Improving Data Quality: Process Mapping Is a Start
Data quality has become a regulatory flashpoint. Not exactly breaking news. Not exactly news at all for any lender paying attention. Even so, too few financial institutions are auditing the accuracy of the data they will need to estimate their reserves under expected loss.One of the early messages from regulators about CECL complexity could be driving such a laissez-faire attitude: if you like your current methodology, you can keep it.
Recent studies tell us the data management practices of both domestic and international financial institutions are not good, and they’re not getting better. It can be demonstrated through an exercise called “process mapping,” a workflow diagram designed to clarify a process or series of parallel processes. Process mapping done correctly allows an entire organization to see and better understand a particular process, then focus on problem areas and identify improvement opportunities.
Lenders Use Mapping for the Loan Decision Process
Lenders can use mapping to gain insight on their loan decision process. The instructions are simply to follow the customer journey from the request for a loan to the decision to fund it to its maintenance. Getting started includes four steps:
- Identify all departments and individuals involved in the loan decision process.
- Create a simple diagram connecting the departments and individuals – a flow chart.
- List the information or data elements entered and/or edited at each point in the process.
- Detail the procedures for entering, editing and updating data.
The following diagram details the loan mapping process applicable for a $10 billion financial institution.
For each step in the chart, answer key questions about how information is gathered. For example:
How is this data created?
- What person(s)?
- What systems(s)?
- Which process(es)?
While the exercise appears fairly simple and straightforward, you might be surprised by how many entry points there are within the loan decision process.
Revelations in Mapping Loan Decision Process
The final step in mapping a typical loan decision process is assembling the documented procedures addressing entering, editing, altering and managing data. Therein lies the nasty revelation. Few institutions have extensive rules for data management; most guidelines are developed by users. Heuristics, or rules of thumb, are also established by those who populate the fields and complete the forms, and are almost never institutionalized.
But mapping doesn’t just unearth the bad. Strengths are also revealed in thoroughly mapping a process. Still, the concern that weaknesses will be discovered, and require action to resolve, is why some avoid mapping.
The arrival of CECL has turned the spotlight on data quality. The accuracy and consistency of the data that support and document your loan activities has grown exponentially more important because of CECL. Your auditors and regulators are focused on it. So are your customers, and even the media. Mapping your lending processes and tracking the data for each point along the way will reveal your data strengths and weaknesses and provide the insights and direction you need to begin the process of data quality improvement.
Annual RMA/ AFS Survey on the State of Enterprise Data to Support Credit Risk Management, June 2016
The State of Data Quality, Experian, 2013