Too often, assumptions are made about credit quality, loan-to-value ratios, loan terms, and loan types based upon anecdotal data. As a result, credit unions can miss out on loan opportunities or suffer high losses due to reliance on accepted risk assumptions rather than on actual performance data.
Point-of-purchase lending is one prime example. Often, credit unions choose not to participate in point-of-purchase lending because they have heard about credit unions experiencing trouble with loans originated at the dealership. But what about the many credit unions that have been participating for years in this type of lending and have healthy, profitable lending portfolios as a result? The fact is that point-of-purchase lending does have risks, but managing those risks enables credit unions of all sizes to participate in long-term, profitable programs.
Another example is in real estate lending. Who would have thought ten years ago that credit unions would have taken huge losses in their real estate loan portfolios? Yet, that is exactly what credit unions have experienced over the last few years.
If a credit union limits its portfolio to just low risk loans, it would be impossible to generate returns that would enable it to maintain profitability. So, what steps can a credit union take to manage risks associated with lending, provide more loans to its members, and maintain healthy and stable returns?
Consider the information that is collected in the loan decision process. Typically, application data includes factors such as time on the job, time at residence, income, and more. Collateral valuation information is also collected, and in the point-of-purchase environment, certain information is collected about the originating third party.
An important but sometimes overlooked factor is what happens with this information after the loan is made. If the information was central to making the loan, why wouldn’t it then be important to use in analyzing the loan’s performance as it matures?
Here are a few key data points that should be collected and stored in the credit union host system. Information that will be useful for later analysis includes:
All credit scores used to make the loan decision or price the loan
The cost to originate the loan if those costs are specific to that loan type (i.e. dealer origination fees)
Loan-to-value ratio for all collateralized loans
Dealer information for auto loans whether they are indirect loans or not
Dealer finance director information, if available, for auto loans
Vehicle manufacturer information
Vehicle warranty and payment protection product information, whether the credit union sold the warranty or not
Values of real estate property
The total value of all liens against real estate property
Credit unions should look to employ technology that enables them to do ongoing, regular analysis of their loan portfolio. Typically, on-board host system reports are not sufficient for providing the necessary consistent analysis. Also, host systems’ data architecture does not allow for tracking historic trends.
Although some credit unions rely on spreadsheets for tracking, they can become very difficult and tedious to manage after several months. The marketplace offers a variety of tools for credit unions that are designed to track historic trends, including specialized software like Lending Insights’ LPMS. But in any scenario, the analysis must be robust and regular, employing Multi-Dimensional Portfolio Analysis (MDPA).
MDPA allows the credit union to slice and dice the loan portfolio by multiple dimensions. For example, users can isolate real estate loans and evaluate losses by year of origination and credit tier at origination, or assess loss severity by vehicle brand and loan underwriter. All of the information collected at origination can assist in evaluating loan portfolio performance from a seemingly infinite number of perspectives. That data contributes to a more refined lending strategy that not only reacts to conditions but offers the prospect of predictability.
Here are a few standard analysis reports credit unions should include in their comprehensive, ongoing portfolio analysis.
Monitoring Risk Concentrations
Credit unions should establish risk thresholds for certain loan products and monitor portfolio growth to ensure that these thresholds are not exceeded. Additionally, when entering new loan markets it important to start small and use performance analysis to determine how fast the credit union should increase concentration thresholds.
Risk Based Pricing and Interest Rate Risk
Use credit union pricing tiers, changes in portfolio yields, and loss data to determine if the credit union is charging the right rate for each risk tier.
Static Pool Analysis
Pool loans together by origination characteristics (most often by date) to view how loans originated under the same circumstances perform over time. This will enable the credit union to detect lending strategies that have a negative impact on loan performance.
Credit Score Migration
Often it is helpful to know if the credit scores of members with unsecured lines of credit are going down. But credit score migration analysis can also help predict future loan performance by tracking credit score volatility after origination, and may actually provide the credit union the opportunity to curtail certain lending activities as downward trends are indicated in the population.
Reacting To Analysis
It’s important that credit unions not only gather and analyze the above data, but react to the trends exposed by the analysis. Remember, numbers don’t lie and even if the analysis disproves assumptions, it’s important to not ignore the facts. FICO reports that if lenders had reacted to the downward trend in credit scores twelve months before the real estate bubble burst, fewer real estate loans would have been originated, saving lenders millions of dollars in losses. But, who would have believed those numbers? Experience led many institutions to believe that real estate values never fall. The facts proved otherwise.
Credit unions are encouraged to react cautiously without overreacting. In some lending environments, volatility in lending strategies can cause disruptions in loan growth, which can also have a negative impact on lending.
Like anything else, portfolio analysis will take practice and can be difficult to get started. However, properly analyzing loan performance is worth the time and money spent to start the process.
This article originally appeared in the Spring/Summer issue of CU Direct’s CU Lending magazine.
Bill Meyer is the Public Relations and Corporate Communications Lead for CU Direct Corporation and the company’s CUDL, Lending Insights, Lending 360, CUDL Retail, and Vero brands. He can be reached at firstname.lastname@example.org.