk of default increases. 3

4 Many Origination Files listed properties as being owner-occupied, despite clear indications that this was not the case. (e) (f) Misstatement of Loan-to-Value Ratios. A property s loan-to-value ratio or combined loan-to-value ratio ( LTV and CLTV ratios) expresses the amount of the Mortgage Loan as a percentage of the property s total value. The higher the LTV or CLTV ratio, the greater the likelihood of default. For many Mortgage Loans, the CLTV ratios were far higher than the ratios represented by DLJ. Missing Loan Documents. Many Origination Files were missing key documents that the Mortgage Loans originators would have needed in order to issue the Mortgage Loans in compliance with underwriting guidelines. 7. All told, the Re-underwriting Review showed that 796 of 806 of the examined Mortgage Loans had defects constituting a breach of one or more of DLJ s representations and warranties under the PSAs. (The breaches for these 796 Mortgage Loans are catalogued in detail in the reports attached as Exhibits D (HEMT Trust), E (HEMT Trust) and F (HEMT Trust). These defects alone and in combination meant that the loans did not have the characteristics that DLJ represented them to have, and that there was a higher likelihood that the borrowers would fall behind on their loan payments or would default on their loans altogether, as indeed many of them did. The increased risk of borrowers delinquencies and defaults could materially and adversely affect the performance of the Mortgage Loan pools, and with it, the interests of the Certicateholders therein, across all of the Trusts. 8. Accordingly, on April 23, 2012, a letter was delivered to DLJ calling for the repurchase of 510 breaching Mortgage Loans from the HEMT Trust. On June 22, 2012, another letter was sent to DLJ demanding the repurchase of 244 Mortgage Loans from the HEMT Trust and 42 Mortgage Loans from the HEMT Trust, which loans were similarly in breach, according to the Re-underwriting Review. 9. In addition to the Re-underwriting Review, the Directing Certificateholders 4

5 undertook a separate analysis of the Trusts Mortgage Loans. In December 2011, the Directing Certificateholders instructed the Trustee to inform DLJ that their investigation found additional breaches of DLJ s representations and warranties, based on an Automated Valuation Model analysis ( AVM ) of the Trusts Mortgage Loans (the AVM Review ). An AVM generates statistically-derived property valuations using sales data for comparable properties in the same geographic area as the property under evaluation. Using the same comparable sales data that would have been available to the Mortgage Loans lenders at the time of origination, the AVM Review conducted retroactive appraisals of the properties backing 1,680 Mortgage Loans in the Trusts. DLJ was informed that the AVM Review found that 1,169 of those loans or 69.6% breached DLJ s representations that were based on the accuracy of the value of those properties, such as CLTV ratios, compliance with the lenders underwriting guidelines, and the accuracy of data provided in the Mortgage Loan Schedule attached to each PSA. Specifically, in November and December of 2011, certain of the Trusts Certificateholders notified DLJ that these Mortgage Loans were in breach and demanded their repurchase; the Trustee also provided DLJ with a copy of such notice. 10. Pursuant to the terms of the PSAs, when DLJ discovers that a loan in a mortgage pool has breached any representation and warranty in such a way that materially and adversely affects the interests of Certificateholders, it must either cure such breach or repurchase the loan from the respective Trust. But despite having been made aware of particular defects in the Mortgage Loans constituting breaches of the representations and warranties, DLJ has refused to cure or repurchase any of them. 11. Also pursuant to the terms of the PSAs, DLJ is required to promptly reimburse the Trustee for any actual out-of-pocket expenses reasonably incurred by the Trustee in respect of 5

6 enforcing DLJ s representations and warranties. Despite this clear requirement and despite the Trustee s delivery to DLJ of invoices related to its enforcement of DLJ s representations and warranties, DLJ has refused to reimburse the Trustee. 12. SPS is a mortgage Servicer for the HEMT and HEMT Trusts and a Special Servicer for the HEMT and HEMT Trusts. It is therefore a party to the PSAs for all of the Trusts. SPS is wholly owned by Credit Suisse Holdings (USA), Inc., the same entity that owns DLJ, and thus is an affiliate of DLJ. 13. As a Servicer, SPS has modified hundreds of delinquent Mortgage Loans in the Trusts. On information and belief, this process would have required re-underwriting of the loans. Similarly, SPS has charged off (i.e., passed losses onto the Trusts) and released from the Trusts hundreds of loans another process which, on information and belief, necessitates an analysis of borrowers Origination Files. Thus, on information and belief, when performing loan modifications and charging off and releasing Mortgage Loans, among other servicing activities, SPS likely discovered DLJ s breached representations and warranties with respect to the Mortgage Loans. However, SPS has not notified the Trustee of any such breaches by its affiliate DLJ, the failure of which is in violation of its contractual obligations to do so under Section 2.03 of each of the PSAs. 14. SPS has also obstructed investigation into the full extent of the defects in the mortgage pools by refusing to afford the Trustee reasonable access to the Origination Files and related servicing documents, to which the Trustee is entitled under the applicable PSAs. Exs. A, B, C, Section SPS conduct has contributed to the losses caused by the breaches of its affiliate DLJ. These breaches are material and adverse to the interests of the Certificateholders. Had SPS 6

