The key to great analysis: Good data

Over the last several years, many credit unions have made tremendous progress in the way they use analytics to drive lending decisions and pricing.  However, one of the roadblocks we often see with credit unions is that they do not have quality data that is necessary for the analysis process.  The key is to incorporate data management strategies that ensure a credit union has quality data for analysis.

Involve Your Entire Team

Data management and analysis is not contained in a single job function.  As your credit union gets a handle on data management, mining and analysis, it will find that there are opportunities for individuals from varying disciplines to take part.  Often, we want to intuitively go to IT for all data and reporting functions, but IT professionals may know very little about what lenders or marketers want to get from the data and reports.

Credit unions should establish a “Data Task Force,” if you will.  This task force should consist of lending, marketing, sales, IT and operations individuals.  Each individual should come to the team with an idea of what data they are currently collecting, and where that data is being stored.  This will be essential when it comes to creating a common data warehouse.  Most importantly, there should be an owner of this project in the credit union.  The owner should head up the task force and document everyone’s data resources and needs.

Data Diving

The first initiative of the Data Task Force should be to determine what data is being collected today, where the data is being stored, the storage life of the storage facility, and the accessibility of the data.  For example, many lenders that use a Loan Origination System (LOS) collect information in that system that does not get transferred to its core data processor when a loan is funded.  Data stored in the core is often archived and becomes inaccessible after a short period of time.  Even if your credit union has a data warehouse facility today, it may not be easy for department leaders to access data from that warehouse.

In the process of determining what data you have available today, you should also determine the quality and purity of that data.  Bad or incomplete data will not be very useful to  your credit union in the future.

Data Purification

If data purity issues are discovered once a credit union has made an inventory of available data and mapped where that data can be found, then a project should be initiated to purify that data.  It may seem time consuming at first, but once the project has been completed, the results will pay off in a big way later on.

A good example of a data purification project might be with real estate Loan-To-Value (LTV) ratios.  Your credit union may not have originally stored each loan’s LTV in electronic form in your data storage facility. As a result, some loans have LTV’s available, others do not.  A complete analysis is going to require filling in the gaps – even on loans that have already paid or charged-off.  If these values exist in a paper format, it will require staff to manually input these values.

Data Opportunities

Once current data resources have been identified and purified, your team may want to discuss opportunities for capturing new data types that haven’t been collected in the past.  For some credit unions, this may be something as simple as documenting an applicant’s time on the job or at their residence. Further, it may be identifying new risk attributes, such as dealer personnel in an indirect environment, or type of job or business for a person’s employment.  Increasingly, these external factors have been identified as potential risk factors.

There are two issues that arise when considering new data opportunities:  The first being that there is a temptation to collect all possible data; the second is where to put the collected data.  With regards to the first issue, it is recommended that you validate that each data point collected actually has a potential impact.  As to the second issue, it is recommended not to rely on the core data processor as your credit union’s storage facility.  Perhaps a simple data warehouse that consists of simple spreadsheets is a good place to start.  In fact, storing AIRES files is a great first solution.

The fact of the matter is that financial institutions are relying more and more on internal data to drive business decisions.  To maintain a competitive edge, credit unions must employ effective data strategies.