Daniel Aunon-Nerin, head of credit portfolio management at Swiss Re Corporate Solutions, outlines the evolution of credit portfolio management and the transformative impact of modern technological advancements on this critical financial function.

 

When I tell people what I do, they often look puzzled: “Credit portfolio management? What is that?” Even finance specialists, with their extensive knowledge of the industry, usually follow up with: “Is it about buying and selling bonds?”

Well, kind of.

I try to simplify it by saying: “I provide essential strategic support to the business.”

Most people accept this explanation with a polite nod, not aware that credit portfolio management (CPM) has become a crucial function for financial institutions exposed to credit risk. This evolution began in the 1990s when banks started using statistical models – including credit scoring tools and portfolio optimisation techniques – in their loan portfolios to assess creditworthiness, diversify risk and optimise risk-return trade-offs. The functional evolution of credit portfolio management continues to this day and will only accelerate over the near term as data analytical capabilities and technology keep advancing.

At the heart of CPM lies the initial steps: origination and risk assessment.

At Swiss Re Corporate Solutions’ Credit & Surety, we provide insurance to banks in support of their trade and asset-backed financing to their clients. Consequently, our portfolio consists of a set of payment obligations triggered by specific events of default. These obligations are illiquid and non-tradable. Our underwriters analyse individual transactions and assess their respective risk/return characteristics, while pricing actuaries construct sophisticated models to quantify these risks.

Yet a complex challenge arises when trying to develop an understanding of how these individual transactions impact the broader portfolio. This is where the credit portfolio manager steps in.

The aggregate impact of each transaction – past, present and future – on the overall profitability of the portfolio is analysed in detail. This analysis is the foundation of the CPM role: determining performance measurement.

However, the journey does not end there. Depending on the underlying portfolio, losses might differ in frequency and severity. While losses may be infrequent in typical credit portfolios, they often exhibit a clustering behaviour called dependency. This is the most important and challenging risk to manage. Correlations and other nonlinear dependencies need to be analysed, and simulations and stress tests performed.

CPM also involves complementing our analyses with tools that allow us to shape the future of the portfolio. Decisions must be made based on the portfolio’s maturity profile, with consequences unfolding over time. Therefore, having forward-looking views on the main risk and return drivers is essential.

Once these views are established, transparent communication with the underwriting community is crucial. This often involves capital allocation to individual businesses. As conditions might differ from expectations, proactive risk mitigation techniques are also implemented to control losses in certain scenarios. Continuous portfolio monitoring ensures agile decision-making based on real-time insights.

While the building blocks of successful CPM will likely remain consistent, advances in data analytical capabilities and technology will enhance the quality of CPM contributions.

Initially, CPM work focused on regulatory adjustments, globalisation and financial product innovation.

Today, the focus has shifted.

“Our clients are now looking for ways of reducing paperwork to become more efficient, increase security and reduce fraud – and more recently efficiently incorporate increasing ESG requirements,” says Veronica Assandri Foldnes, global head of commodities & energy transition finance at Swiss Re.

The industry is moving towards digitisation. McKinsey & Co’s report The multi-billion-dollar paper jam estimates that it takes up to 50 sheets of paper to document one single shipment and that 30 billion pieces of paper are exchanged annually. The solution? A wholesale shift towards digitisation of trade documents that could represent a cost savings of up to US$6.5bn for the industry and increase trade volumes by US$30-40bn. In parallel, the adoption of the legal entity identifier (LEI) represents a crucial step towards standardisation and digital homogenisation.

These advances serve to drive down fraud, which is estimated to cost market participants billions of dollars annually. Yet, despite the clear benefits, the task is proving more difficult than expected as participants struggle to align amidst a patchwork of regulations and competing interests. The big change is going to take a bit longer to materialise.

Financial institutions, banks and insurance companies have long recognised the power of advanced data analytics, and the industry-wide disruption created by rapid advances in artificial intelligence (AI) has definitely accelerated developments in this space. A recent survey by The Economist showed that 85% of IT executives in banking have a clear strategy for adopting AI in the development of new products and services, seeking to balance business benefits with regulatory complexity and retain customers’ trust.

Swiss Re has for some years partnered with big data software providers to help extract the best from our data.

Julie Yost-Zihlmann, head of data curation & reporting at Swiss Re, says: “Putting the data in order was a clear goal. However, we decided to empower the business to drive the process, selecting the most important data that can move the needle. We provided mixed teams with the means, but the business is responsible for bringing that data together from the different tools and making it ready to be processed by users and AI applications, creating a single version of truth. Our strategy is clear and simple: first, get the right data in the right format; then select the biggest opportunities for analytics and finally form decentralised teams united by a central governance.”

These advances are also changing how we manage our credit portfolio in practice. Data collection is becoming more efficient through the use of optical character recognition (OCR) tools to read financial statements, notes and contracts.

Application programming interfaces (APIs), a software intermediary that permits two applications to talk to each other, allows seamless data transfer and enables us to efficiently connect with platforms hosting multiple terabytes of data.

Risk assessment, too, is enhanced by AI-powered algorithms that can analyse vast amounts of data from different sources, such as news and sentiment analysis, providing comprehensive insights into borrower creditworthiness. This method far surpasses traditional financial analysis, therefore leading to better-informed, more accurate risk assessment. Such technology can also analyse textual data from customer interactions and credit and market reports to assess borrower behaviour and sentiment trends, enhancing both credit decision-making and risk monitoring.

AI-powered systems can continuously analyse performance, monitor key risk indicators and alert us of emerging risks or deviations from expected outcomes in real time. Similarly, we use several AI techniques, such as machine learning and predictive analytics, to forecast forward-looking views that we incorporate into our models for portfolio steering. Future efforts will be dedicated to portfolio optimisation, where AI-driven algorithms can help optimise credit portfolios based on risk-return objectives, market conditions, forward-looking views and risk appetite.

“In the future, AI-powered decision engines could potentially handle automated decision-making for underwriters,” says Michael Lum, global head of political risk insurance at Swiss Re. “However, for now, the focus remains on developing tools that allow us to improve operational efficiency in our processes by efficiently summarising vast quantities of heterogeneous information and automating processes as and when possible, thus reducing turnaround times and enhancing customer experience.”

The AI revolution is not just reshaping customer interaction; it is and will continue to fundamentally alter how credit portfolios are managed and monitored. From digitisation to AI, the financial landscape is undergoing a transformation that promises to redefine CPM for generations to come.

I am very excited to follow this journey, but will people understand what I do?