7 complied with its duty to notify the Trustee of breaches in DLJ s representations and warranties earlier, the Trustee could have taken action sooner to stem losses sustained by the Trusts. Had SPS complied with its obligation under the PSAs to afford the Trustee reasonable access to the Origination Files (see Exs. A, B, C, Section 3.07), the Re-underwriting Review could have analyzed a larger number of Mortgage Loans from the HEMT and Trusts, in order to demand that DLJ repurchase breaching ones. 16. Therefore, the Trusts, acting through the Trustee (acting, in turn, at the instruction of the Directing Certificateholders), now bring this action for breach of contract, specific performance, and declaratory judgment to enforce the obligations of DLJ and SPS under the PSAs. PARTIES 17. Plaintiffs HEMT Trust, HEMT Trust and HEMT Trust are New York common law trusts established pursuant to the PSAs. U.S. Bank National Association is a national banking association, organized and existing under the laws of the United States with its principal place of business in Minnesota, and serves as Trustee for each of the Trusts. 18. Defendant DLJ is a Delaware corporation with its principal place of business in New York, New York. It is a wholly owned subsidiary of Credit Suisse Holdings (USA), Inc. and is primarily engaged in the purchase of mortgage loans. DLJ acted as the Seller for all of the Trusts at issue in the Complaint. 19. Defendant SPS is a corporation organized under the laws of the State of Utah with its principal place of business in Salt Lake City, Utah. It too is a wholly-owned subsidiary of Credit Suisse Holdings (USA), Inc. JURISDICTION AND VENUE 20. This Court has personal jurisdiction over DLJ and SPS pursuant to CPLR 301 and 302 because (i) DLJ s principal place of business is located in New York, New York, and (ii) 7

8 the transactions giving rise to the Trusts claims took place within the state of New York. In addition, the Trusts were formed under New York law and the PSAs are governed by New York law. DLJ and SPS likewise regularly transact business in this State. 21. Venue is proper in New York County pursuant to CPLR 503 because DLJ has its principal offices in and is a resident of New York County. FACTUAL BACKGROUND I. DLJ CREATED SECURITIES OUT OF THREE POOLS OF NEARLY 23,900 RESIDENTIAL MORTGAGE LOANS. 22. This case concerns residential mortgage-backed securities, which are securities secured by a pool of residential mortgage loans. 23. In the securitizations at issue here, DLJ bought the almost 23,900 Mortgage Loans, with an aggregate principal balance of over $1.2 billion, from third party originators, which had made (or originated or underwrote ) the Mortgage Loans to individual borrowers. DLJ then resold such Mortgage Loans to Credit Suisse First Boston Mortgage Securities Corporation ( CS Corp ), the Depositor, pursuant to three Assignment Agreements (one for each of the Trusts), between and among DLJ and CS Corp, dated February 28, 2006 (for the HEMT Trust), June 30, 2006 (for the HEMT Trust) and August 30, 2006 (for the HEMT Trust). 24. The Depositor conveyed the Mortgage Loans to each of the Trusts (also known as depositing the Mortgage Loans), and assigned all of its rights to the Trusts, pursuant to the PSAs. (Exs. A, B and C.) 25. The securities (Certificates) were then issued, backed by the Mortgage Loans, each of which represented an interest in the Trust, to Credit Suisse Securities (USA) LLC, the underwriter for the securities, which in turn offered them for sale to the public. 26. The Trusts are administered by a variety of entities, including the Trustee, which is 8

9 empowered to bring this action; and several servicers, including SPS, which acts as Servicer (for the HEMT and HEMT Trusts) and Special Servicer (for the HEMT and HEMT Trusts), servicing the Mortgage Loans on behalf of each Trust. Through common ownership by Credit Suisse Holdings (USA), Inc, SPS is an affiliate of DLJ. 27. In sum, the Trusts hold the Mortgage Loans on behalf of and for the benefit of the Certificateholders, and the Trustee (on behalf of the Trusts) stands in the shoes of CS Corp with respect to the enforcement of DLJ s obligations under the Assignment Agreements and PSAs, including the cure, repurchase, and indemnification obligations arising from breaches of representations and warranties. II. DLJ S REPRESENTATIONS AND WARRANTIES REGARDING THE MORTGAGE LOANS. 28. Because payments made to Certificateholders depend on payments made by borrowers on the underlying Mortgage Loans, the credit quality of the Certificates hinges on the characteristics of the underlying Mortgage Loans, as represented by DLJ, and the availability of recourse against DLJ if it breached its representations. These and other factors directly affect the cash flow generated by the Mortgage Loans and thus the value of the Certificates. 29. Important information about the characteristics of the Mortgage Loans is contained in the Origination Files. These files normally contain, among other items, the borrower s loan application; verification of the borrower s income, employment, assets, and debts; the borrower s credit report; an appraisal of the mortgaged property s value; and a statement of the property s occupancy status. 30. Investors in the Trusts did not have access to the individual Origination Files, and therefore they assessed the credit risk of the Mortgage Loans and thus the credit risk of their securities through DLJ s representations and warranties concerning the Mortgage Loans. 9

