Weekly lecture notes and Google Colab notebooks, posted as the course progresses. See the full syllabus for grading, readings, and course policies.
Register here for a WRDS account. You will need it for your project. Use a UMD account.
New to Google Colab? Read the Colab primer before running the weekly notebooks.
Course Structure
The course is organized around four questions rather than a chapter sequence:
- 1. Is it a good strategy? (sessions 1–6) — market efficiency, factor models, data mining and research protocol, luck vs. skill, shrinkage with many overlapping signals, and the factor zoo at the research frontier (IPCA and its critique).
- 2. Executing cheaply (sessions 7–8) — portfolio construction and risk models, trading costs, shorting, and capacity.
- 3. Executing with an ETF (session 9) — the active vs. passive debate.
- 4. Macro implications (sessions 10–12) — inelastic markets and price efficiency, then crowding and fragility: the August 2007 quant unwind, ETF liquidity mismatch (August 2015, bond ETFs in March 2020), and herding from common signals, including AI-driven trading.
Schedule
| Week | Dates | Theme |
|---|---|---|
| Week 1 | Oct 19 & 21 | Is it a good strategy? Market efficiency; basic models & factor choice |
| Week 2 | Oct 26 & 28 | Fundamental factor models; data mining & research protocol |
| Week 3 | Nov 2 & 4 | Case #1: GMO; ridge and shrinkage; are characteristics covariances? (IPCA vs. Kozak & Nagel) |
| Week 4 | Nov 9 & 11 | Executing cheaply: portfolio construction (Case #2: Martingale); trading costs & capacity |
| Week 5 | Nov 16 & 18 | Executing with an ETF: active vs. passive; inelastic markets & price efficiency |
| No class Nov 23 & 25 — Thanksgiving recess / university holiday | ||
| Week 6 | Nov 30 & Dec 2 | Macro implications: crowding & fragility — historical episodes, then AI-driven herding |
| Week 7 | Dec 7 & 9 | Team presentations of backtesting projects |
| Term B ends Dec 10 | ||