Madison Associates LLC

While interest rates are at historic lows, many lenders are mindful that rates will not stay low forever. Increasingly, they are concerned about their ability to measure the effect of future interest rate increases on coverage ratios. The objective is to be able to quickly and easily get an understanding of potential exposure and risk.

Their first effort is to simply identify when loans will reprice or mature with indications of:

• The loan balance that will reprice,
• The current rate and spread,
• The index on which the loan reprices, and
• The current coverage ratios.

With this information in hand, clients are able to identify situations where the coverage is thin and will be significantly adversely affected by higher rates. They can also identify any concentration of repricing in future time periods — for example, do a lot of loans reprice in three years?

Once the repricing schedule is available, analytics can be applied to it to automate measuring the impact on future coverage ratios.

• The current balance can be amortized down to the repricing date to better estimate the impact of higher rates on coverage.
• Potential exposure from utilization of available lines can be measure.
• Interest rate projections can be established, future loan payments can be recalculated, and stressed coverage ratios can be presented.
• Projected coverage ratios that are below a user specified threshold can be identified for further analysis.

While all the information necessary to perform this analysis is resident in most core systems, the ability to create reports that present the information and perform the analytics is more difficult to achieve. Systems that easily provide this information to clients allow them to better measure this component of credit risk.

Stress testing the commercial loan portfolio has been an important function for many years. Increasingly, we find clients are stress testing more granular subsets of the total portfolio. Here are some of the criteria we find clients using. See if any would make sense for you. The easier it is to select the desired sub group the more likely you are to conduct more granular analysis, and the less staff time is required to get the desired result.

Department and Broad Portfolio Types. These are very broad categories such as commercial real estate (CRE) loans, C&I loans, construction loans, asset based lending, etc. In many cases there is a lot of overlap between departments and portfolio types. Not only would the appropriate stress values differ, but the particular ratios that you many want to test will differ. For Example, you may want to stress LTV and DSCR for CRE loans but DSCR and loan interest rates for C&I.

Geographic Criteria. Different markets and geographic areas may have substantially different economic conditions, some may be strong while others quite weak. Geographic criteria would include region, state, market or city.

Collateral or Loan Type/Loan Category. Needless to say, different collateral and loan types exhibit quite different risk characteristics.

Loan Programs.  Loan programs may be directed at different customer groups and/or have different terms and structures that merit separate analysis.

Origination Period. Loans originated in periods when underwriting criteria were especially aggressive may exhibit heightened risk.

Participation Loans. Purchased participations were originated by another lender, so even though your organization also reviewed the underwriting of the loans, servicing will be by another lender and they will have more direct contact with borrowers and performance information. And Participations sold, while underwritten and serviced by your organization, may be larger implying worse consequences if they go bad but they may also have been underwritten against more rigorous criteria. Either way, the ability to look at participation loans separately from the rest of the portfolio is valuable.

Rate Reset Dates and Indexes. The particular characteristics of floating rate loans give rise to risks associated with higher loan payments if interest rates increase. The ability to easily identify situations where variable rate loans may reprice upward and evaluate the resulting potential DSCR is also valuable.

Stress Testing Loan Covenants

Overview: It is important to be able to stress test many variables, applying various techniques to stress testing the portfolio. Stress testing loan covenants is one such approach that many find useful.

For a long time the Madison System has enabled clients to identify loans that violate loan covenants, or come close to violating the covenants. We have recently added a new feature to one such report – the ability to stress LTV and DSCR covenants by changing the value of collateral or businesses supporting loans to stress LTV, and by changing the income of collateral or businesses to stress DSCR (Debt Service Coverage Ratio).

In this report users may determine the amount of the stress to either collateral value or income, or both, and the Madison System will then recalculate the LTV and DSCR for each loan to display those that violate the covenants which are highlighted in red, or those that come within 15% of violating the covenants which are highlighted in yellow. The report can be set to display only violations, only violations and near violations, or all loans.

Stress Testing Participation Loans

Overview: Participation loans often have characteristics quite different from other loans in the portfolio, and therefor it is useful to be able to stress test those loans separately from others. A recent addition to the Madison Report library makes it easy to stress test participation loans separately from other loans in the portfolio.

There are many factors that make participation loans different from others in the portfolio. They are usually larger loans, participations purchased were originated by another lender and are typically serviced by that lender, and the underwriting standards applied to these loans may be more rigorous than others in the portfolio.

Consequently, when stress testing the portfolio it may be useful to exclude participation loans so that the other loans in the portfolio may be evaluated. Or it may be useful to exclude only participations purchased so that stress tests are applied only to loans that were originated by the client organization.

Finally, while participation loans may have been underwritten with more demanding criteria, they also tend to be larger loans, so that if a problem occurs, the consequences are more pronounced. That being the case, it is useful to be able to select only participation loans, or only participations purchased, or only participations sold and test those loans.

The ability to easily include or exclude participation loans from stress tests provides added value to stress test reporting.