F2R Replication v1.2 — Paper v0.16

Replication update accompanying A Second Opinion for the Portfolio: How a New Forecasting Model Changes Cross-Asset Allocation Decisions, paper v0.16.

What's new in v1.2

• Adds Figure 6, comparing cumulative net wealth across three portfolio configurations using 112 monthly observations from May 2017 through August 3, 2026.
• Reports terminal normalized wealth of 3.62× for the incumbent ensemble, 4.32× for the selected C3 hybrid, and 4.50× for the alternative C2 configuration.
• Extends the public reproduction code and numerical validation to cover Figures 1–6.
• Includes the revised 10-page manuscript and preserves the original research evidence, Figures 1–5, and historical v1.1 release.

Interpretation and limitations

Figure 6 displays monthly-observed cumulative net wealth, not a reconstructed daily return path. The paper's volatility, Sharpe ratios, and maximum drawdowns remain based on 2,327 daily portfolio return observations.

The Chronos-2 analysis is a retrospective technology counterfactual. Its pretraining information is not established as point-in-time available throughout the historical evaluation period.

The release supports reproduction from frozen derived research evidence. It does not rerun model training, redistribute raw market data, or reproduce the full historical portfolio simulation.

Paper: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=7460378

Research platform: https://research.slackquant.com/research/second-opinion-portfolio/

F2R Model-Adoption Replication v1.1

Public paper-level replication release for A Second Opinion for the Portfolio: How a New Forecasting Model Changes Cross-Asset Allocation Decisions.
This release reproduces Figures 1–5 and validates the principal reported numerical results from frozen derived research evidence. It does not rerun Chronos-2 inference, refit the conventional forecasting models, redistribute raw market data, or expose the live F2R production configuration.