Data Visualization in Credit Risk Management: A Snapshot

According to Wikipedia, “Data visualization is viewed by many disciplines as a modern equivalent of visual communication. A primary goal of data visualization is to communicate information clearly and efficiently to users via the statistical graphics, plots, information graphics, tables, and charts selected. Effective visualization helps users in analyzing and reasoning about data and evidence. It makes complex data more accessible, understandable and usable”.

Our experience has been that data visualization is a powerful tool that enables you to take a vast set of data, analyze it, quantify it, and present it in a visually-appealing and easy-to-understand manner. Not surprisingly, data visualization has become essential for gaining business intelligence and insights, expanding business beyond traditional practices, and achieving a competitive advantage in the marketplace.

As financial institutions seek to leverage their data assets to manage credit risk more effectively, they face a number of barriers. Data quality is one of the top challenges that we find across financial institutions. This challenge becomes magnified as the volume and sources of data grow. It should not come as a surprise that data quality is fundamental to sound credit risk management, as data is often the basis of credit risk decisions. It is therefore critical for the data to be correct and reliable. Secondly, financial institutions have seen significant changes since the last economic crisis. Organizational structure, including newly developed business groups or silos, makes it difficult to collect data across the organization. This results in fragmented data and an incomplete view of enterprise risk. Technology is a third impediment. The list of technology complaints is long: multiple systems trying to accomplish the same goal but producing different results, disconnected legacy applications that have been adopted through mergers and acquisitions, lack of knowledge about these systems, dependency on IT for problem resolution and decision support, etc. The result of these challenges is inconsistency throughout the organization, lack of transparency, and slow turn-around time for analytics and reporting.

Historically, banks have relied on IT and third-party consultants for data management, data aggregation, and decision support, but increasingly financial institutions are turning to data visualization tools to help them aggregate, analyze, and glean insight from their data. According to Qlik, a business intelligence and visualization software provider, data visualization tools are being used by thousands of financial institutions across the globe, including 47 out of the top 50 financial institutions. In the 2012 Big Data @ Work survey conducted by IBM, 71% of the 124 respondents from the financial sector reported that use of big data and analytics (including data visualization) creates a competitive advantage for their organization. Financial institutions, ranging from commercial banks to asset managers to insurers, use data visualization to address a variety of market needs, such as regulatory compliance, portfolio analysis, benchmarking, and model development.

Banks can use data visualization to reduce validation effort and to support regulatory discussions. Imagine that you run stressed loss estimation analytics at your bank. You use a third-party model as a challenger to forecast losses for your C&I portfolio. Your C&I portfolio consists of a significant number of mid-sized borrowers from the trade sector, while the model that you are using is developed off a data set dominated by large corporate borrowers. You can use data visualization to demonstrate and document that the model is sensitive to midsized borrowers and performs well on your portfolio.

Figure 1. Industry distribution

Source: Moody's Analytics

Figure 2. Industry EDF median

Source: Moody's Analytics

Portfolio managers, risk managers, and underwriters at banks and other financial institutions use data visualization to uncover risk concentrations and to inform loan approval decisions. Real-time dashboards can highlight portfolio improvement opportunities. Suppose you discover that your portfolio is overly concentrated in Real Estate counterparties, which is higher risk than your Northwest portfolio. You next evaluate your exposures in Utilities, managed by a separate line of business within your organization, and find that you have minimal exposure to this sector, even though the Utilities sector has historically been the least risky in your relevant markets. A question to consider – have you been making the right lending and investment decisions given this new insight?

Figure 3. Risk distribution by geography

Source: Moody's Analytics

Figure 4. EDF trend

Source: Moody's Analytics

Using data visualization for benchmarking Data visualization is an effective tool for benchmarking, a common practice at financial institutions. Data visualization enables you to quickly answer questions such as “How does our current portfolio compare to our portfolio a year ago?” or “How does our portfolio compare to industry benchmarks?” If you find that interest rates on your new CRE originations have decreased over the past year, then you might suspect that lending practices have changed. However, if you see similar trends in your peers’ portfolios, then you might conclude that an external force (e.g., macroeconomic factors) is at play. Benchmarking helps you understand your portfolio results and trends in a broader context, allowing you to respond more appropriately.

Data visualization is also useful for model development. Data visualization tools make it easier for you to visualize large amounts of data across multiple dimensions and to identify trends and relationships in your data. For example, data visualization could help you identify correlations between dependent and independent variables in your retail portfolio that were not previously considered. Using this new information, you could improve the performance and predictive power of your credit risk models.

Figure 5. Rating distribution

Source: Moody's Analytics

Figure 6. EAD by rating count and EDF

Source: Moody's Analytics

At Moody’s Analytics we have invested in and adopted data visualization tools for data quality checks, model development, benchmarking, and business reporting. We also offer data visualization functionality as part of our products. As a leader in enterprise risk management solutions, Moody’s Analytics is combining its in-house expertise, proprietary data, and credit risk analytics to provide innovative new products that will help our clients bridge the gap between data and credit risk management.

Business transformations driven by data visualization could include real-time data quality dashboards, a comprehensive view of risk across the organization, and self-service analytics that cater to business users and decision makers. Data dashboards allow practitioners to easily pinpoint portfolio outliers and identify potential data quality issues. Data from multiple systems and sources can be linked via a data visualization tool which will serve as a hub for customer intelligence. The hub will facilitate reporting at the enterprise-level rather than across disconnected lines of business, breaking down silos and producing new business insights. IT dependency and long service request queues will be a distant memory, as business users will have direct access to empowering data and analytics for decision making, root cause analysis, and reporting. Last but not least, self-service data visualization will free resources to work on more important business goals.