The ramifications for bank IT are huge, since stress testing will mean updates to the data that tracks and measures capital reserves, loan performance and the governance of mortgage origination and servicing. "With Dodd-Frank in place, many institutions are going through a significant transformation of the risk and compliance infrastructure, and that's where the solid infrastructure foundation can play an important role," says Eric Ebel, senior director of Moody's Analytics.
Moody's Analytics, which provides stress testing modules and thus has skin in the game, has upgraded its RiskFoundation product to include integrated regulatory reporting and data modeling to accommodate new Dodd Frank-mandated stress tests (DFAST) and the 2013 Comprehensive Analysis and Review (CCAR) data and reporting requirements. Ebel recently did an interview with BTN in which he discussed some tips to build out a stress testing infrastructure to handle added disclosure requirements.
Banks have a number of options when choosing a partner to run stress testing programs, including Misys, Sungard (SDA), Algorithmics, Sophis and Wall Street Systems.
"Every firm that operates within capital markets risk and regulation has stress testing technologies that they offer," says Medy Agmi, a Celent analyst.
Taken in tandem, Dodd-Frank and Basel III will require most banks to hold more capital, change risk weightings, increase capital for off-balance sheet items and improve risk analysis. That's where stress testing and data management come in. While there have been delays in implementation (the original Basel III implementation date in the U.S. was supposed to be Jan. 1, 2013 — a date that has obviously come and gone) and Dodd-Frank-based regulations are still being tweaked, both laws will include provisions for mandated stress testing against systemic risks and economic shock. And the capital reserve requirements will likely involve added stress tests to ensure compliance efforts are up to date and on target.
"There is a lot of data to collect from the entire balance sheet, and that will come from different systems within the bank. With all of the resources necessary to comply, centralizing the data is very important," Ebel says. He suggests a three-pronged approach to building a stress-testing infrastructure.
First, a centralized data foundation can be helpful, particularly as data management becomes complicated due to increasingly stringent and granular regulatory requirements. This centralized infrastructure should consolidate multiple sources of internal and external data, including loan-level balance sheet characteristics, borrower and property-level risk and financial data, economic and market insights, and other information that may impact a bank's cash flows, profit and loss, capital, liquidity and risk-adjusted performance.
Second, Ebel says the infrastructure should enable group-level and subsidiary-level analysis to ensure consistency and reduce overall cost of ownership. It should allow banks to manage different corporate management hierarchies for different products in parallel and link them through mapping mechanisms. Third, balance sheet, income, losses and risk weighted assets should be managed at the appropriate level of granularity and linked through aggregation algorithms.
Finally, Ebel says stress testing should not be limited to annual or mid-cycle regulatory exercises, but should be incorporated into business practices, and should factor in a wide variety of institution or business-specific assumptions in addition to regulatory scenarios. Stress testing outputs should also be timely to the current economy, taking macro and micro factors into consideration.
Agmi says a flexible, centralized stress testing infrastructure can help accomplish the goal of the stress tests, which is to enable a bank to demonstrate to regulators and other parties it has made adequate risk and reserve provisions to absorb unprecedented future shocks to the economy.
"Prior to the 2008 crisis, stress testing had been constrained by historical data and there were few forward-looking elements," says Agmi.
Agmi says while the 2008 crisis had some similarities to prior crises, such as the hedge fund crisis of the late 1990s, there was a confluence of systemic economic risks that has not been encountered before. "It's not just one factor that is stressed or shocked at a bank, it's multiple factors at the same time, so now we are seeing stress testing of different variables around a financial institution, whether it's around funding, liquidity or counterparty default," Agmi says.