Skip to Content

An effective early warning system for the post-pandemic banking sector

Grigori Tymchenko
September 7, 2020

Background

The COVID-19 pandemic has led to postponed deadlines for the implementation of the new Basel IV capital requirements for the banking sector. By doing so, regulators intended to relieve credit institutions of the impact from the expected organizational and administrative burden associated with Basel IV implementation, thus enabling banks to concentrate on the business operations. The urgent operational tasks include, in particular, the identification of zombie companies – those for which default risks have increased due to the difficult economic situation – and to perform measures to reduce these expected risks. These zombie companies are no longer able to offer innovative products and services, have sales difficulties, and are kept afloat by constant refinancing efforts. Regulators are aware that a significant proportion of these companies are currently on bank books, but the entire extent of this situation is uncertain. European banks are clearly aware of the problem, and they are working on early warning systems to identify these zombie companies, but they are often limited by a lack resources and know-how.

Our tools for identifying zombies

Finding zombie companies manually among thousands of corporate clients would be a Sisyphean task for analysts, but existing technology can help. Our data-driven Finance, Risk, and Compliance team has developed a two-tiered solution that uses real-time data from corporate databases and online sources to analyze and present a company’s current economic performance.

  • Phase 1

The goal here is to identify and quantify the risk-relevant signals, such as insufficient product quality, personnel bottlenecks, supply chain disruptions, current sales difficulties using data mining methods (sentiment analysis, topic modelling) from the unstructured text data. In addition, both balance sheet analyses and cash flow forecasts are prepared. The analysts are automatically notified as soon as the business risk of a corporate customer has increased significantly and cost coverage and thus debt service is at risk.

  • Phase 2

Once potential zombie companies have been identified, the risk of default should be quantified. This is done by analyzing and forecasting key financial figures. Our engine collects macroeconomic and company-specific financial data from over 42 million companies worldwide. Based on this data, scenarios are generated that serve as input for modelling default probabilities and other credit risk parameters. The use of forward-looking data leads to a high selectivity of credit risk models (PD, LGD, CCF), which has been proven over the years as required by supervisors.

Recommendations for implementation

When integrating such early warning and risk quantification systems, the implementation of regulatory requirements must be a priority. In our recent blogs, we specifically address the new regulatory challenges posed by Basel IV reform. Thus, we show which innovative approaches can be used to optimize credit risk management, for example by tokenizing risk positions as described in our first article. Furthermore, we address the question whether the use of IRBA is still worthwhile and how the benefits can be analyzed. In the third article, we determine how high the capital requirement will be as a result of the introduction of new output floor rules. In addition, Basel IV requires the appropriate discriminatory power of risk models. We discuss three possible adjustments in the fourth article. Small parameter adjustments to the risk modelling input floor have a big impact on RWA – we examine the exact impact in the fifth article. Finally, article six discusses the most important success factors for innovative risk management against the background of the aggravating circumstances caused by the pandemic, and this article deals with effective early warning systems for regulator-compliant risk management. Thank you for reading!

Capgemini Invent addresses the CxO data strategy and supports its clients in data-driven value creation.