Back-testing Interest Rate Risk Models in a Static Rate Environment

6/23/11

By Charles J. Garrison, CPA

A typical reaction to the current stable rate environment is that back-testing of the Interest Rate Risk Model (IRRM) would be a fruitless exercise. In addition to changing interest rates, however, there are many other variables in an IRRM including beta factors, decay rates and behavior patterns of customers and bank management. Now is the time to isolate and adjust these other variables through the use of back-testing to refine the assumptions so that when rates finally move, the only remaining variable is the interest rate itself.

The following assumptions should be reviewed as part of a back-testing process:

• Driver rate relationships – This measurement (commonly known as the beta factor) is the correlation in rate between a shift in the base rate (generally the Federal Funds rate or Prime rate) and the various interest bearing assets and liabilities. The correlation is indicative of the change in rate of a given financial instrument as compared to a change in the base rate. For instance, if management has used a 1.6:1 ratio for loans, and a .4:1 ratio for savings deposits, this indicates that loan rates will change by 160 basis points (bp) and deposits will change by 40 bp given a 100 bp shift in the base rate. A time lag indicates the correlation of timing between a change in the base rate and a change in the rate of a particular financial instrument. These relationships should be reviewed in light of management’s past behavior and expected behavior during future rate changes. Beta factors or time lags may have changed from historical expectations as many banks will re-price assets faster than liabilities when rates rise to improve the net interest margin.
• Non-maturity deposit rate sensitivity – This assumption, also known as a decay rate, is critical as it is indicative of customers' reaction to a change in the base rate (by a decision to remain in the same deposit instrument or the same financial institution). Positive decay rates indicate a migration from a deposit type, while negative decay rates indicate an increase in a deposit instrument. The decay projection is an indicator of the volatility of instruments given a change in the base rate. For instance, in a falling rate environment, customers may elect to move out of non-maturity deposits, and into long-term fixed rate deposits until such times as rates have stabilized. As rates reach a trough in customers’ perspective, they are likely to keep funds in a non-maturity instrument until such times as rates in long-term instruments are more attractive. Customers’ behavior pattern and the corresponding decay rates should be reviewed and tested to control for normal deposit migration during a period of static rates.