10 31. In the PSAs, DLJ made representations that each of the Mortgage Loans had been originated according to underwriting guidelines designed to ensure the quality of the Mortgage Loans. For example, DLJ represented that the Mortgage Loan complies with all the terms, conditions and requirements of the originator s underwriting standards in effect at the time of origination of such Mortgage Loan. (Exs. A, B, C, Schedule IV (iv).) It represented that each Mortgage Loan at the time it was made complied in all material respects with applicable local, state and federal laws, including, but not limited to, all applicable predatory and abusive lending laws. (Exs. A, B, C, Schedule IV (xx).) Relatedly, DLJ represented that the Mortgage Note and the Mortgage are not subject to any right of rescission, set-off, counterclaim or defense... (Exs. A, B, C, Schedule IV (ix).) 32. In the PSAs, DLJ provided specific data points about each of the Mortgage Loans through a Mortgage Loan Schedule ( MLS ) 1 and represented that the information set forth therein was complete, true and accurate in all material respects as of the Cut-Off Date. (Exs. A, B, C, Schedule IV (v).) 33. DLJ specifically represented that [n]o Mortgage Loan had a combined loan-to-value ratio at the time of origination of more than 100%. (Exs. A, B, C, Schedule IV (xxiv).) 34. A full list of DLJ s representations and warranties appears in Schedule IV to each of the PSAs; they are incorporated here by reference. (Exs. A, B, C, Schedule IV.) 1 The MLS includes important information, such as the balance remaining on a property s senior lien, the original appraisal value of the property, whether the loan is a balloon loan, the loan s current interest rate, the borrower s debt-to-income ratio, the borrower s monthly income and FICO score, whether the property is owner-occupied (i.e., the borrower s primary residence), the original balance of the loan, the original LTV ratio, property type descriptions, and purchase price. 10

11 III. A REVIEW OF THE MORTGAGE LOANS AND ORIGINATION FILES SHOWS BREACHES OF DLJ S REPRESENTATIONS AND WARRANTIES REGARDING THE MORTGAGE LOANS. 35. After the securitization transactions closed, the Mortgage Loans in the Trusts experienced high rates of defaults and delinquencies. As of December 2012, for example, the HEMT Trust has suffered losses of some $204.2 million, or 35.3% of the Trust s original principal loan balance; the HEMT Trust has suffered losses in the amount of $235.9 million, or 58.9% of the Trust s original principal loan balance; and the HEMT Trust has incurred losses of $292.8 million, or 55.7% of the Trust s original principal loan balance. In the first 12 months alone, the HEMT Trust realized $7,876,670 million in losses; the HEMT Trust $22,497,880 million; and the HEMT Trust $40,737,305 million. That losses were experienced so early in the life of the securitizations is strongly indicative of defects in the underwriting and origination of the Mortgage Loans. A. Forensic Review of the Origination Files 36. Directing Certificateholders instructed the Trustee to obtain certain Origination Files from the Servicers. After hundreds of Origination Files were requested and received from certain of the Servicers (but notably not from SPS), forensic reviews of these files were undertaken in April and June Beginning in April 2012, a forensic review was conducted of 806 Mortgage Loans in the Trusts, many of which were defaulted or delinquent. The review, also known as a re-underwriting, found that 796 of the 806 Mortgage Loans breached one or more of DLJ s representations and warranties (510 from the HEMT Trust; 244 from the HEMT Trust; and 42 from the HEMT Trust). The breach rate of the reviewed loans, therefore, was an astonishing 98.8%. 38. Important to the loan underwriting and therefore re-underwriting process are 11

12 documents known as the guidelines. These are the fundamental credit rules that each loan originator is required to follow in deciding whether to make a loan to the borrower and are linked to DLJ s representation that the Mortgage Loan complies with all the terms, conditions and requirements of the originator s underwriting standards in effect at the time of origination of such Mortgage Loan. (Exs. A, B, C, Schedule IV (iv).) These guidelines are not publicly available documents, so the Directing Certificateholders instructed the Trustee to obtain copies of the applicable guidelines from DLJ, the entity who represented that the Mortgage Loans complied with those guidelines. However, DLJ refused to provide them Due to DLJ's refusal to provide the requested guidelines, the Re-underwriting Review tested DLJ s representation that the Mortgage Loans complied with all the terms, conditions and requirements of the originators guidelines at the time of origination with guidelines that, on information and belief, are materially similar to the originators guidelines. Additionally, the Re-underwriting Review found loan approval forms in the Origination Files, which indicated the exact guidelines used for each Mortgage Loan; the Re-underwriting Review then used these loan approval guidelines in examining the Mortgage Loans. 40. All of the guidelines used in, and the breaches uncovered by, the forensic review are described in the reports attached hereto as Exhibits D, E, and F. 3 The reports summarize each of the examined Mortgage Loans, how those Mortgage Loans failed to comply with applicable 2 In addition to directing the Trustee to request the guidelines from DLJ, the Directing Certificateholders also instructed the Trustee to request the applicable guidelines from the Servicers. The Servicers were unable or unwilling to provide guidelines to the Trustee. 3 More detailed reports were provided to DLJ, beginning in April Because the reports include private information about specific borrowers such as addresses and employment status the complete reports could be offered to the Court after the parties and the Court have had the opportunity to discuss a protective order. 12

