Reinforcement Learning for Micro-Level Claims Reserving
On the map: Deep reinforcement learning methods → Adversarial robustness and RL in finance
Abstract (arXiv)
Outstanding claim liabilities are revised repeatedly as claims develop, yet most modern reserving models are trained as one-shot predictors and typically learn only from settled claims. We formulate individual claims reserving as a claim-level Markov decision process in which an agent sequentially updates outstanding claim liability (OCL) estimates over development, using continuous actions and a reward design that balances accuracy with stable reserve revisions. A key advantage of this reinforcement learning (RL) approach is that it can learn from all observed claim trajectories, including claims that remain open at valuation, thereby avoiding the reduced sample size and selection effects inherent in supervised methods trained on ultimate outcomes only. We also introduce practical components needed for actuarial use -- initialisation of new claims, temporally consistent tuning via a rolling-settlement scheme, and an importance-weighting mechanism to mitigate portfolio-level underestimation driven by the rarity of large claims. On CAS and SPLICE synthetic general insurance datasets, the proposed Soft Actor-Critic implementation delivers competitive claim-level accuracy and strong aggregate OCL performance, particularly for the immature claim segments that drive most of the liability.
Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Placed in Adversarial robustness and RL in finance, one of 15 directions in Deep reinforcement learning methods. Builds on 16 of its 61 references that are in the map. First posted on arXiv 12 Jan 2026.
Builds on
16 of 61 references.
References · 16
- 2017Proximal Policy Optimization Algorithmscited by 15,215
- 2018Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actorcited by 4,885
- 2013Playing Atari with Deep Reinforcement Learningcited by 3,274
- 2017Mastering Chess and Shogi by Self-Play with a General Reinforcement Learning Algorithmcited by 734
- 2017Population Based Training of Neural Networkscited by 367
- 2018Meta-Gradient Reinforcement Learningcited by 159
- 2021Recent Advances in Reinforcement Learning in Financecited by 110
- 2018Deep Hedgingcited by 108
- 2019Learning an Adaptive Learning Rate Schedulecited by 29
- 2019Deep Reinforcement Learning for Imbalanced Classificationcited by 24
- 2021Autocalibration and Tweedie-dominance for Insurance Pricing with Machine Learningcited by 18
- 2021Stochastic loss reserving with mixture density neural networkscited by 5
4 more
- 2018DeepTriangle: A Deep Learning Approach to Loss Reservingcited by 4
- 2020Individual Claims Forecasting with Bayesian Mixture Density Networkscited by 4
- 2025Adaptive Insurance Reserving with CVaR-Constrained Reinforcement Learning under Macroeconomic Regimescited by 1
- 2025On the use of case estimate and transactional payment data in neural networks for individual loss reservingcited by 1
Built on by
Citing papers per year: 2026: 1.
1 paper.
- 2026From Chain-Ladder to Individual Claims Reservingcited by 0
- 2026From Chain-Ladder to Individual Claims Reserving17 Feb 2026
Similar, not cited
Close in wording, no citation either way
- 2025Dynamic Reinsurance Treaty Bidding via Multi-Agent Reinforcement Learning
- 2025Continuous-Time Reinforcement Learning for Asset-Liability Management
- 2025A Hybrid Framework for Reinsurance Optimization: Integrating Generative Models and Reinforcement Learning
- 2025ClauseLens: Clause-Grounded, CVaR-Constrained Reinforcement Learning for Trustworthy Reinsurance Pricing
- 2026From Control Boundary to Insurance Claim: Reconstructing AI-Mediated Losses Through the CER Framework
- 2025Reinforcement Learning for Durable Algorithmic Recourse
- 2026Auditing the Risk Claims of Distributional Reinforcement Learning
- 2024Reinforcement Learning applied to Insurance Portfolio Pursuit