Strawberry Labs — Research

Preface

This book is a research programme in econophysics. Its central claim is a hypothesis: that the onset of systemic instability in equity markets is a phase transition whose approach leaves measurable signatures in the changing structure of the market’s dependency networks, and that those signatures precede the volatility- and correlation-based measures in standard use. We set out to state that hypothesis precisely, to build the empirical instrument capable of testing it, and to test it honestly.

Strawberry Labs is the research programme of Todd B. Adams. The SBFoundation platform is its empirical instrument: an end-to-end, point-in-time, survivorship-corrected, cost-aware research substrate that acquires and models market data, engineers and validates factors, and carries surviving strategies through to paper execution. Two things must be kept distinct throughout, and this book keeps them distinct. The platform, and the single-layer factor and comovement measurements it already runs, are established — built, exercised on live data, and reported with their uncertainty. The multiplex-network, spectral early-warning, and graph-learning programme at the heart of this research is proposed — a falsifiable plan of work, hypothesised and designed to fail cheaply if the data do not support it, but not yet built and not yet tested.

The book gathers, in one place:

• the research programme — the research proposal, the research questions, and the factor catalogue and models;
• the account of the platform as a research substrate, and why its point-in-time, survivorship-corrected, overfitting-audited discipline is what makes a surviving result credible;
• the empirical findings, the methodology, and the experiments that test the questions;
• the working journal and the publications;
• a set of appendices — the platform-depth and whitepaper documents, and a generated Reference Library of the academic sources the work builds on.

For an exhaustive, honest inventory of what the instrument can and cannot do — separating what is built, what is on the development path, and what is deliberately out of scope — see the platform capability map appendix.