13 underwriting guidelines, and how they breached DLJ s representations and warranties. What follows is a summary of some of the typical types of these breaches. 1) Borrowers Incomes and Employment Were Misstated. 41. A borrower s income drives his or her ability to repay a Mortgage Loan; the borrower must earn enough money every month to make the required loan payment. A borrower s ability to repay a loan is also affected by his or her employment status; the borrower must have a job that will provide a steady stream of income for the foreseeable future, putting the borrower in a financial position to make the required loan payments over time. If the borrower s income is not commensurate with the amount of the loan payments, and/or the borrower does not have steady employment, then the likelihood of default increases. 42. DLJ represented that [t]he Mortgage Loan complies with all the terms, conditions and requirements of the originator s underwriting standards in effect at the time of origination of such Mortgage Loan. (Exs. A, B, C, Schedule IV(iv).) Similarly, the MLS, whose completeness, truthfulness and accuracy DLJ represented and warranted, contains a representation specifying the borrower s monthly income for each Mortgage Loan in the Trusts. (Exs. A, B, C, Schedule IV (v).) 43. The forensic review found that many borrowers had listed incomes or occupations that were facially unreasonable, should have set off red flags in the Origination Files, and/or were not verified. For example: (a) (b) A mortgage application stated that the lender approved a loan to a borrower employed as a polisher with a stated monthly income of $9,300 per month. $9,300 a month, or $111,600 per year, for a polisher is not reasonable and should have put the lender on notice for a misrepresentation. When an auditor contacted the borrower s employer at origination, the employer verified that the borrower s actual monthly income at the time was $1,548, or less than 20% of the amount represented to the lender. (HEMT Trust.) A mortgage application indicated that the borrower was an HR Supervisor at a medical center in California with a stated monthly income of $15,

14 However, the file s verbal verification of employment showed that the borrower was in fact merely an assistant at the medical center, for whom a monthly income of $15,600, or $187,200 per year, was plainly unreasonable. The U.S. Bureau of Labor Statistics reported that the average monthly salary at the time of origination at the 90 th percentile for the position of medical assistant in the same geographic region was $3,600, or less than 25% of the borrower s stated income. (HEMT Trust.) (c) (d) (e) A mortgage application indicated that the borrower was a sales representative at a car dealership in California with a monthly income of $37,253, or $447,036 per year. A monthly income of $37,253 for a Sales Representative is unreasonable and should have caused a prudent lender to question the borrower s income and employment. An auditor subsequently contacted the borrower s employer at origination who verified that the borrower was actually employed as an Accounting Clerk, not a Sales Representative, and that the borrower s actual income at origination was $1,461, or less than 4% of the stated income. (HEMT Trust.) A mortgage application indicated that the borrower was employed as a manager at Rite Aid with a monthly income of $19,500, or $234,000 per year. A monthly income of $19,500 for a manager at Rite Aid is plainly unreasonable. The U.S. Bureau of Labor Statistics reported that the average monthly salary at the time of origination for a worker at the 90 th percentile with the borrower s position in the same geographic area was $4,821 per month. (HEMT Trust.) A mortgage application stated that the borrower was employed in Florida as a nail technician with a monthly income of $5,500. Although the borrower stated that she was employed as a nail technician for four years with the listed employer, the file contained a written verification of employment stating she was employed for 2.5 years there. The file contained no evidence that the discrepancy was investigated. The borrower s bankruptcy petition reflects $12,700 in gross annual income in 2005, or $1,058 per month. The borrower s actual DTI ratio was therefore 213%, grossly exceeding the 50% maximum based on applicable underwriting standards. (HEMT Trust.) 44. Such misstatements of borrowers income and employment breach many of DLJ s representations and warranties, including without limitation, its representations about the Mortgage Loans consistency with the originators underwriting guidelines and that the information on the MLS was complete, true and accurate in all respects. Where there is a breach of a representation and warranty regarding a borrower s income and/or employment, the risk of 14

