Ortec Finance has developed a new Scenario-Based Machine Learning (SBML) approach that can optimize any type or combination of investment objectives while managing complex non-linear constraints. The resulting optimized portfolios are then evaluated using a stochastic scenario approach in the traditional way—delivering both innovation and transparency.
In the first case study, we explored how SBML can overcome the limitations of traditional optimization techniques, including the ability to generate an efficient frontier while integrating complex constraints.
This second case study, developed in collaboration with M&G Investments, pushes optimization further by introducing 3D optimization within GLASS, allowing insurers to use three objectives simultaneously:
- Maximize assets above insurer liabilities
- Maximize expected cumulative surplus in the worst 5% of scenarios
- Minimize the Market Risk SCR Charge as a percentage of surplus assets
All while constraining the Market Risk SCR Charge itself. The result: a new frontier of efficient portfolios that reflect a true balance across competing objectives.
Why It Matters
A key strength of SBML is transparency. Unlike “black box” AI, SBML portfolios can be analyzed with GLASS’s existing tools, including contribution analysis, which explains why certain asset classes are weighted more heavily. This gives insurers and asset managers confidence in both the process and the results.
This case study highlights how SBML takes Strategic Asset Allocation into the next dimension:
- 3D optimization enables insurers to optimize across multiple objectives simultaneously.
- All SBML-generated portfolios adhered to constraints while outperforming traditional approaches.
- Contribution analysis provided transparency, showing how results are achieved and why they are superior.
The study demonstrates the significant advantages of SBML—more optimal portfolios, better trade-off management, and faster, higher-quality results than traditional methods can provide.
Learn how Ortec Finance and M&G Investments applied 3D SBML optimization to push the boundaries of Strategic Asset Allocation.
SBML will be available as a GLASS module by the end of 2025, offering insurers, asset managers, and consultants a powerful new way to optimize portfolios.
If you are interested in reading the case study in full, you can download it here.