Predictive Credit-Risk Models as Managerial Interventions in Chilean Life Insurance Companies*

Purpose: This study examines which type of predictive credit-risk model, used as a managerial intervention, is most effective in improving solvency stability and risk-adjusted profitability in Chilean life insurance companies.The study addresses delayed managerial recognition of issuer-level credit deterioration in long-term, relatively illiquid investment portfolios. Design/Methodology/Approach:The study uses a Rapid Evidence Assessment (REA) to synthesize evidence on predictive credit-risk model types used in insurance and comparable institutional investment settings.Searches were conducted across academic and specialist databases and institutional sources.Studies were screened using explicit inclusion and exclusion criteria, critically appraised using the TAPUPAS framework, and synthesized through qualitative coding and thematic analysis within a CIMO (Context-Intervention-Mechanism-Outcome) framework.The final evidence base comprised 13 studies.Findings: The evidence indicates that predictive, issuer-level models can improve forwardlooking solvency management, risk anticipation, capital positioning, and portfolio performance when their outputs are interpretable, timely, and aligned with managerial decision processes.Structural, Merton-type approaches appear particularly actionable for Chilean life insurers because they combine issuer-level differentiation with relatively strong interpretability.However, the effects are conditional on data quality, market liquidity, regulatory requirements, governance, and managerial responsiveness.Practical Implications: Life insurers should implement predictive credit-risk modelling through phased pilots focused on high-materiality issuers, develop standardized signal-toaction protocols, and align model outputs with solvency, compliance, actuarial, and governance processes.Predictive models should be complemented rather than replace professional judgment.Originality/Value: The study integrates credit-risk modelling, insurance solvency, evidencebased management, and contingency theory to assess predictive models as managerial interventions rather than solely as technical forecasting tools.It provides context-sensitive implications for Chilean life insurers operating under long-duration liabilities, concentrated portfolios, and illiquid corporate bond markets.