Washington State Department of Financial Institutions

Webinar Frequently Asked Questions

1.  What mortgage variable interest rate indices are commonly used?  What should management consider when choosing an index?

The Prime rate is the most commonly used index.  The reason Prime rate is most often used is that it is easily recognized and tracked by member borrowers.  Other commonly used indices are the US Treasury yield curve, the 11th District Cost of Funds and LIBOR.  The 11th District Cost of Funds (COFI) has been used in the past, but due to its regional nature and lagging status it has fallen out of favor.

Credit unions should take into consideration how the other side of the balance sheet is pricing and try to match interest rate indices as well as possible.

A question not asked, but which is important – is how often should the variable loan rate be adjusted?  A quarterly adjustment will probably result in the least amount of interest rate risk, because deposit rates often change quarterly.  There are marketing considerations as well, because members will compare the frequency of the variable rate adjustment.

2.  What level of interest rate shocks are examiners looking for?

Credit unions should consider several interest rate shock scenarios.  The standard industry-wide interest rate shock is a 300 basis point plus/minus shock.  This gives examiners a comparison of relative risk against industry standards and others in the industry.  Because market interest rates are at historically low levels, the Division of Credit Unions recommends that higher interest rate shocks should be performed to simulate rising interest rates.  The maximum positive rate shocks should be at levels that might be achieved in the future.  Short term interest rates (Fed Funds) have been between four and nine percent more than 80 percent of the time, since the early 1970’s.  Currently, the Fed Funds rate is at 1.75%.  Regulators would like to see higher interest rate shocks than 300 basis points, at least in ramping models that go out at a minimum of two years.

3.  What are examiners asking management to do to validate the  ALM model?

Validation is a useful training tool in setting ALM model assumptions.  Because the accuracy of the ALM output results are mainly dependent on the assumptions entered into the model, validation is a critical part of the ALM process.  It is important that these assumptions closely approximate the characteristics of your credit union.

There are a variety of ways to validate your credit union’s ALM model.  One method is to perform back-testing or post period comparison of the ALM model results to your credit union’s actual financial performance.  Another method is to have an independent third party model validate your credit union’s ALM model.  This is done by analyzing and reviewing your credit union’s ALM model assumptions and inputs for appropriateness, inconsistencies and conformance with past performance.  The independent third party will probably use a different ALM model in its validation process.

4.  What is the "17-4 Test” that I hear examiners talk about?

The “17-4” test is a simple model examiners use to assist in setting their examination scope for the ALM review.  It attempts to measure the projected change in net worth after a +300 basis point instantaneous and parallel interest rate shock.  The change in net worth is determined by calculating the approximate price loss on mortgages due to the rate shock.  Securities subject to FAS 115 are also included in the analysis.  The name comes from the following two assumptions:

  1. 30 year, fixed rate loans that have not amortized yet will immediately devalue 17% with a 300 bp rise in interest rates; and
  2. ARMs (with no amortization and no first rate reset) will immediately devalue 4% with a 300 bp rise in interest rates.

Note: Hybrid loans (3-1, 5-1, 7-1 and 10-1 ARMS) with five or more years to the first reset are considered to be fixed under this calculation.

FAS 115 investments are devalued using the following devaluation percentages:

Held-to-Maturity & Available-for-Sale Investments:

< 1 Yr. 3%
1 – 3 Yrs. 6%
3 – 10 Yrs. 12%
> 10 Yrs. 20%

Trading Investments:

All 6%

The “17-4” test does not consider the effect liabilities will have on your credit union’s interest rate risk.  As previously stated examiners use the 17-4 test to help set their ALM exam scope.  Because the 17-4 test is a very simple model, it is unacceptable to use only the “17-4” test in measuring your credit union’s interest rate risk.  For more information—consult the NCUA Excel sheet on the 17 – 4 Gross test:

http://www.ncua.gov/ALManagementInvest/AssetValuationWorkbook.xls.

5.  The examiner recommended during the last examination that board members receive ALM training.  Does this webcast count towards complying towards this?

Yes it does; however, this is not complete.  We expect that ALM training will provide directors with the knowledge to ask questions, challenge assumptions, and be generally knowledgeable about ALM concepts and assumptions.  The training we expect for ALCO members is more extensive than the training for Board Directors.

