Don’t Short-Change Yourself on Core Deposit IRR Hedging!

William J. McGuire, McGuire Performance Solutions
April 11, 2005

The March 31, 2005 quarter end is the first time in many years that both short-term and long-term interest rates are up sharply. Because of the rate moves, look for net interest income (NII) and net economic value (NEV) interest-rate risk (IRR) profiles to be adversely affected. This will be especially the case for regulatory hot-button NEV IIR, where higher base-case discount rates will erode value and slower prepayments will lengthen assets, magnifying losses in rising rate scenarios.

Ad hoc core deposit repricing and average life inputs, which typically understate the IRR hedging potential of member-provided core deposits, will magnify these tendencies. This will lead some credit unions to make bad balance-sheet management decisions based on their “excessive” IRR, when in fact they do not have any such risk. Loss of income, turning away members seeking fixed rate mortgages and other long-term loans, and general loss of member benefits are the unfortunate direct off-shoots of these poor decisions.

How do you avoid short-changing yourself like this without inadvertently adding risk? It is simple: Know current market conditions and your members, understand the basics of core-deposit IRR hedging, and quantify all core-deposit behaviors. This is really just fundamentals and common sense, but with a substantial payoff.

Current Core-Deposit Market Conditions

As of the end of March 2005, Fed moves have pushed short-term rates up by 175 bp since June of 2004. Longer-term rates have followed recently; they are up 75-125 bp since late 2004. Core-deposit rates paid have increased by much less. For example, national average repricing has been just 2 bp for share drafts, 3 bp for shares, and 18

---

1 Core deposits are defined here as share drafts, shares, and traditional MMDA. Note that high rate and indexed MMDA categories are not included. Nor are share draft or share categories that have similar high-yield specifications. Core deposits for this article are balances that do not have a primarily investment orientation. Note that is also excludes balances in any category that might be just “parking” until stock markets go back up, yields on bond funds return, etc.

2 Deposit rates are provided by Datatrac, Inc. of Milwaukee, WI, a nationally recognized expert in the field of deposit and loan pricing information.

---

bp for traditional MMDA since mid-2004. So far, balances supplied or underlying retention behaviors are also holding steady.

Information on the supply stability of core deposits is largely anecdotal. But favorable behavior is definitely being seen in the client data provided to MPS for use in its core-deposit analyses and updates. With the exception of investment type MMDAs, where large inflows of “surge balances” from 2001 and 2002 are now flowing out again, core-deposit balances have largely held steady since June of 2004. There may be some weakening in growth, but even that is not clear in the many MPS credit union studies conducted or updated since mid-2004.

Insights into the retention stability of existing core-deposit balances is also either anecdotal or available only for MPS core-deposit analysis clients. All signs, however, point to a remarkable stability of core-deposit retained balances, especially in checking and traditional savings oriented balances. This implies stable average lives, with longevities that clearly depict a large degree of member indifference to changing financial incentives.

What is perfectly clear is that current credit union repricing is generally well below historic speeds. Table 1 presents a comparison of actual versus forecasted credit union repricing since June of 2004. Actual rates paid are the national average for credit unions. Forecasted rates paid data are produced from a special MPS statistical analysis of credit union pricing covering the period 01/97-06/04. The analysis quantified typical repricing by category over that time. The forecasts are projections of national credit union pricing based on the estimated pricing equations;the forecasts show rates paid if credit unions were repricing today consistent with their average historical tendencies.

As is clear in the data, current core-deposit rates paid are well below what they would be had credit unions been setting rates paid upward in accordance with historic average behaviors. The fact that current overall market rates paid are also lagging (bank/thrift repricing is less than prior behavior) sets an umbrella in place under which individual credit unions can lag their rates paid also.

---

1 The MPS category level equations are components of a simultaneous equations model of credit union pricing for core deposits and CDs. Similar forecasts can be produced at the credit union or market level, or for individual competitors, providing that sufficient weekly time series data (e.g. from Datatrac) are available.

---

Interest rates are up sharply; credit union rates paid are lagging even their historically slow pace of repricing, while balances remain steady. How can all this be happening at once? Aren’t members supposed to react to adverse changes in financial rewards by pulling their core deposit funds out of credit unions? To those in the know, current behaviors are no surprise. The reasons are as follows.

