Many community banks have been in touch with their borrowers to assess their financial condition during this precarious economic environment. Bankers are also segmenting their portfolios based on some of the most affected industries, including restaurants, hotels, retail, manufacturing, and others.
But there are other elements in addition to the industry of the borrower that contribute to an elevated level of credit risk during this COVID-19 crisis, and they require banks to take a deeper look into the details of the loan. As someone who works with bank loan data all day long, I can understand how difficult it can be for banks to gather all the specifics for their loan-level information. But it is essential during the COVID-19 crisis for banks to triage their loan portfolios beyond the industry of the borrower by identifying these critical data fields and quantifying their impact using the appropriate stress testing techniques.
The good news is that banks do not need nearly as much data as they think they do and most of everything they need should already exist in their core processing system. Here are five critical loan-level characteristics that banks must consider in addition to borrower industry in their stress tests and in their day-to-day oversight of the loan portfolio that will allow them to get out in front of problems before it’s too late:
1. Maturity Dates and Interest-Only Expirations
Pay specific attention to those loans of an affected industry that have a maturity date within the next 12 to 24 months, or those structured with a balloon payment at the end. Many businesses might have planned to replace or refinance their loans at the end of the term, but this will be extraordinarily difficult if they are in an environment where banks no longer want to lend. More complications can arise with borrowers if they are having cash flow problems, making it difficult to qualify them for refinancing or new loans if they can’t meet underwriting standards. Many banks may be resistant to making such risky loans, with good reason.
Loans with structures such as expiring interest-only periods that would dramatically increase the borrower’s debt-service requirements would also represent an elevated level of credit risk in this environment. Many borrowers simply will not be able to afford to see their debt-service triple in size while their cash flows are under pressure from the pandemic’s financial fallout.
2. Risk Rating
Banks spend a considerable amount of time assigning risk ratings to each loan. Even if two loans have similar characteristics, with a similar debt service coverage ratio (DSCR), they may still have a different risk rating value for a variety of reasons. One reason for a loan to have a better risk rating can be the borrower’s liquidity. If a borrower has a large amount of savings, those savings can potentially subsidize disruptions in cash flow from the underlying businesses or property associated with the loan. In these cases, the personal guarantee actually has value to the bank. This can potentially make their risk rating one level stronger as they may be better equipped for financial changes. The point is that too many banks are focused on the NOI and DSCR of the borrower. While those metrics are critically important, the risk rating will often be highly correlated to (or even trump) those metrics because of examples such as the one in which the borrower has other sources of liquidity that can be relied upon to ensure servicing of the debt.
3. Loan-to-Value Ratio
The current climate has illuminated a large cash-flow problem within many industries. That being said, a loan with a cash flow problem but also a low loan-to-value (LTV) has a greater margin of safety to the bank than a similar loan with a high LTV. Let’s say there are a group of similar problem loans in an industry that is declining, and the bank needs to sell the collateral. It would be safer to absorb the loss for a loan with a lower loan-to-value ratio, everything else being equal. The property can be liquidated below the previous appraised value, while still covering the principal balance of the loan. The types of collateral is also important. Receivables and inventory can lose their value far more rapidly than real estate if not managed and overseen by the bank very carefully.
4. Debt Service Coverage Ratio
The higher the debt service coverage ratio, the better. A high debt service coverage reflects a stronger ability to service debt obligations. When income is halted or revenues decline, those with a higher debt service coverage ratio will most likely have more flexibility and a higher cushion to withstand the downturn. Still, don’t put too much emphasis on DSCR for two reasons: First, DSCR is more applicable and reasonable when there are about 10% to 15% shocks to revenue for a borrower, but if we are talking 60% to 80% shocks, it won’t be as meaningful. Second, it is important to consider the different types of industries with the same DSCR, as they have different cost structures. For instance, compare two types of real estate: a strip mall and a multifamily property. If the anchor tenant in the strip mall decides to move, the rest of the businesses will be negatively affected. However, with multifamily, the revenues are more spread out and diversified across occupants. From a risk perspective, a multifamily loan with a 1.25 DSCR is much better than a strip mall loan with a 1.75 DSCR. Therefore, don’t prioritize the DSCR, but it can be a useful indicator and help you stratify risk if your bank has a large concentration in one or two risk ratings within the loan classification system.
5. Vintage
Lastly, consider the origination date of the loan, especially in real estate loans. If there are two similar loans, but one originated in 2014 and one was made in 2019, the older may have a greater margin of safety because the appraised value of the collateral was based upon 2014 standards. Since then, property values have appreciated, but the bank may not have needed to order a new appraisal since the loan was originated. This means that the LTV in your loan tape is actually overstated for this loan because you are not recognizing the appreciation in the value of the collateral. This can be estimated though using proper techniques.
The current loan-to-value ratio for the first loan is significantly stronger than the second loan, which means these loans have dramatically different risk profiles despite their similarities at the surface. Taking into account the current situation, the first loan will have a higher cushion and margin of safety over the second loan. It is important to consider the origination date, as well as the last appraisal date, which in many cases is the origination date. Vintage is far more important for real estate collateral than other types of collateral such as receivables and inventory.
By taking these five characteristics into account, banks will have a better idea of which specific loans to look out for in their portfolios. Even if all these characteristics are in the core of the loan data, it is also important to analyze the intricate relationships among all the loans your bank is holding. These pieces can be brought together in one holistic analysis, such as a stress test, which can bring efficiency and a method to triage the loans for problems. Pandemic stress testing is critical to helping community banks CEOs navigate safely through this crisis. Segmenting by industry is a good start, but you need to go further to identify those credits that are most vulnerable to the pandemic.
--Radhika Gupta is an Invictus Group Bank Data Analyst
