“Not everything that can be counted counts,” said Mike Donovan, VP, Strategic Risk Analytics & Credit Portfolio Management at Canadian Imperial Bank of Commerce (CIBC). He was the second speaker to address the September 19, 2013 evening meeting of the Toronto chapter of GARP regarding the set of Principles for Effective Risk Data Aggregation & Risk Reporting released by the Basel Committee in January 2013.
CIBC, like other Canadian banks, is adapting to the heightened risk management data requirements and building the foundation for future sustainable growth. Donovan used the opening quote by Einstein to remind the audience that big data does not immediately translate into great insight.
CIBC stores and mines on the order of 100 terabytes of risk data. There are four frontline segments of banking that require support and governance from technology & operations: risk management, administration, finance, and corporate development.
“The key success factor is to get the front lines involved,” said Donovan. Governance and infrastructure “involve a tremendous amount of work.”
For the latest round of feedback on the new Basel Committee on Banking Supervision (BCBS) regulations, only nine groups world-wide, including the Canadian Bankers Association (CBA), gave feedback. He emphasized that it was very important for the industry to provide feedback to the regulators, for example, on issues surrounding acquisitions.
At CIBC a working group was struck that included stakeholders from credit, market, operational, and liquidity risk. The self-assessment was done through spreadsheets, he said, and now they must do gap remediation. “A lot of the governance and policy issues will be done up-front, to be followed by technological changes,” said Donovan.
“What will the end state look like?” Donovan asked. Diffuse accountability will transform into clear ownership and accountability for data. During question period, he elaborated on the responsibilities of ownership: “You need human capital to clean data and enforce procedures for collecting it.”
Multiple data taxonomies will become a single consistent taxonomy. For example, the three terms arrears, past-due, and delinquent are in use in different banking areas but mean basically the same thing.
Currently, lines of business manage risk aggregation with tools in Excel; this will change to automatic high-speed aggregation. Multiple reporting sources are not aligned or connected; this will change to a “unified data model, fully supporting reporting automation.” Currently, usable data dictionaries are only in place for subject matter experts. With the new changes, the data dictionaries will be both accessible and useful—for thousands of data elements.
In brief, the bank realizes “risk data is a strategic asset,” said Donovan. He closed on a quotation by management guru Peter Drucker: “what gets measured gets managed.” ª