BasePoint Analytics

This message contains graphics. If you do not see the graphics, click here to view.



	Fighting Fraud with Frank May 14, 2007 Investment Banking: Reducing Fraud Exposure

"...up to 70% of early payment defaults could have been avoided with better pre-funding or pre-purchase reviews..."

Welcome to the first edition of "Fighting Fraud with Frank." I am Frank McKenna, the Chief Fraud Strategist with BasePoint Analytics. BasePoint develops advanced pattern recognition models that root out mortgage fraud by analyzing patterns within the loan application.

One of the best parts of working for a company like BasePoint is that we get the chance to work directly with many lenders and investment banks to understand the issues that they face with fraud, early payment defaults and rep and warrant issues involving misrepresentation. We get a chance to see what strategies and practices are most effective in stopping fraud. I firmly believe that sharing information on best practices for fraud management will be one way to help the mortgage industry solve the issues of fraud that exist today. I want to begin sharing some of those best practices with you on a regular basis. To that end, I will be sending out periodic updates on new fraud trends, and practical tips on how to address them in this newsletter.

The Secondary Market is Paying Close Attention to Fraud
Perhaps one of the biggest changes in the market recently is the intense focus on fraud by the secondary market. The secondary market is seeing first hand that some of the loans originated by lenders still include misrepresentations. A key component at play here is the fact that misrepresentations in loan applications clearly affect the early performance of the loans. Our studies based on historical loan data suggest that up to 70% of early payment defaults could have been avoided with better pre-funding or pre-purchase reviews.

Investment Banking firms are taking matters into their own hands now. Rather than relying solely on lenders to root out fraud on their behalf, they’ve taken direct responsibility for keeping risky loans out of their portfolios. Many of the top investment banks and due diligence companies have recently adopted predictive analytics technology to determine the specific level of fraud in loan pools and, as a result, they are rejecting significantly more suspect loans before they buy.

Due Diligence Now Incorporates Specialized Fraud Reviews
In the past, investment banks have relied primarily on credit risk factors to predict risk in their samples before they invested in a pool. They used debt-to-income ratios, credit scores – primarily credit and compliance-related metrics. Today, they’re opening it up further to understand true fraud risk. This is a dramatic change that could be the impetus for the establishment and adoption of much-needed fraud management standards in the mortgage industry. Investment banks are instructing their due diligence partners to conduct a specific review for suspicious loans and potential misrepresentation so they are buying loan pools that are as “clean” as possible.

6 Steps Investment Banks Can Take to Reduce Fraud Risk
Here are the 6 initiatives I believe present the biggest opportunity for Investment Banks to control fraud:

Sample Out Your Fraud Risk - Use predictive analytics to identify loans most likely to contain fraud risk. This enables you to review the smallest number of loans to keep costs down while still uncovering the richest concentrations of fraud. Creating a smart sample allows you to concentrate the fraud risk to a reviewable population which is then scrutinized much more carefully. Smart samples can typically identify 60% or more of the fraud risk within just a few percent of the total loans in the bid tape.
Standardize the Due Diligence Review - Insist on a standardized review process with your due diligence company that identifies and reports back specifically on fraud. Before a loan purchase is completed, make sure that the information provided by the borrower for highest risk loans have been vetted by an experienced underwriter. For example, is the income really reasonable given the occupation, age and number of tradelines on the borrower’s credit report?
Align with Lenders' Pre-Funding Fraud Checks - Lenders with better pre-funding fraud reviews have lower rates of early payment default and fewer rep and warrant issues related to fraud. By working with lenders to institute better practices up-front, investment banks will be ensuring better quality loans in the future.
Autopsy Review – Investment banks that review their “rejects” and non-performing loans on a frequent basis are able to identify changes in fraud patterns sooner. Updating responses to fraud based on the most recent data patterns is critical given the dynamic nature of fraud.
Measure Performance – Understanding the effectiveness of tools and solutions is critical. For example, in measuring the performance of fraud and EPD reduction products, a variety of factors need to be evaluated including, detection rate, false positive rate, and review rate. It’s a mistake to select or measure a solution on only one or two of these criteria.
Fraud Consortium – Share data on common fraud patterns with the industry. Most mortgage fraud is not an isolated incident. Fraud rings operate by re-using information and perfecting the same scheme over and over again until it is detected. By sharing information on bad brokers, appraisers, employment issues and fictitious identities Investment Banking firms can weed out a lot of the most egregious fraud.

Case Study – Smart Sampling Works
We put the best practice of smart sampling and standardized review to the test for an investment banking client recently. The results were so compelling that several investment banks and due diligence companies have adopted the practice on a go forward basis.

What we did was relatively straight forward:

We used a predictive fraud model to select the sample and identified the top 5% of loans which were most likely to contain fraudulent misrepresentations

We then worked side by side with the due diligence underwriters on those files performing a special fraud investigation on those files

The result of this effort was that 50% of the loans we reviewed did in fact contain misrepresentations. The due diligence process was straight-forward and repeatable, and the false positive rate was better than 1:1, meaning that for every two loans reviewed, one contained fraud.


	Frank McKenna, Co-founder and Chief Fraud Strategist of BasePoint Analytics Frank helped develop and introduce advanced predictive technology to detect mortgage fraud. Frank’s vast experience in fraud management has enabled him to identify unique and effective tools to manage lender risk through pattern analysis and evaluating other parties in the transaction, such as mortgage brokers.
	Additional Information Ask Frank Send your questions, comments, and/or ideas for future discussion topics to Frank. The Facts on Fraud Broker Magazine