GUIDE / 03

[ RESEARCH PROTOCOL ]

THESIS BEFORE CODE

A SYSTEMATIC LOOP FOR TURNING A MARKET OBSERVATION INTO SOMETHING TESTABLE, REVIEWABLE, AND EASY TO DELETE WHEN IT FAILS.
HYPOTHESIS
STRUCTURED SPEC
GENERATED CODE
BACKTEST
REFINEMENT
REPEAT
  1. OrderResearch discipline
  2. [01]

    Observe the market first

    Start from a falsifiable market observation, not a code idea or an indicator shopping list.

  3. [02]

    Structure the specification

    Write the thesis in the wizard. Resolve material ambiguities before asking a model to produce code.

  4. [03]

    Read risk before results

    Inspect execution assumptions and bounded risk containment before looking at any backtest output.

  5. [04]

    Change one dimension

    Refine one dimension at a time, preserve the prior version, and keep the change attributable.