What’s Driving Member Core-Deposit Behaviors?

A key feature of core deposits is that members have more than just a financial motive to provide funds to your credit union. The non-rate related value attached to checking in share drafts, or the certainty of yield and guaranteed value of shares, or the convenience of branches, direct deposit, automatic bill paying, and other services, create powerful nonrate paid-related incentives to maintain balances. It is these non-financial motivators, often lumped together as the “ambiguous influence” on member core-deposit behavior, that are key to explaining the supply stability and long life of core deposits despite their typically unattractive levels of financial compensation.

The ambiguous influence is what creates IRR hedging in core deposits. Members are willing to give up immediate financial advantage (in higher rates paid) in exchange for service, convenience, and relationship related value. This is true even when financial advantages are eroding because of rising interest rates. In fact, history offers clear instances of how member core deposits behave as interest rates rise. In 1994, when the Fed moved short-term rates over 200 bp in about a one-year period, most credit unions repriced core deposits by far less than the interest rate increases. Core deposit balances held firm, however, especially in share drafts and shares (growth slowed, of course). A similar story is told between mid-1998 and the end of 2000, when short-term rates again took off, the yield curve inverted, and stock markets boomed. Core-deposit growth certainly slowed to almost nothing, but not many credit unions lost appreciable funds from traditional core deposits.

Will history repeat itself in this period of rising interest rates? The betting should be that it will, if conditions are considered. First, almost all financial institutions are suffering squeezed margins. The only way to rebuild these is to control interest expense, i.e. maintain repricing discipline. This has kept a lid on rates paid overall and reduced the number of “pricing outliers” who might upset the status quo. Second, current very low rates paid make any increase – even if it is very large in percentage terms—sseem so small as not worth the trouble. With inflation low, the financial drive of higher rates paid is just not strong enough given the minimal dollar gains to members.

Third, it is also possible that there has been a shift in member liquidity preference. Statistically significant increases in share draft and share category balances supplied and retention behaviors were observed in many analyses of MPS client data in fourth quarter 2001 and first quarter 2002. This is assumed to be due to the great deal of uncertainty created by unusual events during those periods. But much uncertainty remains today, in both political and economic conditions, implying that more balances will likely be in credit unions for non-rate paid related reasons for awhile.

A further uncertainty type argument regards changes in the way members view deposits as investments. Credit union core deposits offer “principal protection” in their contract terms, via their federal share insurance. This is not so for many competing financial products, in particular stocks or bonds. As important segments of the population age, a natural change in investment portfolios is toward greater percentages of principal-protected holdings. The desire for guaranteed value, not highest rate paid, may thus be a further reason for deposits to stay in credit unions as interest rates rise.

So we now know current market conditions, and what drives the core-deposit behaviors of our members, and we expect favorable recent trends to continue unabated. How does this relate to IRR hedging? Once again, it is relatively simple but some assembly is required.

The Basics of Core-Deposit IRR Hedging

Three types of core-deposit behavior are relevant to IRR hedging: total balances supplied, repricing, and retention of existing balances. Together they determine how sensitive core-deposit interest expense and value are to changes in interest rates. Less sensitive interest expense and more sensitive values (remember, core deposits are a liability) define better core-deposit IRR hedging.

To understand the behavior of core-deposit interest expense as interest rates change, you need to quantify: a) sensitivity of funds supplied as interest rates change (member behavior), and b) your repricing decisions (credit union behavior). Stable supply and limited repricing are the ideal paring from an IRR perspective because they define slow changes in core deposit interest expense as interest rates rise. This behavior parallels the slow changes in interest income from fixed-rate assets, thus hedging these assets in income at risk IRR assessments. Proving stable supply across interest rates also affirms the favorable liquidity behavior of core deposits, something that examiners are very interested in right now.

To understand the value-related behaviors of core deposits, you need to quantify the retention patterns of existing balances through time. This requires following a fixed set of individual member accounts over time and analyzing runoff, preferably across alternate interest rate levels. Recent runoff behavior forms the basis for estimating average life (the typical amount of time core deposits stay in the balance sheet). By comparing the all-in cost of core deposits with a given average life to the cost of alternative finding with a similar term, the present value (also known as economic value) of the balances is computed.

