Regulators are forcing a flawed measurement of interest rate risk on financial institutions - and overlooking a much simpler solution.
Responding to section 305 of the Federal Deposit Insurance Corp. Improvement Act of 1991, regulators have set a timetable for the implementation of rate-risk measurements.
New reporting rules will go into effect in March 1996. Regulators will then issue rules establishing an explicit capital adjustment for rate risk.
Traditionally, regulators analyzed the repricing of assets and liabilities to get a sense of how rate changes might provoke short-term changes in net income. Now, they are examining the potential impact of rate changes on equity capital, or "economic value."
The regulatory measurement of economic value hinges on the present value approach of market value accounting. This technique adjusts all interest- bearing assets and liabilities to current value by discounting their expected future cash flows using current interest rates.
The difference between the present values of assets and liabilities becomes an adjustment to equity capital, with the result representing the market value of equity capital.
In a stable rate environment, the market value approach results in modest adjustments to equity capital. However, if rates spike upward - and they will in time - resulting capital adjustments may be alarming. They also may be misleading and irrelevant. This stems from two flaws in the methodology:
The horizon problem. This results from differences in the maturities of interest-bearing assets and liabilities.
While most loans and investments have definite yields and maturities, some important categories of deposits do not. This severely complicates the process of measuring the impact of rate fluctuations.
When rates fluctuate, changes in asset values can be determined fairly readily. Not so for liabilities such as demand deposits, which have no stated maturities and whose values can only be guessed.
This maturity imbalance causes exaggerated adjustments to capital following major rate shocks.
The matching problem. This results from differences in the amounts of interest-bearing assets and liabilities on the books. It principally stems from the omission of demand deposits - a funding source without interest cost - from the calculation.
When rates rise or fall, more assets reprice than liabilities, further exaggerating the effect on capital.
A better approach exists for measuring economic value.
Remember, the purpose of such measurements is determining the long-term effect on net interest income from changes in interest rates.
With this clue, the solution should be obvious:
Simply, assume a bank will reinvest and renew contracts in such a way as to maintain its current asset-liability posture, then project net interest interest income over a period of sufficient length to eliminate the horizon problem.
Specifically, cash flows are projected into the future, and, when applicable, adjusted for changes in interest rates.
Net interest income is calculated for each period. Then it is discounted, using present value analysis.
Various scenarios for potential changes in economic value can then be developed by repeating the exercise and assuming rate shocks of varying magnitudes.
This method is superior theoretically. And, most importantly, it is a more practical approach.
First, this method is based on projections that are clear to the managers of financial institutions. Second, it measures economic value in terms of discounted net interest income, not by the discounted value of financial instruments.
Third, it provides consistent results unaffected by horizon and matching problems. Fourth, it is independent of the murky issues of devising theoretical maturities for nonmaturity deposits.
Absent the alternative approach, bankers and their regulators will be wedded to a misguided process.
Forcing the situation, regulators will include demand deposits in rate- sensitivity models despite the fact that these liabilities carry no interest rate. And they will use assumptions to manufacture maturities of demand and other nonmaturity deposits.
To be sure, these fixes lessen the horizon problem by lengthening the estimated maturities of interest-bearing liabilities. And they reduce the matching problem by narrowing the differences between amounts of interest- bearing assets and liabilities.
But the adjustments are more fudge than fact.
Worse, regulators will ask financial institutions to take the responsibility for assigning maturities to nonmaturity deposits, and to document justifications.
This directive will provoke an unusable response: Financial institutions will work backward. First they will determine which maturity structure mininimizes the appearance of interest rate risk. Then they will work to justify the model.
In practice, assumptions will be selected that minimize the horizon problem. The bias of the matching problem, an inherent limitation of the supervisory model, almost always will remain.
Banking regulators should shelve their methodology for interest rate risk, focus on a simpler approach that is understandable to the managers of financial institutions, then restrict the regulatory burden only to those institutions at risk.
Most importantly, banking regulators need to understand the flaws of their supervisory model before using it to establish a capital charge for interest rate risk.
Mr. Rieke is first vice president of Principal Financial Securities Inc., Dallas.