RANDREW RESEARCHCONTROLLED EVIDENCE / V1.0
RESEARCH SYSTEM / ARG-R01 / LIMA — WORLDWIDE1 / 4

00 / CONTROLLED EVIDENCE / STRUCTURAL QUESTIONS

MODELS ARE EASY.VALIDATION IS THE HARD PART.

I study forecasting, counterfactual evaluation, simulation, and applied machine learning—especially where observational accuracy can conceal structural failure.

COUNTERFACTUALSFORECASTINGSIMULATIONMACROECONOMICS
RESEARCH THESIS / ARG-R01

The central research program: create environments where the quantity we care about is actually observable, then test whether our usual model-selection criteria deserve trust.

BASEBCRP + INDEPENDENTMETHODCONTROLLED ENVIRONMENTSOUTPUTPAPERS + SYSTEMSSTATUSACTIVE