Vary the discount rate up and down and the resulting changes in present value define the effective duration of core deposits – the percentage change in their value per 100 bp change in interest rates. The larger is effective duration, the more core-deposit values appreciate as interest rates rise. This appreciation offsets long-term asset value depreciation, and that is the value at risk IRR hedging provided by core deposits.

With all of the pieces in place, you can now ensure that you do not short-change yourself on core-deposit IRR hedging. Here are the final steps.

Measuring and Putting Core-Deposit IRR Hedging to Work

Because historic member and credit union actions are the only guide to future behaviors (the core-deposit contract offers virtually no insights), all core-deposit behaviors need to be statistically quantified. This is the only way to ensure the right IRR-related inputs are input into your ALM model, resulting in accurate IRR analyses. How to quantify core-deposit behaviors is a story for another time, but note that the level of statistical analysis required is very high due to the number of behaviors involved, the potential for balances to move among categories, and regulatory mandates for advanced level and fully documented inputs for ALM models. Many moving pieces and strong compliance mandates take the ground out from under simple approaches to measuring core-deposit behavior, so beware of anyone suggesting that quick and easy solutions are adequate.

With quantified repricing and lag inputs for rates paid, and quantified runoff data by scenario, your ALM model is empowered with the exact right inputs to calculate NII and NEV IRR. All of the interest expense stability and value related hedging power is fully represented, to provide the right answers to how your credit union’s funding side interacts with asset side behaviors.

In most cases, comprehensively quantified core deposit inputs deliver far more IRR hedging than do inputs based on ad hoc estimates or simple analytics. Re-optimizing the balance sheet in light of the new information leads to performance gains. Further, because of the proven advanced level statistics that now underlie your IRR assessments, you can also be more confident in building your balance sheet strategies. More performance again, and enhanced member benefits, are now yours.

But one last thing. You can’t just “set it and forget it” when it comes to core-deposit IRR inputs. You need an ongoing program, not just a report, to be successful and ensure regulatory compliance. The steps in a core deposit measurement, deployment, and monitoring program are as follows.

Your program starts with knowing your members, your credit union’s advantages with respect to core deposits, and your core-deposit market or markets. This process includes quantifying historic behaviors using an advanced statistical methodology and forecasting all three types of core-deposit behaviors forward in multiple interest rate scenarios. With the quantified ALM model inputs so produced, you can measure NII and NEV IRR with high precision. Rework your balance sheet for higher performance and reap the benefits.

Each month, compare forecasted retained balances to actual retention to provide an early alert of any changes in this key behavior. Every quarter, document your monthly comparisons for ALCO and the Board. Once a year, or more often if you are relying on core-deposit behaviors heavily (e.g., to support high percentages of long -term, fixed-rate assets), update the statistical analyses to keep the forecast equations up to date. At this point, also back-test prior period forecast data against actuals for the last 12 months and report the findings to ALCO and the Board. Every three years, or more often if conditions are rapidly changing, complete a new statistical study.

Finally, it is a good idea to add a core-deposit contingency plan to your ALCO policies. This can be done as an appendix to an existing contingency funding policy or as a stand-alone document. Specify the conditions under which action must be taken (e.g. core-deposit balances fall below a given percentage of funding, or pricing spreads move adversely by a defined amount, or retained balances fall below some percent of forecast values) and what actions will be taken (e.g. borrowing wholesale funds at a given term, raising rates paid, or enhancing the ambiguous influence). The plan doesn’t have to be elaborate; it just needs to show that you have considered the issue and are prepared to act quickly and decisively in a time of need.

Closing Comments

There is a thick vein of value embedded in most credit union balance sheets, in the core deposits. Because members look beyond just rate paid for value, powerful IRR hedging behaviors are built in to most credit union core deposits. Putting this advantage to work requires some effort, but few strategies provide anything close to the attainable ratio of performance to cost. And while we’re at it, some of that performance gain can be used to add even more member convenience, service, and relationship to your franchise. Which of course enhances core-deposit behaviors by adding to the ambiguous influence, and so on and so on! William J. McGuire is president and CEO of McGuire Performance Solutions, Inc. Contact him at 480-556-6771 or Wmcguire@mpsaz.com.