Technologies and techniques

Many lenders concerned with fraud use predictive-scoring technologies to detect the often-hidden patterns of fraud and risk that otherwise might slip through the cracks. Technology that combines historical-pattern detection based on a large consortium of good-performing and fraudulent loans with behavioral trends for entities involved in the application can help lenders identify the riskiest loans.

This type of technology differs from the database-validation and collateral-risk tools that use comparisons between application information and third-party data sources. Advanced pattern-recognition scoring techniques, long used in the credit and debit-card industry, enable lenders to achieve greater predictiveness and to review and verify only the highest-scoring loans.

Lenders diligently review applications that receive high-risk scores to verify the borrower's income, employment, occupancy status, assets and identity, as well as to validate that the collateral is valued appropriately.

These technologies also can identify potential fraud issues with other parties in the transaction, such as brokers or individuals within a broker's organization. Lenders use pattern-recognition technology to determine which brokers they should put on watch or cease doing business with.

Before they submit a loan for processing, brokers can use the same techniques that lenders have found successful to find fraud or insufficient-payment capacity. By simply submitting application information, brokers can retrieve risk scores and reports and act upon them before submitting the file to the lender.

If used correctly, these scoring technologies can help prevent a great deal of fraud attempts. By taking advantage of these technologies, brokers can maintain their reputations and stop problems before they start.

Risk scores and credit data

Although credit scores are used as a baseline for determining borrowers' ability to pay on time, they are not a complete assessment of risk. Fraud scores and early-payment-default scores, which can be calculated using predictive-analytics technology, often can better identify the fraud risk.

These scores identify specific applications or loans that should be reviewed further. These scores may better predict a loan's default risk than a credit score alone.

Also, when used as a research tool on high-risk applications, credit reports become a valuable tool. Most credit reports contain a variety of data elements that can be used to assess the borrowers' fraud risk.

There are five fraud types in particular that brokers can help uncover if they leverage the credit report's full value:

1. Piggybacking: This is when borrowers are added as an authorized user on another person's credit account, rather than building their own credit history. Although piggybacking is technically legal, lenders are scrutinizing those borrowers' credit reports much more intensely and denying credit if it appears that it is not the borrower's own payment history.

Often, borrowers can sign up with full-service solutions on the Internet whereby they pay a fixed sum to have three to five credit accounts added to their credit report for a period of six to nine months. If done correctly, piggybacking can artificially boost a person's credit score by 100 points or more.

If borrowers are listed as authorized users on more than 50 percent of their accounts, it may be a piggybacking scheme. Another indicator of potential piggybacking is if the date the credit file was established is newer than the oldest date open on their accounts.

2. Straw borrowers: These are people whose good credit is used by someone with bad credit -- whether knowingly or unknowingly. One indication of a straw-borrower scheme is if borrowers have less than four tradelines on their credit report and a good credit score.

Page: 1 2 3 Next