In one of our last posts, we talked about how marketers can use data. As part of the growing usage of data and analytics, we want to highlight the different ways key executives can leverage data to drive their business. In this post, we discuss five ways that lending analytics can be used in any bank or credit union.
A CLO (Chief Lending Officer) often has a wealth of information at their fingertips. Modern loan origination platforms have mountains of information. Unfortunately, much of the information contained in the LOS systems goes unused by lending teams. With the growth of non-traditional lenders, like LendingClub, credit unions and community banks must become more analytically driven to remain competitive lenders.
Five Ways CLOs Can Use Lending Analytics
Below we identify five ways that CLOs and their lending teams can use data and analytics:
1. Concentration Risk
Would a wise investor put all of his funds in one sector? One company? Of course not. Just as investors must ensure their portfolio has adequate diversity, lenders must ensure their risk is not overly concentrated. Concentration risk might involve stratifying your loan portfolio by geography, product type or credit tier to ensure risk is appropriately spread across a variety of variables.
For example, if your lending portfolio is heavily concentrated in a particular county where business is booming, what happens if business collapses in that county? What might have looked like high credit quality loans could change dramatically and quickly if significant layoffs or business bankruptcy occurs. Similarly, CLOs want to strike a proper balance amongst their credit tiers. Buying A and A+ paper might reduce charge-offs, but this will negatively impact net interest margin by reducing the average yield for your portfolio.
Developing KPIs to constantly monitor concentration risk is something all lenders can and should be doing today. The reporting process should be automated to ensure accuracy and timeliness of information.
2. Segmentation Analysis
Segmentation analysis is another great way for CLOs to get started with lending analytics. When integrated with profitability data, CLOs can identify which customer segments are driving the most revenue and profits.
Segmentation analysis can manifest in a variety of forms: lenders can segment customers by profitability, by geography, origination channel, or combinations of these and other factors. Proper communication with your bank or credit union’s marketing team then enables your organization to develop strategies to effectively target and lend to key segments.
For organizations with a mature data and analytics program, integrating lending data with other data sources can prove invaluable and open up additional segmentation opportunities. Overlaying loan application data with third-party data (like US Census Data) can identify new opportunities to expand and grow your loan portfolio.
3. Loan Performance Analysis
Your lending team probably spends a significant portion of their time generating new loans. But how are you monitoring loans that have already been funded? How are these loans performing over time?
Proper monitoring of a loan’s performance throughout its life is critical to successful data-driven lending. By monitoring loan performance over time, you will start to uncover key insights like how many loans are being charged off, are in default, have been paid off and other outcomes. The results may highlight common characteristics of high-performing loans and of low-performing loans. This information can then be used to manage your underwriting criteria along with your loan growth strategy.
4. Data-Driven Pricing Models
Companies like Deep Future Analytics are starting to bring this idea to many banks and credit unions. By leveraging your historical lending data, statistical models can be applied to optimize pricing while maintaining (or even reducing) your organization’s credit risk. Based on past loan performance, your bank or credit union can identify whether or not you can offer better rates on C or D quality loans while ensuring adequate ALLL. By understanding your past charge-off history and integrating with loan performance, this form of lending analytics can keep your CFO and your CRO (Chief Risk Officer) happy.
We anticipate a sharp increase in the number of community banks and credit unions performing this type of price modeling as data and analytics become more commonly leveraged throughout the industry. Large financial institutions have been doing this type of work for years; as data and analytics technology becomes more practical for smaller organizations, these advanced models should become more frequently adopted.
5. Non-Traditional Risk Metrics
FICO score have been the dominant number used in credit decisions for the majority of financial institutions. But, as 2008 showed, credit scores are not as a flawless as they are often believed to be.
Modern banks, like SOFI, are using non-traditional metrics to fund and refinance the massive amount of student loan debt outstanding. Their lending analytics models rely on factors like where you want to college and your current employment and income status. As more and more millenials “say no to credit cards”, traditional banks and credit unions should consider how to integrate these non-traditional metrics into their underwriting criteria.
While we don’t expect smaller organizations to completely overhaul underwriting criteria using non-traditional risk factors, coupling these risk factors with traditional credit score-based underwriting models could generate additional lending opportunities.
Lending analytics are becoming more common with community banks and credit unions. For banks and credit unions to remain competitive, a greater adoption of these data-driven techniques must become a part of any strategic plan. While we understand that jumping right into, for example, social media mining for credit decisions is a big leap for most organizations, laying a foundation to continuously improve lending analytics is critical, if not necessary, to remain competitive in the next decade.