15 default is greater, and thus the value of the Mortgage Loan is diminished. Therefore, such breaches by DLJ materially and adversely affect the value of the Mortgage Loans and the Certificateholders interests in the Mortgage Loans. 2) Borrowers Debts Were Misstated. 45. A borrower s level of debt impacts her ability to repay a loan; the less debt the borrower has at the time she takes out the loan, the more funds she will have available to make her loan payments. The applicable underwriting guidelines required underwriters to consider an applicant s installment debts, revolving debts, and home equity lines of credit in determining the applicant s liabilities. A borrower must not have a debt load that would make it onerous or impossible for her to make the required payments. The greater the borrower s debt, the greater the likelihood that she will default. 46. In the MLS, DLJ represented the completeness, truthfulness and accuracy of the borrower s debt-to-income ratio for each Mortgage Loan in the Trusts, which encompasses representations about the borrower s monthly debt obligations (the numerator of the ratio). (Exs. A, B, C, Schedule IV (v).) 47. The forensic review found that many borrowers had misrepresented their debts, frequently not disclosing recent home purchases. For example: (a) (b) A mortgage application failed to disclose all of the properties owned by the borrower at the time of origination and thus portrayed a false picture of his ability to repay financial obligations. On April 14, 2005, one day prior to the subject loan s closing date, the borrower purchased a property in Arizona with liens totaling $267,800 and a monthly payment of $1,570. The borrower s ownership of this property was not reflected on the loan application at the time the subject loan closed. (HEMT Trust.) A mortgage application failed to disclose all of the properties that the borrower owned at the time of origination. Three months prior to the closing of the subject loan, the borrower purchased a property in Illinois with liens totaling $270,000 and a monthly payment of $1,652. Then, one month prior to the subject loan s closing, the borrower purchased yet 15

16 another property with liens totaling $196,500 and a monthly payment of $1,308. The borrower s ownership of these additional properties was not reflected on the loan application for the subject loan. (HEMT Trust.) (c) A mortgage application failed to disclose that, weeks before the subject loan closed, the borrower purchased a property with liens totaling $317,200 and a monthly payment of $2,842. The borrower s ownership of this property was not reflected on the loan application at the time the subject loan closed. (HEMT Trust.) 48. Such misstatements of borrowers debts breach many of DLJ s representations and warranties, including without limitation, its representations about the Mortgage Loans consistency with underwriting guidelines and that the information on the MLS was complete, true and accurate in all respects. Where there is a breach of a representation regarding a borrower s level of debt, the borrower s ability to repay such a Mortgage Loan is not what it has been represented to be; the risk of default is greater, and thus the value of the Mortgage Loan is diminished. Therefore, such breaches by DLJ materially and adversely affect the value of the Mortgage Loans and the Certificateholders interests in the Mortgage Loans. 3) Debt-to-Income Ratios Were Misstated. 49. A key measure of a borrower s ability to afford mortgage payments is his debt-to-income ratio (or DTI ratio), which compares the borrower s monthly debt obligations to his monthly income. The higher the DTI ratio, the larger the percentage of his monthly income that a borrower must devote to debt payments. Thus, the higher the DTI ratio, the greater the likelihood that the borrower will default. 50. The underwriting guidelines set out maximum DTI ratios for different types of loan products in a series of matrices; the maximum DTI ratios for products listed on those matrices 16

17 ranged from 45%, to 50% or 55%. 5 DLJ warranted the accuracy and truthfulness of the MLS, and made representations regarding DTI ratios for Mortgage Loans in the Trusts, as set forth on the MLS. 51. The forensic review found that the true DTI ratio for many of the borrowers was actually higher than that represented and warranted by DLJ on the MLS, and was often much higher than the maximum DTI ratio allowed by the applicable underwriting guidelines. For example: (a) (b) (c) A Mortgage Loan was made subject to a maximum DTI ratio of 50% to a borrower whose stated monthly income was $8,380. However, the borrower s publicly available Chapter 7 bankruptcy petition, filed one month after the subject loan closed, reflected a total of $5,106 in gross annual income, or $ per month in The borrower s DTI ratio at his actual income level was therefore %, grossly exceeding the 50% maximum under the applicable underwriting guidelines or under any other conceivable set of guidelines and the DTI figure represented by DLJ on the MLS. (HEMT Trust.) A Mortgage Loan was made subject to a maximum DTI ratio of 45% to a borrower whose stated monthly income was $5,850. However, the borrower s petition for Chapter 7 bankruptcy reflects a total of $16,992 in gross annual income, or $1,416 per month in 2005, when the subject loan was closed. Combined with the borrower s undisclosed property purchases at the time of the subject loan s closing date, the borrower s actual DTI ratio was %, significantly exceeding the 45% maximum allowed and the DTI figure represented by DLJ on the MLS. (HEMT Trust.) A Mortgage Loan was made to a borrower who stated he was employed as a painting company manager with a monthly income of $6,000. However, the file reveals that the borrower s initial loan application stated that his actual income was $3,500 per month, which was changed two months later to nearly twice that amount. Indeed, the employment information section of the borrower s credit report at origination lists the borrower s employment as McDonald s employee. The borrower s DTI using the $3,500 per month salary was approximately 70% higher than that allowed by any 5 For example, one set of guidelines stated when compensating factors are used, under no circumstances may a total DTI exceed 55%. (Ex. G, MortgageIT Alt-A Guidelines Version June 21, 2005 as revised on March 20, 2006, ) (emphasis in original.) 17

