Asset liability management attempts to simulate what will occur to net interest margin, income and capital when rates change. The simulation is reliant on assumptions of how fast and how much deposit account balances and costs will change and the speed at which loans and investments pay down.

Has the practice of ALM morphed into treating the worst possible outcome as a reality?

Many credit unions feel compelled to shock the balance sheet to “keep the regulator off my back.”  A balance sheet shock assumes dramatic rate increases that are instantaneous, large, parallel and sustained. Historically, large increases have occurred but they were never instantaneous or parallel and hardly sustained. Understanding the impact of a shock on the balance sheet is an exercise given enormous significance in spite of the near impossibility of actual occurrence. 

Investment decisions at many credit unions seem to be driven by the unlikely shock mentality (we will describe this in relation to banks below).

Ironically, the perceived risk to income and capital from rising rates has resulted in less credit union income and capital in the declining and flat rate scenarios.

What Has Occurred in the Past?

The two largest rate spikes in the last 25 years (and among the largest in history) are summarized in the accompanying table. (click to expand) For perspective, also review the graphs depicting credit union and bank industry cost of funds as Fed Funds change.  Note the number of months from trough to peak:  24 months and 15 months – hardly instantaneous!

In the first 12 months of Fed rate increases beginning 2004, bank and credit union deposit costs changed just 9% and 6% of the Fed funds rate change respectively.  Trough-to-peak, the change in cost of deposits is 45% and 31% respectively of the total fed fund movement, but that took two years which allowed financial institutions to receive some benefit from adjusting asset pricing as well.

Banks and credit unions both significantly lag the adjustment of their deposit costs versus movement in the fed funds rate. As you note the actual 1993-1995 percentage changes (Implied Beta, see table above), consider what change in deposit costs are being calculated in your credit union's ALM simulation.

In discussions with credit unions, we are hearing expectations of deposit cost changes ranging 35%-50% of Fed funds in the first year.  To expect such change is to imply that margins are more at risk than what has been the case historically (see table), the expectation leads to a shorter target for asset duration than necessary, which in turn, leads to foregone income.

Banks and Credit Unions Behave In Unspoken Unison

Net interest margin at financial institutions has been eroding since the early 1990s. The erosion is due primarily to an oversupply of lenders combined with technology that makes rate shopping easy for consumers, which have led to the commoditization of product offerings. Will the managers universally struggling with margin erosion for years take action that erodes margin further simply to be in concert with Fed action? 

Depositors with large balances tend to be older and more risk averse. Those depositors are 10 years closer to retirement (or retired) since the last rate spike (2004). Will depositors suddenly chase yield by leaving insured deposits for the stock market (currently at 17,800) or the bond market?

Did the consumer move funds in the above two rate spikes? Barely. In the first year of Fed action, core deposits as a percentage of total deposits for credit unions went from 49.4% to 49.5% in 2005 and from 56.3% to 54.4% in 1994 (core deposits are regular shares and drafts). Interestingly, total shares grew in both time periods.

With compliance costs accelerating at breakneck pace, margins are not expected to support a shift from previous pricing behavior and while consumers' behavior indicates acceptance.

IRR (Mis)Treatment Is Another Good Reason for Regulatory Standardization

The proposed risk-based capital rule, the creation of a separate office of supervision for the five largest credit unions, and the recently finalized liquidity rule are examples of the trend standardizing regulatory treatment of banks and credit unions. Unlike bank capital regulation (based on managing credit risk), credit union regulation is based on avoiding IRR when rates rise. This focus ignores the fact that bank and credit union balance sheets have evolved to become substantially similar as they pursue the same households with the same financial products and creates a competitive disadvantage for credit unions. This is because the “rate shock” mentality has spilled over to asset decisions in the credit union's effort to conform to regulatory focus on IRR. How much this affects balance sheet strategy is debatable and distinct to each credit union, but it's clear credit unions have targeted an unnecessarily short duration in the investment portfolio and left significant income on the table.

The graph to the left (click to expand) shows the differences in investment portfolio yields for credit unions and banks with assets exceeding $100 million over the past 11 years.   If credit unions matched banks in investment yield, the additional income represents $22.9 billion, according to NCUA and FDIC call reports using yield difference and year ending credit union investment portfolio balances for the 11 years. Although some credit unions already match banks' investment yield, the difference in aggregate is material and implies a systemic cause. 

We Become What We Measure

There are four main reasons for this performance difference. All four reasons can be tied to regulatory perception of IRR, which large credit unions increasingly struggle to balance with the need for more income.

First, banks leave less funds in overnights, and instead more “fully invest” the balance sheet. Banks manage to longer investment portfolio duration, rely more on mortgage-backed securities and less on callable agency securities, and more actively manage the portfolio. In countless discussions with credit unions, we are told that regulatory scrutiny on mortgages and long term assets has led to overweighting on callable agency securities (whose characteristics can mean worse extension risk than MBS and no monthly cash flow).  Banks and their regulators seem to have struck a balance between earnings generation and IRR management.

Credit unions need a level playing field on regulation. The NCUA has already authorized the investments that allow for competitive investment performance and should complete the regulatory parity to banks in regard to IRR. If retained, $22.9 billion incremental investment income over the 11 years would have raised the net worth ratio by 225 basis points, to 13.08% for credit unions greater than $100 million in assets.

What's Next?

The NCUA announced that a new IRR regulation will be released soon. Additionally, a key “agency priority” in the 2015 Performance Plan is “to develop a proposal for a separate interest rate risk component for complex credit unions' risk based net worth requirement.” This performance goal appears to reintroduce IRR into RBC after the final rule is approved.

NCUA should strongly consider the above data regarding foregone income and capital and the historical evidence that financials were reluctant, in unison, to raise their cost of funds until absolutely necessary.  

IRR regulation and supervision should focus on outliers that feature multiple balance sheet characteristics of concern (not just one). Supervision should not include treating all credit unions as if they are outliers.

Perhaps there are other risks requiring more attention given that no credit union failure has been attributed to interest rate risk and while recognizing the impact of IRR regulation on income and capital accumulation. 

Peter Duffy is managing director with Sandler O'Neill & Partners LP in New York City. He can be reached at [email protected] or 646-427-1490.