Overview
Ortec Finance partnered with AllianceBernstein to explore how Scenario-Based Machine Learning (SBML) can optimize US insurance portfolios under complex, non-linear constraints—something traditional methods like Markowitz struggle to handle.
For a US insurer regulated under US RBC, the study focused on maximizing book yield and minimizing C1 asset risk. SBML successfully produced an efficient frontier and incorporated a key constraint: keeping asset risk below 0.5% of the portfolio’s book value.
As private markets become more prominent in insurance investing, liquidity is increasingly important. SBML’s flexibility enabled the addition of a third liquidity-related objective, producing efficient portfolios across multiple liquidity levels.
Key Insights
- SBML handles non-linear regulatory capital metrics with ease.
- Supports multiobjective optimization, improving precision in portfolio design. Enhances evaluation of trade-offs between risk, return, and liquidity.
- Applies to both book valuation (US RBC) and market valuation (Solvency II) frameworks.
- Suitable for insurers and asset managers seeking more sophisticated optimization tools.
SBML will be available for licensing by the end of the year.
About GLASS
GLASS is a leading ALM and strategic risk management solution providing powerful scenario-based analytics for institutional investors globally.
Download the Full Case Study
Access the complete report with detailed analysis and portfolio results.