6.  What ALM models are being used?

There are many different ALM models to choose from.  It is important that your credit union closely evaluate the different ALM models and choose a model that will best meet your credit union’s needs.  Cost should not be the main determinant when choosing a model.  While we do not endorse any model, the following is a sampling of ALM models that Division examiners have seen:

We have also seen several “home-made” ALM models that are effective.  As you heard in the webinar, Gerald Hickman, President of Quimper Community Federal Credit Union, designed one that is very effective for smaller credit unions of similar asset size and balance sheet complexity.

The Division of Credit Unions’ website links to the Paragon Consulting Group income simulation model which your credit union may find useful.

7.  ALM model assumptions preloaded by 3rd parties tend to be wrong.  How does the Division of Credit Unions view this?

Pre-loaded assumptions do not take into consideration the characteristics that are inherent in your credit union; such as how dedicated your member’s are to depositing money at your credit union.  Also, it is important that the ALM model assumptions correctly reflect the “optionality” in your credit union’s balance sheet.”

Regulatory agencies believe it is the credit union’s responsibility to set the ALM model assumptions.  If running the model and inputting the data and assumptions is performed by an independent third party, then it is management’s responsibility to make sure the ALM consultant receives the data and assumptions, which reflects your credit union’s activities, strategies and inherent characteristics.

8.  What are “core deposits”?

Core deposits are defined as those deposits that are not interest rate sensitive.  In effect, this means that share accounts will stay at your credit union regardless of how dividend rates move.

When modeling interest rate risk, it is extremely important that core deposits be modeled properly.  How core deposit rates are modeled (either interest rate sensitive or not) is the most critical ALM modeling assumption, which has the greatest impact on the ALM output reports.  Overstating the amount of core deposits will result in understating your credit union’s interest rate risk.

 9.  Do you take into account the equity markets when performing ALM?

No, generally interest rate risk is based on the bond market, and equity markets are too hard to predict.  The equity markets do impact ALM, because changing equity markets will affect liquidity by impacting how willing members are to deposit money into your credit union rather than investing in the stock market.  Additionally, the equity markets impact housing prices.  If investing in housing is generally viewed as a better investment then investing in the stock market then this will drive-up the price of houses.

10.  Is credit union borrowings frowned upon?

Borrowing that is planned for and is done for strategic purposes is not frowned upon.  However, borrowing that occurs with no prior planning will be scrutinized and criticized.  Borrowing is often a less expensive means to acquire funds when compared to offering higher CD rates to attract deposits.  Additionally, borrowers can be repaid immediately, whereas, member CD deposits will remain on your credit union’s books until they mature.

11.  Is NII better than NEV?

No, neither is better than the other.  They are just different ways of measuring interest rate risk.

NII measures the changes in expected net interest income given changes in interest rates.  NII is measured over a period of time (12 months or more), and includes the credit union’s own growth projections in assets and liabilities.  These growth projections can be as detailed as your credit union’s ALM model allows.  These growth projections must have a direct relationship to your credit union’s organizational goals (i.e. loan and share growth).

NEV measures the market value of your credit union’s balance sheet based on today’s book of business.  Technically, NEV is a calculation of the discounted present value of all your balance sheet’s future cash flows.  It does not address any balance sheet growth assumptions, nor does it consider the expense structure of your credit union and how this impacts your credit union’s ability to generate future net income.  These are inherent flaws in NEV analysis.  While NEV does not consider your credit union’s expense structure and its ability to generate net income, it can be extremely helpful in determining the magnitude of risk, and the direction of risk in your balance sheet when compared period to period.

From an conceptual and ease of understanding perspective, we often find that credit union representatives find that NII is easier to understand than NEV.  This is probably due to NII focusing on the income statement and its format is more readily recognized and understood.

12.  Can I use an Excel spreadsheet, instead of an ALM model, for my (small credit union?

Yes; however, if you are considering developing your own interest rate risk management spreadsheet, it will require an advanced knowledge of ALM concepts.  We recommend that you ask other credit unions what they use.

13.  Do you recommend attending ALM vendor workshops?

Yes, additional training is useful.  Also, it is important to understand ALM vendor model offerings so your credit union can choose the ALM model that will best suit your credit union’s needs.  Overall, improved understanding of interest rate risk is perhaps the single most important factor that could improve credit union measurement and management of interest rate risk.