Stress testing programs – Multifamily
Stress testing programs should consider their tests in a one-year time horizon measured in both a mild and severe scenario. Mild scenarios should test items to where they honestly think they will be in one year. Severe scenarios should try and replicate conditions that represent the worst possible outcome for the portfolio but are still logically possible (do not blindly make up severe scenarios).
Without going into every possible stress testing scenario, here is a stress test example for a multifamily portfolio. This example’s logic can be transferred to other concentrations:
Multifamily
These tests should measure the impact on DSC, which is their ability to repay the loan, during a number of economic occurrences. The only source of revenue for multifamily projects is tenant revenue and the health of the projects is based on changes in rental payments and vacancy rates. Using changes in Net Operating Income (NOI) to reflect vacancies, rental concessions or a reduction of average lease payments for units, tests should be designed to reflect expected changes over the next one year. If the majority of a portfolio contains certain Class B or even C properties, it might be expected that rents will need to be reduced to compete with other projects in the area. Reducing NOI by 10%, for example, might reflect that multifamily projects would have to reduce rents by 10% to remain competitive and in equilibrium with their market. Reducing EBITDA at the borrower level would measure the impact of other borrower income on the DSC.
A severe analysis of a multifamily might try and replicate things like previous recessionary experiences, changes to the Section 8 program (either overall or at specific projects), the rapid increase of superior new projects in the market or even scenarios if all multifamily projects took a one notch reduction (Class A to Class B).
Using vacancy rate as a proxy for changes in NOI, mild tests could be developed to reflect the reduction of renters if home ownership increases or movement of renters into new competing projects. The mild test could be designed to select a vacancy rate that would move the portfolio average DSC to 1.2. A severe example would use a vacancy rate that would move the DSC to 1 or even below.
