Asset-liability management (ALM) is at the core of banks' financial strategy. Its objective is to ensure balance between resources (customer deposits) and uses (granted loans). This management relies on generally sophisticated mathematical models. However, the sharp rise in interest rates in 2022 and 2023 exposed the limitations of current models, sometimes causing hundreds of millions of euros in losses for certain institutions.
Faced with this challenge, quantum computing could offer an innovative solution. By rapidly processing complex combinatorial problems, it could provide banks experimenting with it a decisive competitive advantage.
ALM: A critical challenge with existing model limitations
Mathematical ALM models are essential for banks to anticipate interest rate, liquidity, and credit risks. They help secure the balance sheet, optimize profitability, and comply with regulatory constraints by simulating different market scenarios. However, their ability to predict unexpected shocks remains limited, as demonstrated by the sharp rise in interest rates in 2022-2023 within an inflationary context. This situation significantly disrupted banks' balance sheets.
- On the savings side: Some French banks saw billions of euros in deposits disappear within months. According to the Banque de France, in 2023, the net flow of household investments fell to €109.5 billion, a sharp decline of €56.5 billion compared to 2022. This drop was largely due to massive withdrawals from sight deposits, contrasting with the strong accumulation observed in previous years. Customers favored higher-yield investments (such as Livret A and life insurance) over checking accounts.
- On the credit side: Early loan repayments declined, as paying off low-interest loans became less attractive in a high-rate environment, while demand for new loans sharply decreased.
In response, banks had to seek new sources of financing in the markets, often at higher costs, while increasing competition among themselves to attract or retain customer deposits. The rise in interest rates also increased the cost of customer deposits.
Overcoming current limitations: the potential of quantum computing in ALM
Today's ALM relies on advanced mathematical models and various algorithms to analyze the balance between a financial institution’s assets and liabilities. It incorporates market assumptions, economic projections, and scenario simulations to anticipate risks related to interest rates, liquidity, or fluctuations in financial instruments’ value.
The Black-Scholes model, for example, is commonly used in ALM. It evaluates the value of options and derivatives based on key parameters such as underlying asset price, market volatility, and time to maturity. In ALM, it specifically helps anticipate the impact of market fluctuations on financial products sensitive to interest rate changes. For instance, when simulating an economic shock or a sharp rate fluctuation, Black-Scholes allows banks to assess the evolution of their options and adjust risk management accordingly.
The model can be solved using various methods, including Monte Carlo simulations. Recent studies, including one from February 2025, have shown promising results regarding the potential advantage of quantum computing in these algorithms. By leveraging entanglement, superposition, and interference, quantum computing accelerates simulations by processing multiple scenarios in parallel, improving random sampling, and reducing the number of required simulations.
The rise of quantum computing thus opens highly promising perspectives for ALM. Quantum computing operates on principles radically different from classical computing. Quantum computers can process a vast number of possible solutions simultaneously.
In the ALM context, quantum computing offers the following advantages:
- Managing complex combinatorial problems: ALM models must integrate numerous explanatory factors, including market variables, customer behaviors, and regulatory constraints. Quantum computing could analyze millions of different scenarios simultaneously.
- Enhancing simulations: Scenario simulations, which take a long time on classical computers, could be significantly accelerated through quantum algorithms. Today, modifying a single variable may require hours of calculations on traditional systems.
- Predicting customer behaviors: Banks could better anticipate deposit movements and investment decisions based on market trends or geopolitical contexts by integrating a broader range of input data, such as complex behavioral models. In France, the bancassurance model adds an additional layer of complexity in modeling asset transfers between banks' balance sheets and insurers' balance sheets.
Thus, quantum technology could optimize ALM models, improve crisis anticipation, and refine response strategies.
A progressive approach to integrating quantum into existing models
However, the transition to a quantum-based ALM model will not happen overnight. The technology, while promising, remains immature. Additionally, training and integration costs are high. Finally, banking models are complex to adapt, and translating them into quantum algorithms represents a key challenge.
Given this context, a three-phase approach is recommended for banks considering a quantum initiative:
- Experimentation phase: Compare quantum algorithms with existing simulations (e.g., Monte Carlo) to assess speed and accuracy.
- Model enhancement: Gradually integrate behavioral data and additional variables (customer profiles, sensitivity thresholds, etc.).
- Full integration: Once the technology matures, embed quantum algorithms into the entire ALM framework to improve responsiveness and prevent liquidity shocks.
We offer to support banks in adopting quantum computing for ALM
Integrating quantum computing into ALM is not just about computational power—it requires a deep understanding of banking challenges and translating these challenges into practical quantum algorithms.
Sopra Steria aims to assist banks by:
Identifying business problems suited for quantum computing: With deep industry expertise, Sopra Steria helps financial institutions pinpoint key friction points that could benefit from quantum capabilities.
Translating challenges into operational algorithms: Its experts collaborate with banking teams to design tailored solutions while considering regulatory, economic, and technological constraints specific to the industry.
Leveraging a cutting-edge partner ecosystem: Through a network of leading technology partners, Sopra Steria provides banks with privileged access to the latest advancements in quantum computing while ensuring smooth integration with existing systems.
Conclusion: anticipating tomorrow’s challenges
Although inflation appears to have stabilized and recent sharp interest rate shifts seem to be behind us, future economic crises could emerge, particularly due to ongoing geopolitical tensions. In this context, ALM remains a crucial issue for banks, as even small improvements can translate into financial impacts worth tens or even hundreds of millions of euros—especially in today’s inflationary and high-interest rate environment.
Given this reality, optimizing ALM models is essential, making the exploration of quantum computing particularly relevant. Starting now—before it reaches full potential—ensures that banks can reap its benefits as soon as the technology becomes viable.