DATATOMIC / RESEARCH DESIGN NOTE
Financial stress without hindsight
Turning the seismic analogy into a question that can fail a test.
EXPLORATORY NOTE · PROPOSED INQUIRY
An introductory framing of the question and possible next steps. No completed study or experimental result is reported.
An analogy is a starting point
The seismic idea asks whether accumulated stress, transmission through a network, and cascading failure offer a useful way to describe financial instability. The difficult step is translating those words into observable quantities before examining a crisis. If every past event can be given a convincing story afterward, the model has not yet earned predictive value.
Specify the proposed test
A proposed protocol would choose a data universe and forecast horizon, define a stress measure, and record a rule for issuing an alert. Evaluation would move forward through time, using only information available at each date. Conventional volatility and simple risk measures would provide comparison baselines.
Count the uncomfortable cases
The evaluation should report false alarms, missed events, lead time, and performance across different market conditions. The stress measure should be tested under small changes to thresholds and data choices. If useful results depend on one hand-picked period, the result would need further investigation.
A falsifiable next step
The next proposed artifact is an evaluation specification that can be reviewed before results exist. This note describes research design, not a trading signal, completed backtest, or demonstrated forecasting capability.
CONTINUE THE INQUIRY
Read the research direction