Working manuscript · October 2026 · Not peer reviewed
Quantem
Auditing timing, costs and graph value in relational stock prediction
An audit of timing, transaction costs and graph evaluation in relational stock prediction, separating controlled evidence from recovered historical outputs and conditional replay.
Research question
What do graph-based stock-prediction results establish once information timing, transaction costs, baselines and execution provenance are examined separately?
Work and approach
My contributions span research design, implementation, experiment execution, analysis and manuscript preparation, using AI tools throughout the workflow.
- Separate controlled comparisons from recovered historical outputs and conditional accounting replay.
- Audit input timing, transaction-cost units, passive baselines and graph-null comparisons.
- Preserve historical exhibits while recording implementation repairs and the interpretations they do not validate.
How to evaluate the work
The manuscript distinguishes what a controlled test demonstrates from what a retained output or constructed-clock replay can support. Reproducing accounting mechanics under stated assumptions does not establish actual first-feed availability or investable returns.
Evidence and limitations
The working manuscript is titled Auditing Timing, Costs, and Graph Value in Relational Stock Prediction. This page summarizes the audit and my contribution; it does not present the older dashboard as the reviewed manuscript.
Original execution bindings for the recovered historical summaries remain uncertain. The conditional replay is small and uses constructed clocks; it does not establish long-run alpha. Graph non-detections do not prove equivalence, and the audit does not establish a market-wide absence of predictive value.