Auditing a Mortgage Book
Background
An audit team will assess the credit quality of a mortgage portfolio and conformance with the organization’s standards. The audit team needs to provide an opinion about credit risk controls, fraud risk, and the degree of conformity of the mortgages to standards for securitization. A significant portion of the loans was originated not by the organization itself, but by mortgage brokers, who may have different data quality standards. As with other businesses involved in the mortgage industry, obtaining funding through securitization is critical for the continued operation of the business and compliance with the required standards is critical. As well, business management asked for specific recommendations to enhance fraud risk controls as they have found a growing number of fraud cases.
The Challenge
The audit group was working on multiple objectives, including assessing risk management controls, compliance with standards, and assessing the procedures for fraud mitigation. The data analysis component of the project involved a data set contained not only a large number of loans but also a significant number of variables for each mortgage, which made it particularly difficult for the audit team to examine using desktop tools such as spreadsheets.
Our Approach
To answer the multiple questions that the audit team brought up during our exploratory session, we provided analyses to understand the data at different levels of aggregation:
For loans with a particular credit rating, we created loan profiles that included profitability and risk indicators (including accrued interest and interest rates, balance, and loan-to-value ratio), as well as performance indicators (delinquency status, missed payment amounts, and frequency, among others). Based on these profiles, we were able to provide a view of what a typical loan would look like within a specific credit rating, and we were able to produce samples of abnormal loans.
We created profiles by, both, creditor and the property’s address or geolocation, which allowed us to identify people who had acquired several properties and, in some cases, identify if these individuals were operating as a business instead of being a retail customer. As well, we tested for properties that may have been used as collateral for concurrent loans under different creditors’ names.
Another set of analysis focused on the officers adjudicating loans and the brokers who were originating them. These set of analysis helped the audit team to identify inconsistent applications of the credit standards.
The Benefit
The analysis covered a large and complex data set, which would have been practically impossible to analyze using desktop tools such as spreadsheets or small databases within the allotted time for the audit. Although we worked with a single audit team, different auditors focused on separate lines of investigations: credit risk assessment, fraud risk management, and broker/adjudicator operational risks. We provided the team with high-value audit samples, which motivated valuable discussions with the business line staff, and that resulted in insightful recommendations.
Contact us