18 conceivable set of underwriting guidelines and higher than the figure represented by DLJ on the MLS. (HEMT Trust.) 52. Such misstatements of DTI ratios breach many of DLJ s representations and warranties, including without limitation, its representations about the Mortgage Loans consistency with underwriting guidelines and that the information on the MLS was complete, true and accurate in all respects. Where there is a breach of representation regarding DTI ratios, the borrower s ability to repay such a Mortgage Loan is not what it has been represented to be; the risk of default is greater, and thus the value of the Mortgage Loan is diminished. Therefore, such breaches by DLJ materially and adversely affect the value of the Mortgage Loans and the Certificateholders interests in the Mortgage Loans. 4) Properties Owner-Occupancy Statuses Were Misstated. 53. Whether a borrower occupies a mortgaged property impacts the chances that he will repay the Mortgage Loan. Borrowers who live in mortgaged properties are less likely to walk away from those properties and to default on their mortgage obligations than are borrowers who buy residential properties as investment vehicles or as vacation homes. If a property is not owner-occupied, then the likelihood of the borrower defaulting increases. 54. The guidelines required appraisals to state whether the property was owner-occupied. (Ex. G, MortgageIT Alt-A Guidelines Version June 21, 2005 as revised on March 20, 2006, ) If the property was not going to be occupied by the owner, then the guidelines prohibited issuing the Mortgage Loans in many situations. (See, e.g., Ex. H, DLJ-Credit Suisse First Boston Second Lien Piggyback, revised June 21, 2005) (second home properties with greater than one unit or co-ops are not eligible.) The guidelines also used occupancy status to determine the level of reserve funds a borrower was required to have in the bank. (See, e.g., Ex. G, (borrowers of loan amounts of less than $650,000 occupying 18

19 properties had to have 2 months of PITI the sum of their monthly principal, interest, taxes, and insurance payments but borrowers for investment properties had to have 6 months).) 55. DLJ made representations and warranties regarding the owner-occupancy status of Mortgage Loans in the Trusts, as reflected on the MLS. 56. The forensic review found that many of the subject properties listed as owner-occupied were not, in fact, occupied by the borrower. For example: (a) (b) (c) A Mortgage Loan that closed in April 2005 was represented as owner-occupied, but a January 2012 credit report reflected that the borrower continued to reside at the previous address as indicated in a loan application in Texas from March 1999 through January 2012, which is after the subject loan closed. The borrower s driver s license records indicate the borrower s address as the previous address in Texas, as does a QC credit report from January (HEMT Trust.) A Mortgage Loan that closed in January 2006 was represented as owner-occupied, but the QC credit report in the Origination File shows that the borrower would continue to reside at the departure address from November 2004 past the subject loan s closing date. A May 2012 credit report confirms that the borrower continued to reside at its previous address in Granada Hills, California after the subject loan s closing date. (HEMT Trust.) A Mortgage Loan was represented as owner-occupied, yet the Origination File contains a letter from the borrower to the effect that he was giving this property to his sister and brother-in-law. An April 2012 credit report confirms that the borrower continued to reside at his original residence in Vista, California after the closing of the subject loan. (HEMT Trust.) 57. Such misstatements of properties occupancy status breach many of DLJ s representations and warranties, including without limitation, its representations about the Mortgage Loans consistency with underwriting guidelines and that the information on the MLS was complete, true and accurate in all respects. Where there is a breach of representation regarding property occupancy status, the borrower s ability to repay such a Mortgage Loan is not what it has been represented to be; the risk of default is greater, and thus the value of the Mortgage 19

20 Loan is diminished. Therefore, such breaches by DLJ materially and adversely affect the value of the Mortgage Loans and the Certificateholders interests in the Mortgage Loans. 5) Combined Loan-to-Value Ratios Were Misstated. 58. Another key measure of a borrower s likelihood of default is the loan-to-value ratio (or LTV ratio), which expresses the amount of the Mortgage Loan as a percentage of the total appraised value of the property. For example, if one borrows $160,000 to buy a house worth $200,000, then the LTV ratio is $160,000/$200,000, or 80% the remaining $40,000 of the house s value reflects the borrower s 20% equity stake in the house. A combined-loan-to-value ratio (or CLTV ratio) the ratio most pertinent to second-lien trusts, such as the subject Trusts expresses the amount of all of the Mortgage Loans secured by a single property as a percentage of that property s appraised value; that is, it accounts for any additional liens. 59. A borrower with a sizable equity stake in a property has more to lose than a borrower with a small stake; a borrower with a large equity position has financial incentives not to default and lose his equity. Conversely, a borrower with little or no equity stake in a property has little financial incentive to avoid default. And if a borrower with a small equity stake does default, then there is a greater risk that a resulting foreclosure will result in a loss for the lender or, in this case, for the Certificateholders who are entitled to the Mortgage Loans cash flows. The higher the LTV and CLTV, the greater the likelihood that the borrower will default. 60. The underwriting guidelines restricted the types and amounts of loan products available to borrowers based on the subject properties CLTV ratios. (See, e.g., Ex. G, MortgageIT Alt-A Guidelines Version June 21, 2005 as revised on March 20, 2006, ) (CLTV ratios in product matrices are maximums on Mortgage Loans).) The guidelines set out the maximum CLTV ratios allowed for loan products in a series of matrices; in general, the greater the amount of the loan, the lower the maximum CLTV ratio. These ratios ranged from 80% 20

