CHICAGO-It's "painful" to capture data on credit, ALM, operations and liquidity each quarter-then scrutinize them amidst 200 economic factors in six economic scenarios, three years into the future.
"But at least we do it," said Mona Leung, CFO at Alliant CU here.
The quarterly enterprise risk management (ERM) process takes one month to complete and produces a visual analysis of the $8-billion CU's financials, in the light of six "fully-loaded" Moody's Alternative Scenarios, Leung explained.
"We can think about how to adjust our capital reserve over the next three years to support the various probabilities illustrated in these scenarios," she said.
For example, if the country were to experience Moody's Scenario 1, indicating an economic recovery, Alliant can calculate the effect on net income, Leung continued. "We can discuss what mitigations would be applicable in terms of our capital reserves. We can have a conversation about what types of products we'll introduce to affect income."
Some credit unions may be tempted to restrict themselves to considering just the most critical economic factors and a limited number of scenarios when running a stress test-"that can be dangerous," Leung said.
"The world doesn't work that way," she suggested. "You have to look at all the factors and how they change and affect your CAMEL ratios in simultaneous scenarios."
Alliant begins the complex ERM analysis each quarter in statistical software, correlating business drivers, from unemployment to delinquency-"everything that might upset our balance sheet," said Leung.
An Alliant analyst then enters the correlations into Asset Liability Manager from Fiserv to produce asset/liability predictive risk models. Meanwhile, credit and operational risks are forecast in Enterprise Miner from SAS. "Asset Liability Manager and Enterprise Miner don't talk to each other," Leung said.
Data gets a little stale once the month-long analysis is complete, Leung added. "We had just finished running an analysis this summer when Bernanke said they were going to hold rates flat for two more years. We had to go back and rerun the analysis, which took us an extra two days."
Looking For That Single Tool
Leung suggested she would abandon Alliant's process in a second for a single tool that would stress-test against economic capital models with integrated ALM, credit and operational risk modules and a planning module-"so we can run lots of stress tests on our portfolios in a correlated fashion. And the correlations we run would be deeper and more granular."
Currently, Alliant has done due diligence on such a tool, eyeing products from SAS, Sungard and Oracle, she said. "These platforms are set up to do full-scenario stress-testing, not just running your curves."
The integrated platforms would probably reduce Alliant's month-long process to one week, Leung said.
ERM took center stage after the sub-prime crisis and credit crunch, Leung said. "Our board of directors and regulators have requested prudent efforts to look forward. We started three years ago and have progressed in our complexity."
