Free module · Research & investment
Research & Backtest Design
Turn a hypothesis into a reproducible experiment, then try to disprove it.
data · inference · validation
The building blocks
Research begins with a falsifiable explanation of who might pay you and why. A chart or a profitable sample is a starting observation, not the explanation.
- Define a causal decision and label
- Build a chronological experiment
- Separate selection from final evaluation
Lessons in this module
- Build a point-in-time dataset
- Separate model selection from evaluation
- Account for dependence and multiple experiments
- Make the accounting identity your first test
Practice and apply
- Does the gross edge survive implementation? — Expected gross return is 18 bp per day. Total absolute traded weight is 0.80 of NAV per day. One-way costs are 12 bp per dollar traded. Daily borrow and funding cost 3 bp of NAV.
Work through the practice exercises · Quant development tools