21 to 100%, depending on the product. 61. The guidelines also used CLTV ratios to set limits on how the Mortgage Loans could be financed. The guidelines further used CLTV ratios to determine the minimum amount of documentation a borrower would have to provide and what the borrower s minimum FICO scores could be. (Id., ) 62. DLJ represented that none of the Mortgage Loans in the mortgage pools had a CLTV ratio of over 100%. (Exs. A, B, C, Schedule IV(xxiv).) DLJ warranted the accuracy and truthfulness of the MLS, and made representations regarding the CLTV ratios for Mortgage Loans in the Trusts, as set forth on the MLS. 63. The forensic review found that the true CLTV ratios for many of the properties at issue were higher than those represented by DLJ in the MLS, and were often materially higher than the maximum CLTV ratios allowed by the guidelines. For example: (a) (b) (c) A lender s guidelines permitted a maximum CLTV ratio of 90% for stated income, cash out refinance loans secured by an owner-occupied property for a borrower with a credit score of 680. Even assuming the accuracy of the original appraisal of the property, the subject borrower s CLTV ratio was 10% higher than the guidelines maximum. (HEMT Trust.) A lender s guidelines permitted a maximum CLTV of 80% for a stated income document loan in a combined amount of $842,000 secured by an owner-occupied property for a borrower with a credit score of 709. Even assuming the accuracy of the original appraisal of the property, the subject borrower s CLTV exceeded the maximum by more than 17%. (HEMT Trust.) A Mortgage Loan was made on September 9, 2005 after an appraisal that did not analyze the prior sale of the property less than a year earlier. The property s appraised value at closing represented an 81% increase in value in 18 months. The recalculated appraised value resulted in a CLTV ratio of 125.7%. (HEMT Trust.) 64. Such misstatements of properties CLTV ratios breach many of DLJ s representations and warranties, including without limitation, its representations about the 21

22 Mortgage Loans consistency with underwriting guidelines, the represented number of Mortgage Loans on the MLS with specific CLTV ratios, and the representation that no loan had a CLTV of over 100%. Where there is a breach of representation regarding a property s CLTV ratio, the borrower s ability to repay such a Mortgage Loan is not what it has been represented to be; the risk of default is greater, and thus the value of the Mortgage Loan is diminished. Therefore, such breaches by DLJ materially and adversely affect the value of the Mortgage Loans and the Certificateholders interests in the Mortgage Loans. 6) Key Documents Were Missing from the Origination Files. 65. To underwrite a loan in accordance with applicable guidelines, an originator must have a complete set of documents containing the information relevant to those standards. Some documents are expressly required by federal law (e.g., HUD-1 settlement statement), while others must be examined pursuant to the underwriting guidelines (e.g., employment verification). When a document required by the guidelines or by law is missing from a Origination File, it is possible that the document either was not verified during the underwriting process or was deliberately removed from the file because it did not support making the loan. 66. DLJ represented that [t]he Mortgage Loan[s] compl[y] with all the terms, conditions, and requirements of the originator s underwriting standards in effect at the time of origination of such Mortgage Loan[s]. (Exs. A, B, C, Schedule IV (iv).) CS Corp stated in the Prospectuses that the mortgagor will have furnished [to the lender] information with respect to its assets, liabilities, income[], credit history, employment history and personal information and that, based on such information, a determination is made by the original lender that the mortgagor s monthly income... will be sufficient to enable the mortgagor to meet its monthly obligations on the mortgage loan and other expenses. (Ex. I, HEMT Prospectus, p. 30; other prospectuses substantially similar.) 22

