Congratulations to the winning team of the Swiss Risk Award 2024, Eric Shaanning, Arsen Stepanyan, Colin Hardy and Francisco Nunez, for their groundbreaking project, The Future of Systematic Vulnerability Identification in Asset Liability Management!
The project focuses on a reverse stress testing approach to systematically identify interest rate risks for any bank balance sheet, comply with regulatory requirements, and support boards in discussing asset liability management risks in a holistic manner.
It combines well-known yield curve models with AI/machine learning techniques to analyze and identify patterns in thousands of scenarios.
The approach has been successfully implemented and run in production at a large commercial bank in Switzerland and presented to FINMA/SNB.
It has also been tested on mock balance sheets and published, revealing a key finding: the six regulatory BCBS scenarios leave blind spots in interest rate risk. These results have gained significant attention from BIS, the European Central Bank, the Bank of England, and the IMF.
The Swiss Risk Award 2024 was presented by Anton Seidel, Chair of the Swiss Risk Award on February 14, 2025, at our Flagship Event. The evening featured an inspiring keynote speech by Christian Bluhm, former Chief Risk Officer of UBS, who shared insights from his 25-year career in risk management, highlighting key lessons, challenges, and the evolving landscape of financial risk.
| Finalist in the Top 5: (Organisation) | Project: (Description) | ||||
| Investment by Objectives (IBO) SA Marc Lussy, Nicholas Hochstädter, and team. | Performance Watcher A tool to support the monitoring risk profiles for discretionary mandates, as well as advisory and execution-only services • A participative network where members anonymously share the daily net valuation of their discretionary portfolios • The network comprises nearly 20,000 portfolios, with a calculated value of CHF 58 billion and hundreds of active contributors • Generates composite indices with which members can compare and track their volatility against peers • Offers alert and reporting features to monitor and notify when risk deviations become significant • Users can compare against “neutral” or customized indices | ||||
| Prodaft Halit Alptekin, Koryak Uzan, Can Yıldızlı, Mehmet Ince, and Onur Eski | Blind Spot A risk intelligence platform that aims to address the increasing number of digital supply chain cyberattacks targeting public and private institutions • Seeks to predict incidents using precursor events, as well as attribute incidents to organizations • Monitors the risk levels of vendors, suppliers, and third and fourth parties • Can provide early warning to the companies, giving them actionable time to propagate the risks • Have collaborated closely with the law enforcement authorities in Switzerland, which yielded overwhelmingly positive results | ||||
| Sophia Hannou | Pill Scan A decision-support tool for nurses who need to identify unpackaged pills • A smartphone application based on image recognition • Simply by taking a picture of the unrecognized pill, the AI model, trained with an image databased, will give the name of the medicine with a high level of certainty • As of today, a prototype has been built with 8 pills that are considered confusing by nurses • The results of this identification by Scan pill show a high level of accuracy | ||||
| University of Neuchâtel, Technical University of Munich, University of Neuchâtel respectively Florian Weigert, Sebastian Müller, and Nikolay Pugachyov | Forecasting Mutual Fund Performance A method for aggregating individual fund predictors to distinguish between outperforming and underperforming mutual funds • Constructs a composite predictor by averaging rankings across multiple predictors • Obtains a diversified and less noisy proxy of manager skill that drives future fund performance • Accessible to a broad range of practitioners, especially those without access to extensive computational resources |