Forecasting Non-Life Insurance Premiums and Emerging Risk Factors in a Developing Market: Evidence from Albania, 2010–2026, with Projections to 2032

This paper provides the first monthly frequency analysis and multi-year forecast of the non-life insurance market in Albania. We construct a harmonized monthly dataset of gross written premiums and gross paid claims covering 2010 to mid-2026 from the statistical reports of the Albanian Financial Supervisory Authority, following a documented reconstruction protocol that other researchers can reproduce from the public reports. Seasonal time-series models, validated through a hold-out sample and rolling-origin cross-validation, outperform standard benchmarks and yield calendar-year forecasts with simulated prediction intervals through 2032, cross-checked against residual-bootstrap alternatives. The central path indicates continued expansion of premium volume at a gradually moderating pace, while paid claims are projected to grow faster than premiums, implying sustained pressure on underwriting margins. A covariate-augmented extension estimates an income elasticity of premiums close to unity yet does not improve out-of-sample accuracy, supporting the univariate forecasting model and enabling scenario-conditional projections under alternative income growth paths. Structural-break tests date significant shifts to the motor tariff reform, the earthquake, and the pandemic, and cointegration analysis identifies a single long-run premium–claims equilibrium in which claims adjust to premiums while pricing does not respond to claims experience. We draw implications for supervision, pricing conduct, and the EU-accession regulatory transition.