Dynamic Amortization Governance Model (DAGM): A Retirement Withdrawal Framework

We evaluate retirement withdrawal rules on a criterion aligned with practice: policies must be executable year after year on a single realized wealth path without foresight of future returns. The Dynamic Amortization Governance Model (DAGM) embeds a portfolio-consistent amortization anchor (DARVA) and complementary references—Guyton–Klinger, fixed safe-withdrawal, and constant-percentage rules—inside an explicit corridor governed by a capital-based robustness diagnostic. Using rolling historical simulations on U.S. (1928–2025) and euro-area (1999–2025) data, we show static rules encode non-interchangeable trade-offs among average consumption, downside floors, upside participation, and income variability. DAGM improves joint outcomes on average and maximum real withdrawals versus simpler static benchmarks in most configurations; comparisons with dynamic ARVA are nuanced on floor metrics but favorable on average spending and upside along strong paths. Extensions cover lifecycle horizons from official mortality tables, couple households via annual-rebalance effective horizons, and intentional legacy reserves ring-fenced from spendable wealth. Working paper version 2.1 (June 2026). Research and educational purposes only; not investment advice.