Mortgage fraud may be both more prevalent than lenders know and more insidious than they realize according to Bridget Berg, a principal in Industry Solutions and Property Intelligence for CoreLogic.  In the first of a three-part series in the company's Insights blog, she says mortgage fraud is different from other types in that it takes a long time to present itself and can pop up at multiple points over the life of the loan.  By the time it does, it may be too late to mitigate the situation.

The delay in discovery can lull lenders into a false sense of security. They might assume, if they haven't spotted problems, had a large loss, or have a low delinquency rate that their controls are working and they've had no fraud.

This, she says, may "all be true-until the day it isn't."  She lays out five reasons why lenders should be wary when everything seems fine.

1. Fraud is relatively rare; less than 1 percent of mortgage applications containing material misrepresentations. Most can be categorized as "fraud for property:" that is borrowers misrepresenting their qualifications for the mortgage. More serious fraud schemes - fraud for profit - often involve multiple players, such as appraisers, loan officers and straw buyers, and are rarer. But the potential for large losses, particularly with fraud for profit, makes fraud a serious problem.
2. Fraud is often undetected. When the economy is strong or property is appreciating rapidly, borrowers who committed fraud-for-property may stay current on their loans. However, if the economy slows or home prices correct, these loans are more likely to default. Even then they are likely to be treated as a credit loss rather than origination fraud.
3. Fraud is cyclical and moves with the market and current opportunity. Before the housing crisis, fraud tended toward falsified down payment sources, inflated income, and straw buyers. In its aftermath fraudsters shifted to loss mitigation fraud, such as short sale schemes, REO bid rigging, and loan modification scams.
4. No one likes to air their dirty laundry. Insider fraud involving employees or agents can damage a lender's brand. Information may also be compartmentalized within a company when investigations are shielded by attorney-client privilege.
5. Many types of fraud are detected immediately, such as stolen credit cards which are usually used immediately.  Mortgage fraud is more insidious and may include measures to delay discovery, such as faked employment verifications or initially keeping payments current.  

These create a false sense of security, but Berg says the risk of fraud is always there; shifting over time and operating right under lenders' watchful eyes. Understanding fraud and creating and using ongoing monitoring is necessary to loss prevention.

Berg says she's found industry participants believe that discovering fraud mainly occurs during origination, and verification processes, underwriting, and pre-funding quality control are good tools to identify it before a loan goes on the books.  However, few lenders have installed strong tracking mechanisms to determine how much fraud is being averted, which loans or loan types, and which methods accounted for the detection. "This could be a valuable dataset to focus operational controls and to understand trends vs. treating fraud as a one-off," she says.

The standard post-funding lender QC review is another point where fraud could be discovered, but these reviews are not fraud-specific and the sampling usually involves fewer than 10 percent of loans. QC reviews are usually done in the first two or three months after closing and involve a reverification of credit reports, income, and assets. Thus, any fraud discovered is likely to be undisclosed liabilities or job loss that happened prior to closing. There are loopholes here as well.  While credit reports will usually uncover most new debts, asset and income reverifications may not be returned, or the latter, if part of a collusive scheme, returned with false information again.

Investors typically perform some level of QC in the first few months after acquiring the loan, but again these are not focused on fraud and are also most likely to find issues like undisclosed liabilities and job loss prior to closing. Investors usually have options for indemnifications or repurchase requests.

Loans that go seriously (60 to 90 days) delinquent in their first 6 to 12 months are often routed to a special QC review. Fraud found at this point will represent a much higher percentage of the sample. Unsophisticated schemes, such as a defaulted one-off straw buyer flip, are often detected in this type of review. Indemnification or repurchase requests are likely where reps and warrants are in the contract.

Many investors will perform QC or root cause analysis on loans with large losses, even those older than a year.  "Egregious frauds and schemes" can be detected here, and recourse will be sought from the originator.

Fraud investigations, usually triggered by a tip about a specific loan, or persons involved in originations, can take place at any point in the life of the loan.  The investigation may widen as related loans are identified. Berg said this is the point where most fraud schemes are fully identified and reps and warrants, if applicable, invoked.

In her final post, Berg says the common long discovery delay, with different types of loans likely to emerge at different times, means that, by the time a problem is identified, it may be too late to mitigate the loss or it has grown large. The lag time also gives management a false sense of security about risk levels and can allow poor practices or an undetected scheme to continue for years.  Those types of fraud that are most likely to be discovered early on are also not representative of overall risk.

When a known or suspected mortgage fraud is identified, lenders must file a Suspicious Activity Report (SAR) with the Treasury Department's Financial Crimes Enforcement Network (FinCEN) within 30 days of discovery. These reports are tracked on-line but only the date of the filing is tracked, not when the actual fraud took place.

In 2013, FinCEN provided the chart below, showing the lag between fraud occurrence and its report. It showed the difference between the peak of reporting and the peak of activity was about five years, but the lag is heavily influenced by the economy and by close examination of defaulted loans. The shift in the period around 2008 demonstrates this.

Most secondary market transactions carry a life-of-loan warranty.  When the number of serious delinquencies and foreclosures ballooned in 2008, investors were motivated to discover any misrepresentations so they could ask for a repurchase or reimbursement from the institution that sold them the loan.

Berg offers some best practices which can help lenders protect themselves despite the usual delays in discovery.  Here are her suggestions.

• Add structured data fields within tracking systems to capture the timing and discovery date, the type of fraud and how it was uncovered.
• Develop tracking of fraud by vintage and seasoning age. Report on each origination/ application year separately and compare them using consistent seasoning ages.
• Track cycles by including as many years of history as possible and include both pre- and post-funding findings.
• Track changes in controls or policies that might impact fraud risk, such as income verification policies or reduced documentation programs. Comparing fraud activity to these policy changes can lead to knowing their positive or negative impacts.
• Track fraud by origination channel, third party originators, loan programs, or loan purpose. These may be leading indicators of trends.

Berg says fraud is present, whether a lender sees it or not. Vigilance is important.