Allstate and IBM are exploring how quantum computing could improve insurance portfolio management, with new research examining its potential to optimise homeowners insurance portfolios exposed to catastrophe risks.

The joint study, published on arXiv in May, investigates how a hybrid quantum-classical computing approach could help insurers balance portfolio growth while remaining within defined risk tolerances.

The research focuses on the mathematical “knapsack problem”, an optimisation challenge that involves selecting the most valuable combination of items without exceeding a specified capacity. Applied to insurance, the technique could help carriers determine which policies to write while managing overall catastrophe exposure.

The problem is particularly relevant in homeowners insurance, where natural catastrophes such as hurricanes, wildfires and tornadoes can generate highly correlated losses across thousands of insured properties.

According to Allstate, the insurer currently performs around 100,000 simulations to model potential portfolio outcomes, although rare catastrophe events remain difficult to capture accurately using conventional approaches.

Eric Huls, Chief Analytics Officer and Chief Data Officer at Allstate, said portfolio-level thinking is essential for catastrophe risk management.

“On the homeowner side, it really requires us to be thinking from a portfolio perspective and not just from an individual risk perspective.”

The research used IBM Quantum Heron to generate candidate portfolio combinations, before applying classical computing techniques to refine solutions that exceeded predefined loss limits and guide subsequent optimisation rounds.

The hybrid approach was compared with four classical approximation methods, with each given 30 minutes to solve the same optimisation problem.

For portfolios containing up to 75 items, both the quantum-classical and classical approaches achieved optimal results. On larger and more complex problems, the hybrid method remained competitive and marginally outperformed the strongest classical alternative under tighter risk constraints.

While the researchers said the technology is not yet ready for commercial deployment at enterprise scale, the project is intended to help prepare Allstate for future applications as quantum hardware matures.

The research highlights growing interest among insurers in quantum computing as a potential tool for tackling computationally intensive challenges such as portfolio optimisation, catastrophe modelling and capital management. Although practical insurance applications remain in the early stages, several carriers and technology providers are investing in quantum research to position themselves for future advances in computing capability.