Magazine article American Banker

Simulation Approach Gaining Favor in Assessment of Interest Rate Risk

Magazine article American Banker

Simulation Approach Gaining Favor in Assessment of Interest Rate Risk

Article excerpt

Asset/liability management can be defined as the discipline of measuring and managing interest rate risk, which is the exposure of net income or capital to changes in market rates of interest. At its simplest level, it arises when a financial institution has earning assets and funding liabilities which are repriced with different frequencies, resulting in expansion or contraction of the rate spread relationship between the two as interest rates change.

While the problem of interest rate risk is deceptively simple in the abstract, the day-to-day management of this problem in practice is very complex. Factors such as uncertainty over the outlook for interest rates, volatility in spread relationships between various market rates, and the dynamics of the yield curve combine to make the rate risk problem a multidimensional one which is not effectively managed using conventional techniques. This article will explore a specific example of the simulation approach, which has received increasing use by practitioners in recent years, particularly with the advent of microcomputer technology.

The simulation approach consist of the use of computerized financial models to simulate the financial performance of a bank or savings institution under alternative combinations of interest rate environment and balance sheet strategy. Through active testing of alternatives, this approach attempts to quantify the exposure of the institution's earnings to changes in interest rates and the impace of alternative strategies on that exposure.

A very effective application of this approach consists of constructing three or more specific interest rate scenarios and three or more specific balance sheet strategies. The outcome of combining the alternative stragegies and scenarios can then be summarized in a "decision matrix," which shows the net effect on net interest income for each of the strategy/scenario combinations.

For example, the following decision matrix will show that the institution depicted has some current exposure to declining rates and will benefit from rising rates. Given this example, if management thought rates would rise, they might be inclined to extend some of the maturities of their market liabilities.

However, as shown on the decision matrix (chart 1), this action will have a much more significant impact on the institution's exposure to declining rates than it will have on their benefit from rising rates. While they may still choose the strategy of extending liabilities, they will do it with a greater awareness of the risk associated with that action.

As illustrated in that simple example, the simulation approach can be powerful in that it clearly quantifies the risk/return tradeoff of the strategy being considered. The most important point drawn out by such an analysis is that the risk and return of specific balance sheet strategies are very often not symmetrical. That is, the disadvantage under one rate outcome is not exactly equal in magnitude to the benefit under the opposite rate outcome.

The cause of this asymmetry usually relates to changing spread relationships, shifts in the yield curve, and the impact of rates on balance sheet categories such as core deposits. A good simulation model should capture all of these factors and therefore provide a more precise measure of the rate risk of the institution and a better framework for decision-making.

The results of an active simulation process can be communicated in many different ways, including effective utilization of graphics such as the graph of the risk profile (chart 2) of the "extend liabilities" strategy used in the previous example. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.