RiskExec has recently been updated to include the following enhancements:

Redlining Analysis Module

Analysis Results Gap Metric

Users can now benefit from enhanced Volume, Peer and Market results in the Redlining module, which now includes a Gap metric. This feature indicates the number of additional records needed in the reviewed tracts to match the peer or market percentage for that geography. 

Please Note: While the Gap metric provides valuable insights, it should not be regarded as the single source of truth. Instead, it serves as a helpful reference point for analysis purposes. Users are encouraged to consider additional factors, such as lender and peer performance trends, mortgage market conditions, and product offerings, and exercise critical judgment when interpreting and applying this new Redlining metric.

Filter Analysis Results

Users are now able to filter the analysis results to display only statistically significant results. Additionally, users can suppress rows that yield no results.

Create Respondent Group Peer Sets from Redlining Analysis Results

Users who have conducted a Streamlined or Advanced Redlining Analysis, and opted for RiskExec to create their peer groups, can now easily extract those Respondent Groups Peer Sets to the Peer Analysis module.

Peer Analysis Module

Peer Sets: Add a Comment

RiskExec users are now able to add a comment or description when creating a Peer Set in the Peer Analysis module.

RiskExec Maps

New Layers

Two new mapping Layers have been added to RiskExec Maps:

• Incorporated Places: The United States Census Bureau defines a place as a concentration of population which has a name, is locally recognized, and is not part of any other place. A place typically has a residential nucleus and a closely spaced street pattern, and it frequently includes commercial property and other urban land uses. Incorporated places are defined by the laws of the states in which they are contained.

• Census Designated Places: The United States Census Bureau defines a place as a concentration of population which has a name, is locally recognized, and is not part of any other place. A place typically has a residential nucleus and a closely spaced street pattern, and it frequently includes commercial property and other urban land uses. The Census Bureau delineates CDPs.

New Theme

One new mapping Theme has been added to RiskExec Maps:

• Banking Deserts: A “banking desert” is defined as a census tract without a physical bank branch within a certain geographic radius from its population center, or within the tract itself. A “potential banking desert” is defined as a census tract that will become a banking desert if one bank branch closes within a certain radius from its population center, or within the tract.

Fair Lending Module

Marginal Effect in Logistic/Underwriting Regressions

A Marginal Effect is an advanced statistical measure defined as the partial derivative from a regression equation. It is the instantaneous slope of one of the factors when all the other factors are kept constant. It represents a particular factor's impact on the Regression model outcome. The higher the absolute value of the Marginal Effect, the “more important” that factor is in the regression equation.

Note: To utilize the Marginal Effects metric in analyses run before May 5th, the user must re-run those analyses. 

VIF for Logistic/Underwriting Regressions

A Variance Inflation Factor (VIF) is an advanced statistical measure that reflects the amount of multicollinearity in a regression analysis. A large value for the VIF indicates a factor has a highly collinear relationship with other factors, however, consideration of multicollinearity should be balanced by the requirement that fair lending models accurately reflect lender credit policies. Typically a high VIF on the indicator variable for membership in a prohibited basis group would be a significant concern in a fair lending regression model and a user may consider modifying the model to enhance performance. 

Note: To utilize the VIF in analyses run before May 5th, the user must re-run those analyses.