Best in-house credit risk technology: Generali Asset Management

Milad Fakurianr/Unsplash

Generali Asset Management has secured the award for Best in-house credit risk technology at the Risk Technology Awards 2025, thanks to its pioneering credit risk platform – a solution that brings together predictive analytics supported by quantum-inspired optimisation, systemic risk monitoring and practical usability to significantly enhance in-house credit risk management capabilities.

Generali Asset Management’s solution uses a predictive causality network model that reveals how credit risk propagates across corporate bond issuers. By leveraging relationships in credit spread volatility and breaking it down into its underlying drivers, the platform constructs a dynamic map of risk contagion. Originally applied to a €90 billion portfolio, its capabilities have since expanded to cover corporate and government bonds, now overseeing an impressive €230 billion in assets under management. This scalability illustrates the model’s robustness and its relevance across asset classes.

Judges said:

• “Enhancements to the predictive model using directed acyclic graphs to find relations is impressive. Helps find contagion/systemic risks.”
• “An impressive, research-led solution that blends academic insight with practical, real-world application.”
• “Good extension away from single credit to portfolio modelling.”
• “Really interesting solution – looking forward to seeing how this thinking influences broader market risk approaches.” 

Unlike traditional, often siloed, approaches to credit risk – focused on isolated issuer risk – Generali’s platform takes a systemic view. The tool maps predictive relationships between issuers, allowing portfolio managers and credit analysts to anticipate how adverse credit events in one corner of the market could ripple through to others. 

In practice, this means early warnings on deteriorating credit conditions and a proactive risk mitigation framework. This forward-looking ability proved particularly valuable during live market conditions in 2024 and 2025, as well as historical stress periods, such as the onset of the Covid-19 pandemic and the Russia/Ukraine conflict.

One of the tool’s innovations over the past year has been the introduction of advanced volatility filtering techniques. These allow for better signal extraction from noisy financial markets, improving the accuracy of spread movement analysis. Enhanced network mapping and a new feature selection framework now enable the identification of ‘key actors’ – sectors or issuers with disproportionate influence on systemic risk – helping asset managers prioritise monitoring and intervention.

The two 2024 studies on predictive causality networks and volatility-based credit risk modelling by Qyrana, Piccolini, Kuchler and others underpin the platform’s methodology. However, it’s not just grounded in academic theory: real-world backtesting and live testing show the model consistently outperforms traditional approaches, offering greater precision in identifying credit deterioration ahead of time.

See the full Risk Technology Awards 2025 results

Risk Technology Awards 2025: Tariff turmoil’s tech effects 

Upheaval in US trade policy drove demands for more data and more simulations as supervisors pushed banks to plan for the worst

The credit risk platform’s strength lies not only in its powerful analytics, but also its user-focused design. Interactive features, contextual captions and suggested configurations ensure it is accessible for users, from quantitative risk analysts to portfolio managers.

The tool’s flexibility extends beyond credit spreads. Thanks to its modular data ingestion process, it can easily be adapted to analyse a wide array of financial instruments and identify financial contagion and interdependencies across multiple asset classes.

Credit research teams use the tool to triage issuers and optimise their coverage, while risk managers gain a tool for real-time monitoring of systemic vulnerabilities. Elsewhere, portfolio managers can build more resilient strategies, informed by a granular understanding of cross-issuer relationships and portfolio concentration risk.

The platform now offers a reliable core model and the flexibility to accommodate user-driven combinations of analytical methods. Future developments include the incorporation of non-linear relationships and more diverse financial datasets.

(L–R): Mishel Qyrana, Enrico Piccolini, Enrico Kuchler

Enrico Piccolini, risk manager at Generali Asset Management, says:

“In 2022, during the crisis of a Chinese real estate giant, I was asked about its impact on our portfolios. We had no direct exposure, but we needed to measure the indirect effects on other positions. That moment sparked the idea to build a credit contagion map that could also be used proactively.”

Enrico Kuchler, risk manager at Generali Asset Management, says:

“The most immediate approach would have been correlation matrixes. But my physics background, and my work on their limitations, pushed me to look elsewhere. Inspired by Mantegna and Stanley’s work in econophysics, I wanted a solution based on causality.”

Mishel Qyrana, quant scientist at Generali Asset Management, says:

“We started with Granger causality and refined it using constrained optimisation. A quantum-inspired optimiser helped us tackle the computational demands. Incorporating clustering leverage effects and jump components, the model delivers strong live results and supports risk and investment decisions alike.”

Sponsored content