About this site

RevDog is built around a simple but demanding question: can an investing system be designed to do fewer stupid things than humans and retail quants across market cycles?

Rather than compressing complexity into a single score or narrative, the system enforces discipline through structure — separating eligibility from conviction, fundamentals from macro regimes, and analysis from portfolio construction.

The goal is not to predict perfectly, but to produce outputs that are stable, explainable, and safe enough to be trusted as inputs for real capital decisions.

The guiding principle is error minimization before optimization.

Hard filters, sector-relative comparisons, deterministic logic, and explicit treatment of uncertainty are treated as design constraints, not optional features.

RevDog assumes markets will change, data will be incomplete, and narratives will break — and is engineered to degrade gracefully rather than fail catastrophically.

In that sense, it is less a stock-picking engine and more a system for disciplined decision-making under uncertainty.