Using Modeling Tools

The first step in protecting the credit union is to select an appropriate risk model. Management should strive for model complexity that is equal to the institution's balance sheet complexity. For credit unions with a significant residential mortgage portfolio, it's especially important to consider adopting a stochastic model to calculate optionality. Residential mortgage borrowers possess a valuable call option on rates, which isn't as prevalent in other loan products. Given the market rates, the value of a mortgage option may not be significant as the refinance incentive is low for most newly originated mortgages. This may deter some institutions from taking steps to calculate a more precise option value; however, capturing optionality is a critical component in the design of hedging strategies.

Beyond offering the ability to perform intricate calculations to ensure more precise analytics, an advanced model can provide credit unions with significant resources to aid in strategic planning. In addition to interest rate risk, they face credit and liquidity risk, as well as a multitude of esoteric factors. An advanced model will project credit losses for the life of a loan to estimate the credit risk inherent in a portfolio. These assumptions will also aid in projecting an institution's liquidity for a more holistic view of the major risks facing any depository.

Sophisticated modeling software can also be used to estimate probability of rate movements, allowing balance sheet strategies to be adaptive toward the market environment. Stochastic modeling uses Monte Carlo simulations (originally developed by gamblers) to project the path of interest rates through time. While these simulations are necessary to value optionality in a portfolio, Monte Carlo simulations can be expanded to provide an estimate of the progression of interest rates. With a well-calibrated model, the Monte Carlo simulations will project the path of interest rates over multiple iterations through time. This enables a credit union to discern the assumed probability of a specific index increasing or decreasing a particular amount at a given time.

The reliability of the Monte Carlo simulations in such cases is driven by the underlying assumption of the model. For example, assumptions ranging from the number of iterations each simulation performs to the model's volatility surface can influence the results of the simulation. Since assumptions can have such an impact on results, it is critical to frequently validate the model being used. If the underlying assumptions are deemed reasonable, an institution can use the results of the probability estimation to inform balance sheet strategy to mitigate interest rate risk.

Reporting Interest Rate Risk

Many metrics are used to report interest rate risk, but one that is most frequently referenced is effective duration. Effective duration is the price sensitivity of a security, given a change in underlying interest rates. However, the calculation for effective duration, while commonly referenced, has some flaws. The calculation itself is precise when the observed rate changes have only small variances. As market rates become more volatile, duration analytics lose precision.

Another shortcoming of effective duration is the assumption that market rates move in parallel fashion, which can be a significant shortcoming when trying to develop a hedging strategy. Since rates rarely, if ever, move in a parallel fashion and long-term assets can be sensitive to multiple points along a yield curve, key rate duration analytics are more appropriate to understand the true nature of a portfolio's sensitivity.

Key rate duration reports a security's sensitivity along the entire yield curve by isolating each point along the curve and valuing the security as a single point is adjusted up and down. This results in a significant amount of processing time, emphasizing the need for a sophisticated model. With the resulting key rate durations, a credit union can more efficiently hedge the balance sheet. Key rate duration analytics guide the institution to the appropriate notional allocation for interest rate swaps. By doing so, the balance sheet is better protected against parallel interest rate moves and changes in the slope of the yield curve.

Many factors can affect a credit union's capital position, and interest rate risk is only a single factor. However, it's one that may be mitigated easier than in the past and shouldn't be overlooked. Instruments like interest rate derivatives are more prevalent and accessible today than in previous years. Such products are available as tools to manage an institution's interest rate risk. The first step to hedging risk is to understand the credit union's risk position. Understanding and trusting the accuracy of risk reporting will not occur unless a well-rounded ALM process is developed with a model that matches the institution's balance sheet complexity.

Thomas Griswold is Director, Strategic Solutions Group for ALM First Financial Advisors. He can be reached at 214-451-3491 or [email protected].