news.cuna.org/articles/108600-understand-alms-purpose-and-capabilities

Understand ALM's Purpose and Capabilities

Sophisticated modeling and forecasting tools complement market insights.

December 30, 2015

One obligation credit unions will never escape—especially in this era of heightened regulation—is asset/liability management (ALM). It’s important for credit unions to not only understand ALM from the regulators’ viewpoint, but also to know ALM’s purpose, capabilities, and limitations.

“Credit unions should view ALM in two lights: interest rate risk and liquidity risk,” says Mark DeBree, director of ALM services at Catalyst Corp. in Plano, Texas.

Most credit unions are in good shape on both fronts, according to DeBree, who cites strong balance sheets, earnings, and capital levels, as well as sufficient liquidity and ample additional liquidity sources. But from a risk management view, two factors could stress liquidity in coming years.

“Retiring baby boomers are expected to transfer $30 trillion in wealth to their heirs over a 19-year span,” DeBree says. “Many credit unions have some of those funds on their balance sheets. Should you expect a deposit outflow as these funds change hands, or see a possible uptick in balances?

“Another factor is the potential outflow of the ‘surge deposits’ some credit unions picked up during the Great Recession,” he adds. “They flooded into credit unions seeking safety from other risky assets. I think they’re likely to stay, given that they didn’t leave credit unions during the recent stock market run-up.”

Looking at interest rate risk, Frank Santucci, director of ALM services at First Empire Securities in Hauppauge, N.Y., says the credit union industry as a whole appears fairly well-positioned.

“Based on Call Report data, natural-person credit unions keep about 22% of their total assets in investments less than three years, and the industry averages almost 40% annual loan portfolio turnover," he says. "Even if interest rates go up and assets extend somewhat, credit unions can reprice a substantial percentage of the balance sheet into a rising-rate environment.

“NCUA’s focus on ALM over the past five years has improved credit unions’ performance,” Santucci continues. “The big question today is: Where are rates going to go? What is the interest rate risk on the liability side of the equation? If rates go up, how much pressure will that exert on deposit rates? After rates fell in 2008, it took four years for the cost of funds to decrease from high to current levels.

"If interest rates start rising, we’ll probably see a mirror image reversal of that trend. The industry’s margins should hold up fairly well.”

ALM best practices

Each credit union must analyze its current and historical data—rather than industry or third-party estimates—to make reasonable and supportable assumptions in the ALM process, Santucci advises.

“Many credit unions are ultraconservative and overstate their risk position,” he says. “Our question to them is always, ‘What are your assumptions?’ ALM isn’t a matter of predictions, but rather an exercise in ‘what if?’ You assign probability, not possibility, to your figures. Yes, it’s possible some completely unforeseen event might occur, but what’s the probability of it happening?

“A good portion of proper ALM is stress testing, where a credit union takes a range of likely scenarios and sees how the extremes test its asset and liquidity base,” Santucci continues. “Historical performance offers perspective. The 4.5 basis point increase from 2004 to 2008 led to deposits leaving, competition for share certificates, and rising dividend rates. What did you do then in response? How did your assets react?”

Credit unions should possess ample additional liquidity sources, DeBree says, including arrangements to add nonmember deposits if necessary and create access to lines of credit at corporate credit unions. And data analysis can provide insights into members’ behavior and find flaws in the data.

Focus on track record—rather than predictions—when crafting strategy, advises Travis Goodman, managing director of advisory services at ALM First in Dallas.

“One error credit unions often make is trying to predict rates, and then basing how many loans they’ll make on those predictions,” says Goodman. “But credit unions can make horrible predictions.”

Instead, he says credit unions should explore which asset classes are cheap and have a track record of good returns at their higher end.

“You buy a few loans in an asset class at what you believe is the low end of its price,” Goodman continues. “As its spread widens, you buy more at each five-point gain. If it reaches a certain return and then begins to contract, you unwind it a little at each five-point decline, but don’t get rid of it entirely. This kind of risk outperforms conventional, rate-based ALM.”

What vendors offer

Credit unions stand to benefit from developing an advisory relationship with vendors because “not all of the ideas for succeeding with ALM work for everybody,” Goodman says. “Credit unions should tailor vendor proposals to their organizational culture.”

As credit unions increase their modeling sophistication, Goodman suggests they consider:

Evaluation that runs 200 or 250 paths on, say, auto loan rates. “You can change the factors for each path, or set of paths, or all paths up or down,” he says.

Multifactor prepayment models that can spot loan burnout. “Over the life of a loan you can identify how many opportunities a profitable adjustment of that loan’s rate of return occurred but weren’t capitalized on,” he says.

Seasoning: Older loans will have a different burnout rate than newer loans.

Seasonality. For example, summer prepays might be faster than in other parts of the year.

Loan-to-value (LTV). It might be harder to refinance a loan that has a 96% LTV compared with one at 80%.

Loan balance. The cost to refinance might not be attractive when factored into your models.

Attribution analysis has emerged as an ALM tool, according to DeBree. “This is where you compare your assets and returns at two times of the year—say, June versus September—to identify which factors cause the greatest losses.”

The ALM landscape has dramatically changed over the past 10 years, DeBree observes, because credit unions have become more sophisticated and entered new segments. “But the losses and failures of the Great Recession also significantly changed the ALM landscape,” he says. “Examiners now ask much more of credit unions, and push even the smaller credit unions like never before. The days of just running gap ratio analysis are likely coming to an end.”

Goodman sees five ALM trends emerging in 2016:

  1. Significantly more sophisticated modeling capabilities;
  2. Comprehensive liquidity forecasting;
  3. Easier-to-implement loan-level analysis;
  4. More evolved option-adjusted spread (OAS) framework for modeling and decision making; and
  5. Greater emphasis on derivative modeling and analysis due to recent regulatory changes.

“When credit unions fret about an ‘uncertain climate,’ we say, ‘It’s always an uncertain climate,’” says Santucci. “No one can tell the future. Credit unions are in the risk management business. You accept that some loans will go south. You learn from that and continue trying to give your members higher dividends, services at a reasonable cost, and low-rate loans. Without taking some risk, you can do none of these things.”