RESEARCH DESK / THREE ORIGINAL CASES
A good result deserves
a difficult question.
Before defending a backtest, examine when information arrived, how the model was chosen and what trading would cost.
Allow about 15 minutes. No Python setup or account needed. The cases are educational; they do not recommend securities or trading strategies.
SYNTHETIC RESEARCH RECORD
A feature that arrives too late
A synthetic study predicts tomorrow's return using today's closing price and a daily indicator. The indicator is stamped with today's date but is not released until tomorrow at 10:00. The model trades today at the close.
0 of 3 explanations reviewed. This is practice progress, not a hiring score.
Inspect the full Quant Research Project LabPrefer interview probability? Open the eight-problem diagnosticMethod reference: scikit-learn on data leakage and keeping model choices separate from test data. Cases and cost assumptions are original teaching examples.