ORLANDO, Fla.–Many credit unions need to improve their ALM modeling and their processes around a rising rate environment.
Madonna Ritter, SVP with MountainView McGuire Performance Solutions (which is transitioning to the name Mountainview Financial Solutions), told the CUNA CFO Council meeting that many CUs need to go into greater depth around numerous areas of the balance sheet.
Ritter reminded CFOs that when doing interest rate risk measurement methodologies and measuring earnings at risk they should have a baseline forecast or projection to compare against, so that a comparison is made using constant rates.
“You want to look at your earnings at risk looking at a static or no growth balance sheet, as it really isolates the changes due only to interest rate risk, and that shows you your true profile,” said Ritter.
Moreover, she said a one and two-year time horizon should be used, as some optionality may not kick in until year two.
When it comes to fundamental expectations for IRR assumptions, Ritter recommended:
- Assumptions be reflective of the credit union and its current market.
- Segmentation of the balance sheet needs to align with pricing and behavior
- Strong preference for assumptions based on institution-specific data.
- Minimum expectation for assessment of reasonableness of deposit price sensitivity and decay rates.
- New business inputs need to reflect current contractual and market experiences.
When it comes to Chart of Account segmentation, Ritter reminded that the degree of data aggregation specified within a model’s Chart of Accounts represents a key component of model setup and implementation. Credit unions need to recognize key ALM behaviors and option effects, and break out callable investments; separate loans by product/pricing terms/amortization/method; CDs must be segregated by term; non-maturity share deposits must be separated by tier, and borrowings must be broken out according to options.
Meanwhile, with New Volume Assumptions, Ritter said replacement and new volumes need to adequately reflect current market pricing.
Other advice from Ritter:
Investments
Mortgage-backed and CMOs must be segregated. Investment scenario specific cash flows must be modeled, as must scenario-specific prepayment inputs, said Ritter.
Loans
Industry prepayment inputs or peer inputs can be used, but “you really need to back-test those prepayments and make sure they’re reasonable for your credit unions, because we’ve seen instances where they are way off.”
As for shares, including CDs, Ritter said step-up and bump-up CDs are often not broken out by credit unions, but they must be and they must be modeled. The same holds true for variable-rate CDs.
The other issue with CDs, said Ritter, is early withdrawals, as rare as they may be. “What happens when rates rise? In the past you may not have had a lot of early withdrawals, but now you’re dealing with members who are hearing about rising rates every day. So, if you don’t have early withdrawal inputs, you need to model some what if’s. Put together some numbers and run those.”
Non-Maturity Share Deposits.
Behavior assumptions applied to the institution’s non-maturity share deposit balances SIGNIFICANTLY impact the accuracy of model forecast results, emphasized Ritter. “You need to spend some time on this and do some testing.”
Credit union CFOs should create repricing betas across products and tiers, said Ritter.
“Be aware of both the lowest rate that can be paid and the lowest rate that will be paid based upon culture/strategy,” she said. “What also goes hand in hand with pricing betas is rate floors. If you don’t have them, you want to look at rate floors, even if it’s something you haven’t done before.
And Ritter reminded, “Know your competition; know where members also have funds. And remember your liquidity needs as rates rise.”