Findings Observed from Sampled Narratives

A sample of 1,769 depository institution SAR narratives was reviewed to identify additional trends and patterns reported in those narratives. Comparisons to the findings in the FinCEN report published November 2006 were made whenever possible. The percentages presented frequently do not add up to 100% because not all narratives provided sufficient information to determine >
Types of Fraud

Mortgage fraud is generally divided into two broad categories: fraud for housing and fraud for profit. Fraud for housing was the most common type reported in the sampled narratives (60%).30 Fraud for profit was reported in just over 36% of the sampled narratives.

30 For this study, occurrences are >
Figures 17 and 18 displays the types of participants in these fraud categories and show the frequency of their mention in each category.





Reports describing suspected fraud for housing referenced purchase loans most often, followed by refinance, 2nd trust, and home equity loans. All reports regarding construction loans described suspected fraud for profit. Home equity loans had the second highest percentage of fraud for profit with 2nd trust, refinance, and purchase loans showing the next highest percentages.

Figure 19 illustrates a comparison of the type of fraud by loan type as seen in the sampled narratives.



Loan Types

Loans for purchasing houses, either for a primary residence, second home, or investment, were the most commonly reported loan types detailing suspected fraud, at 72.75%. Other types of loans reported were: refinance (12.04%), home equity (3.5%), 2nd trust (2.37%), and construction (1.07%). Some significant changes were found by comparing loan types reported in FinCEN?s previous mortgage fraud report to loan types reported during the update period. The percentage of fraudulent construction loans and purchase loans reported experienced a decrease while reports of fraud in 2nd trust, refinance, and home equity loans increased.

Figure 20 displays the comparison.



Filers specified that loans were subprime in 79 SARs (0.19%) for the reviewed period. Without this specification, it is not possible to determine whether mortgages described in the remaining SARs were subprime loans.

Filers did not identify any FHA Title One loans in the sampled narratives reviewed for this update report. It is unknown if there was a decrease in reports of fraud in FHA Title One loans, or if the filers simply did not identify the loans as such. Filers did note that six purchase loans and one refinance loan were FHA insured loans.

Figure 21 provides a comparison of loan types for the initial and updated reports.



Filers noted in the sampled narratives that 54 (25.35%) of the refinance loans were ?cash-out refinance.? Additionally, filers noted that 7.41% of the cash-out refinance loans were early defaults; half of those were first payment defaults.

Early Payment Default

Filers reported that early payment defaults triggered suspicion that loans may have been obtained through fraudulent methods in 71 (4%) of the sampled narratives. Twenty-five (35.21%) of those narratives specified a first payment default. Filers reported early payment defaults were moderately more common in fraud for profit (57.75%) than fraud for housing (42.25%).

Figure 22 displays the types of loans where early payment defaults were detected.



Figure 23 provides a comparison of suspected fraud for profit and fraud for housing by loan type.



Stated Income/Low Document or No Document Loans

Filers reported in 69 (3.90%) of the sampled narratives that the reviewed loans were Stated Income, Low Document or No Document loans. Mortgage brokers originated nearly 80% of these loans. Filers reported that fraud for housing (49.28%) and fraud for profit (47.83%) were nearly equally represented in these loans. Nearly 9% of these loans were early payment defaults; 50% of those were first payment defaults.

Figure 24 below displays the types of loans granted as low/no document or stated income.



Fraud Detection

Filers reported they detected the possibility of fraud in various phases of the loan process: pre-finance, post finance audit, loan default; and through reports by victims, law enforcement, and even the borrowers themselves. SARs noting detection during post finance audits also reported that the loans were performing and current at the time the SARs were filed.

Figure 25 below displays a comparison of when the suspected fraud was detected in FinCEN?s initial report to when it was detected in the updated report. The comparison shows that there was nearly a 50% increase in the percentage of SARs specifying fraud detection prior to loan funding. SARs reporting that the filers detected possible fraud after loan defaults increased nearly 23%. As shown in Figure 25, fraud detection by law enforcement increased by 71%. Filers reported they were contacted by law enforcement to report that their customer was under investigation for loan fraud or to subpoena records for their investigation.



As shown in Figure 25 above, there was a more than 59% increase in detection through contact by victims of fraud, mostly identity theft cases. One explanation for the increase in victim reports could be greater consumer awareness of identity theft and greater use of free annual credit bureau checks, resulting in more frequent credit report checks.

Figure 25 also shows a nearly 88% increase in the reports of borrowers contacting lenders to request a change in the Social Security Number associated with their loans. The borrowers were, in effect, revealing that they used a fraudulent Social Security Number at the time the loan was initiated.

Excerpted from Mortgage Loan Fraud: An Industry Assessment based upon Suspicious Activity Report Analysis -April 2008, page 44