23 67. The guidelines set forth additional and specific documentation requirements for different types of loan products. 68. The forensic review found that many of the material documents were missing from the Origination Files. For example: (a) (b) The file for a loan in the amount of $87,750 was missing all credit documents, the first lien mortgage note, a qualified appraisal, a Uniform Residential Loan Application (Form 1003), a flood certificate, hazard insurance, the HUD-1 6 settlement statement from the first lien transaction, a copy of the Mortgage/Deed of Trust, the subject mortgage note, and all legal compliance documentation. (HEMT Trust.) The lender s underwriting guidelines required it to obtain a verbal verification of employment ( VVOE ) to document the borrower s current employment, yet the file s VVOE was missing the borrower s position, start date, and a date the verification was completed if it was completed at all. Although the borrower stated in his loan application that he earned $21,000 per month as a company Branch Manager, his tax returns in the file reflect that his true position was a Mortgage Broker/Owner, not a Branch Manager, and that his actual income at the time of origination was negative $27,355. (HEMT Trust.) 69. The absence of such key documents from the Origination Files breach many of DLJ s representations and warranties, including without limitation its representations about the Mortgage Loans consistency with underwriting guidelines and compliance with applicable state and federal laws. Where there is a breach of a representation regarding Origination File documentation, the value of the Mortgage Loan is diminished. Therefore, such breaches by DLJ materially and adversely affect the value of the Mortgage Loans and the Certificateholders interests in the Mortgage Loans. 6 A HUD-1 settlement statement is a standard mortgage form which itemizes the services and fees charged to the borrower by the lender or broker when applying for a loan for the purpose of purchasing or refinancing real estate. 23

24 B. The Directing Certificateholders AVM Investigation into DLJ s Representations and Warranties 70. In November 2011, before any of the Origination Files were acquired from the Trusts Servicers, an initial investigation into the Trusts Mortgage Loans and the accuracy of DLJ s related representations and warranties was commenced by the Directing Certificateholders, based on the aforementioned AVM analysis. An AVM considers objective criteria such as the condition of a property and sale prices of comparable properties in the same locale as the subject property to determine the true market value of the property in question. The AVM model employed here encompasses a database of 500 million sales, covering ZIP codes that represent more than 97% of the homes occupied by upward of 99% of the population of the United States. 71. Without the aid of the Origination Files themselves, this investigatory phase centered on three of DLJ s core PSA representations: that the information in the MLS was complete, true and correct in all material respects (Exs. A, B, C, Schedule IV (v)); that the Mortgage Loans complied with all the terms, conditions and requirements of the originators underwriting standards (Exs. A, B, C, Schedule IV (iv)); and that no Mortgage Loan had a combined loan-to-value ratio at the time of origination of more than 100% (Exs. A, B, C, Schedule IV (xxiv)). 72. Where the appraised value of a given property collateralizing a Mortgage Loan as represented on the MLS was at least 10% higher than the value reported by the AVM, the investigation identified a breach of DLJ s representation in Schedule IV (v) to the PSAs. Similarly, where an LTV or CLTV ratio as represented on the MLS was materially understated vis-à-vis a recalculated LTV or CLTV (whose denominator, again, is the value of the property) based on the appraisal value reported by the AVM, or where the AVM registered a property value which resulted in a CLTV ratio greater than 100%, the investigation identified a breach of DLJ s 24

25 representations in Schedule IV(v),(iv) and (xxiv) to the PSAs. 73. Of the sample of 406 Mortgage Loans examined in the HEMT Trust, the AVM investigation determined there were breaches of these representations and warranties in 288 (or 70.9% of the loans). Of the sample of 721 Mortgage Loans in the HEMT Trust, the AVM found that 522 of the loans were in breach (or 72.4% of the loans). Finally, of 553 sampled loans in the HEMT Trust, the investigation found 359 were in breach (or 65% of the loans). IV. DLJ FAILED TO REPURCHASE MORTGAGE LOANS AFTER BEING MADE AWARE OF BREACHES OF REPRESENTATIONS AND WARRANTIES REGARDING THE MORTGAGE LOANS. 74. Under the PSAs, when DLJ discovers or receives notice from a party to the PSAs that any Mortgage Loan is in breach of one or more of its representations and warranties, it must cure the breaching loan in all material respects within 120 days or, failing that, buy back the loan from the Trust in question. In relevant part, Section 2.03(g) of the PSAs for the HEMT and HEMT Trusts (and PSA Section 2.03(f) of the HEMT Trust) provides: [u]pon discovery by any of the parties hereto of a breach of a representation or warranty made [by DLJ] that materially and adversely affects the interests of the Certificateholders in any Mortgage Loan, the party discovering such breach shall give prompt notice thereof to the other parties. The Seller hereby covenants that within 120 days of the earlier of its discovery or its receipt of written notice from any party of a breach of any representation or warranty made by it pursuant to Section 2.03(f) which materially and adversely affects the interests of the Certificateholders in any Mortgage Loan sold by the Seller to the Depositor, it shall cure such breach in all material respects, and if such breach is not so cured, shall... repurchase the affected Mortgage Loan from the Trustee at the Repurchase Price... Exs. A, B Section 2.03(g); Ex. C Section 2.03(f). 75. In the RMBS securitization industry, and therefore with respect to these Trusts, repurchase is understood to mean a mechanism to make a purchaser of mortgage loans whole for the seller s breaches of representations and warranties. The repurchase mechanism therefore 25