Abner's Portfolio

National Life Group

Modern Deterministic Scenarios

Recreated and applied a contemporary interest-rate modeling approach from SOA research, then made the expanded scenario workload practical for production actuarial modeling

MG-ALFAAzureExcelVBA

Impact

runtime reduction
up to 90%

Highlights

  • Independently recreated Mark E. Alberts' Modern Deterministic Scenarios (MDS) methodology and validated results directly with the author
  • Applied MDS to asset portfolio projections and compared results against New York 7 regulatory stress scenarios
  • Implemented clustering in MG-ALFA, reducing simulation runtimes by up to 90% and lowering cloud compute costs

Project Narrative

As an actuarial intern at National Life Group, I conducted an in-depth analysis of Mark E. Alberts’ SOA research paper on Modern Deterministic Scenarios, independently recreating his methodology and validating my results directly with the author.

I applied this approach to project the company’s asset portfolios and compared it against the New York 7 regulatory stress scenarios using MG-ALFA.

To address the significant computational demands of running the expanded scenario sets, I implemented clustering techniques that reduced simulation runtimes by up to 90%, generating meaningful cost savings on the company’s Azure compute infrastructure.

The clustering method became adopted for general-use modeling by the actuarial teams.