Time is of the essence when managing a portfolio of loans: The sooner a lender can detect an increase in credit risk, the better they can mitigate losses and aid customers.

Before a borrower formally enters financial distress, we can often observe indicators of increases in credit risk in news or other data sources. The media may for example report on a factory shutdown due to a flood; this directly impacts production and revenue. Monitoring of these Early Warning Indicators (EWIs), however, tends to be manual and is often a hit-and-miss, time-consuming, and inconsistent process.

It is no wonder that lenders are required to implement robust Early Warning Systems (EWSs) to aid portfolio management. In this blog series we discuss the regulatory expectations concerning EWIs and propose our innovative tool, Eagle Eye, as a solution to support the monitoring of customers using near real-time data.

Regulatory expectations with respect to early warning and monitoring frameworks

Earlier in the year, the European Banking Authority (EBA) published its final guidelines on loan origination and monitoring, which is applicable to all credit institutions in Europe. Furthermore, other international regulators such as the Saudi Arabian Monetary Authority (SAMA) have also recently released similar rules making it clear that portfolio monitoring and early warning systems becoming an area of regulatory focus.

In essence, regulators expect institutions to develop, maintain and regularly evaluate relevant quantitative and qualitative EWIs for the timely detection increases in risk. EWIs should cover all portfolios, industries, geographies and individual exposures. Importantly, EWIs should be linked to the monitoring of the institution’s current position regarding its credit risk appetite, strategy and policies also creating a feedback loop.

The key considerations outlined by the regulator is that FIs should:

  • Include relevant quantitative and qualitative EWIs that use internal and external data;
  • Set defined trigger levels and assigned escalation procedures;
  • Apply more frequent monitoring when a triggered EWI event is identified;
  • Ensure the framework is supported by an adequate data infrastructure to ensure relevant and up-to-date information;
  • Have due regard to individual circumstances, where interaction with the borrower is required; and
  • Embed and maintain a robust and effective framework for ongoing monitoring and management of credit risk and credit exposures.

The guidance outlines the indicators to be considered as part of the ongoing monitoring. These indicators can be categorised into five different data clusters:

  • Financial Data describing the borrower’s financial position, including any adverse changes therein;
  • Behavioural Data describing the expected payment behaviour of the borrower, including significant changes therein;
  • Covenant Data capturing the borrower’s compliance to covenants, including expected covenant breaches and late delivery of certificates of adherence;
  • Macro-Economic Data capturing adverse changes in the macro-economic environment that the borrower operates in; and
  • News Data capturing events reported in the media e.g. concerns raised in the reports by the external auditors, or legal action that may significantly affect the borrower’s financial position.

Challenges in developing informative early warning systems

When monitoring indicators, FIs predominantly rely on lagging data feeds such as information on the financial position of the borrower that may only be available and provided on a quarterly basis. By design the EWS should combine the abovementioned five data clusters to give an earlier view of any credit deterioration than the current credit rating model used to risk rank customers on an ongoing basis.

Additional point-in-time information on the borrowers in the portfolio is available through alerts, news and other media sources. Typically, portfolio managers and analysts may monitor these sources; however, due to the sheer volume of information identification of threats is driven by chance (essentially this can be like finding a needle in a haystack).

Leveraging the power of data to bring credit risk early warning systems to the digital age 

In our experience, combining the five data clusters creates a powerful tool to support the early detection of deterioration in credit quality.

The real value is incorporating news data, which helps to give a more up to date view on the other four data clusters (e.g. media agencies may report on earnings announcements before financials are released, or report on events disrupting production lines). Moving to a process that leverage this wealth of information can help develop a near real-time view of risk.

This process requires automated data feeds to digest and aggregate the large volume of data. Moreover, advanced analytics help understand the underlying sentiment of the news data; categorising the threats across multiple jurisdictions and languages, also linking relevant events to the borrower’s credit quality.

We have improved the traditional monitoring process, developing an approach called Eagle Eye that is fast, systematic, and cost-effective.

Our approach is designed to automatically gather and process granular information from various internal and external data sources across the five data clusters. Eagle Eye combines conventional credit data sources with high-frequency news data gathered from the web, helping to predict credit quality deterioration. The solution identifies an objective set of risk indicators using sentiment analysis and artificial intelligence; this is presented to the user for further analysis. This allows the user to focus on the most significant events affecting borrower credit worthiness and is the true power of our EWS.

Catching the worm: Saving you time and money 

The new regulatory guidelines should not be the only driver to develop a fast, consistent, systematic, cost-effective framework to monitor EWIs. By implementing such a EWS, the portfolio manager will have access to a single summary of all threats related to borrowers in the portfolio, supporting proactive risk mitigation. This not only helps save money through reduced losses, but also reduces the manually intensive effort to gather news and market intel on customers. Furthermore, FIs will be able to make more informed and proactive credit risk decisions using timely and insightful data.

Banks have already started revolutionising their early warning systems to help better manage their credit portfolios.