Skip to content

Week 5 — Reading the Frontier I upcoming

July 20 – July 24, 2026 · Chapters 5.1 – 5.5 · Lab 5 · Mentor Session 5 · Problem Set 5

One paper per day. Each Reader's Guide follows a fixed 7-part anatomy: research question · identification strategy · data · table-by-table reading order · what's clever · what's vulnerable · three replication exercises.

This is the week where the program turns from "learn the tools" to "read the literature." By Friday you will have read five top-three-journal empirical-finance papers the way a referee reads them — tables first.

Daily schedule

Block What
9:00 – 12:00 Read Ch 5.1 — Fama & French (1992) Reader's Guide. The death of beta; size and book-to-market.
1:00 – 4:00 Work nb5.1 (FF92 portfolio sorts).
4:00 – 5:00 Office hours
Block What
9:00 – 12:00 Read Ch 5.2 — Fama & French (1993). The three-factor model; GRS test.
1:00 – 4:00 Lab 5 — Replicate FF93 using Ken French's free factor data (offline-safe synthetic fallback included).
4:00 – 5:00 Office hours
Block What
9:00 – 12:00 Read Ch 5.3 — Jegadeesh & Titman (1993). Momentum; overlapping portfolios. Work nb5.3 (momentum backtest + transaction-cost critique).
2:00 – 3:30 Mentor Session 5 — "How a paper actually gets written" (Prof. Gao walks through one of his own from blank page to publication).
4:00 – 5:00 Office hours
Block What
9:00 – 12:00 Read Ch 5.4 — Petersen (2009). Methods paper read as a paper. Work nb5.4 (clustering coverage taxonomy).
1:00 – 4:00 Read Ch 5.5 — Bertrand, Duflo & Mullainathan (2004). Placebo-law method. Work nb5.5 (naive rejection 0.35 vs. clustered 0.05).
4:00 – 5:00 Office hours
Block What
9:00 – 12:00 Wrap PS 5.1 – 5.5.
1:00 – 3:00 Capstone charter — first draft. Each team submits a 1-page charter (question, data, design, threats).
3:00 – 4:00 Friday demo + Week-6 preview
5:00 pm Deliverable cutoff

What to read

The papers themselves are on the Resources → Reading page with JSTOR / SSRN links.

Notebooks: notebooks/week-05/. Lab: Lab 5 — Replicate FF93. Mentor reading: Mentor 5 — "How a paper actually gets written".

What to submit (due Fri Jul 24, 5:00 pm ET)

  1. Problem Set 5 — five Reader's-Guide write-ups (one per paper, in the 7-part anatomy).
  2. Lab 5 notebook — FF93 replication; report your GRS statistic.
  3. Capstone charter — first draft in projects/<team-name>/charter.md.
  4. Journal entry — journal/week-05.md.
  5. Pull request — week-05 — <your name>.

Data we touch this week

  • CRSP, Compustat — for the FF92/FF93/JT93 replications.
  • Ken French Data Library (public, no auth) — for the canonical factor returns.

Next week preview

Week 6 — Reading the Frontier II →. Text-as-data — KPSS patents, Hoberg–Phillips TNIC, Loughran–McDonald sentiment, fair-lending in the algorithmic era — plus the AI co-pilot chapter.