Chief risk officers can begin now—mid-pandemic—to build a risk management strategy fit for the future.
Risk management is emerging as perhaps an unusual candidate for bold change as financial services institutions (FSIs) look ahead and prepare for an eventual post COVID-19 age. Few in financial services have felt the complexities of the pandemic as much as risk managers. As we explored in a previous point of view, fast-changing data, uncertain timelines, conservative and irrelevant model assumptions, and increased sensitivity to aftershocks posed simultaneous challenges for the Chief Risk Officer (CRO).
As FSIs shift gear from firefighting to addressing the aftermath of COVID-19, it is key that CROs build dexterity now.
We should expect heightened risks (particularly credit risks) and dynamic interplay between risk types (especially between credit and liquidity risks) for a long while. Already, provisions for credit losses across the eight largest US lenders amounted to a staggering 244 percent of their combined net income in Q2 2020, up from 175 percent in Q1. European banks face a similar prospect of ballooning bad debt. The total loan losses across US and European Union banking sectors could hit $880 billion from 2020 to 2022.
Potential losses of this scale make the aspirational goal of accurate, real-time risk reporting and decisioning a necessity. As FSIs shift gear from firefighting to addressing the aftermath of COVID-19, it is key that CROs build the dexterity now to both address the fast-moving environment and lay the groundwork for a risk management strategy fit for the future.
Boosting data and analytics efforts
This calls for broadening the range of data inputs and stepping up the use of advanced analytics in risk quantification and front-line risk management practice. We recommend the following steps:
- Harvest existing, external data to help understand the dynamics and nuances of the changing environment and its impact on the bank’s customers. For example, geographic specific data on COVID-19 infection rates help to anticipate local lockdowns. Supply-chain information can help infer the spread of COVID-19 shock waves. At the same time, local unemployment rates and stimulus package uptakes by area and customer type can be an indicator of where risk hotspots exist or are expected to emerge once furlough schemes and mortgage payment holidays come to an end.
- Reconsider industry classification and take a dynamic view in risk decisioning. One focus area pertinent for COVID-19 is on how companies have adapted the nature of their businesses: the restaurants that have morphed into food delivery services, the machine producers now manufacturing personal protective equipment. The industry classification codes used in traditional risk decisioning models may no longer be appropriate.
- Link the broader external information with key internal data points at the customer level, such as early indicators of potential credit stress. A 360-degree view of the customer is key to gaining clarity over the dynamic context in which the customer is operating and the specific risk drivers and threats that the complex environment brings.
- Let the data tell the story by applying advanced analytical techniques across the combined data set to drive detailed discovery and insights at a granular level. Specific use cases include the application of behavioral analysis to understand and classify customers into much smaller sub-populations for focused risk management. The use of “entity resolution” to identify vulnerabilities to ripple effects as stresses pass through connected networks from supplier to buyer, from employer to employee. Across large portfolios of fast-moving data, artificial intelligence (AI) and machine learning techniques may be the only way to really understand what is going on, quickly—and to get ahead of the risks.
- Use the data-driven view to sharpen risk identification, quantification and prioritization by “feeding” frequent output of AI/ML results to practical tooling for front-line risk managers to use in decision making. Examples include dynamic dashboards and alerts that classify customer micro-segments, eliminating bias and providing a forward-looking view of those cohorts likely to thrive, survive or need to be revived. To help prioritization, drilldowns to the individual customers exhibiting “pre-default symptoms or facing the highest external risks are key to highlighting where the best opportunities lie to proactively intervene with support or forbearance strategies.
Immediate uncertainties aside, banks should emerge from the pandemic with even lower interest rates and squeezed profitability—the prospect of business chasing yield at the expense of higher risk looms large. At the same time, new ways of working, consuming and living are emerging from the pandemic. These cast a new contour for financial services and move risk management closer to an inflection point.
FSIs need to rethink and recalibrate for the impact of lasting behavioral changes on risk appetite, risk pricing, and recovery and collections strategies. Building broader and richer sources of data is fundamental for this, as is the push towards convergence and alignment of cross-functional data and process to manage the interplay between risk types.
Advanced analytics may become “core”
While the pandemic has caught many risk managers by surprise with unprecedented complexity and uncertainty, it is also propelling the future of risk management towards one powered by AI.
The need for speed, agility and explainability in applying “atomic” level modeling and scenario analysis is now coinciding with the desire for faster and more automated decisioning. Advanced risk analytics lie at the heart of turning that into reality.
To help our clients respond to the global pandemic challenge and outmaneuver uncertainty, Accenture has created a hub of our latest thinking on a variety of COVID-19 topics. To find out more on the topic of credit risk, please contact the authors.