Empirical Finance from the Ground Up — ASSIP 2026 Cohort
Eight full-time weeks of research-grade empirical finance, built for the GMU ASSIP 2026 group
George Mason University · College of Science · Aspiring Scientists Summer Internship Program (ASSIP) 2026
Empirical Finance from the Ground Up
ASSIP 2026 Cohort · Eight full-time weeks · A research poster at the symposium on Aug 12, 2026
A textbook-and-lab system for the 14 advanced high-school and pre-graduation undergraduate researchers in the ASSIP 2026 Empirical Finance group — empirical finance the way it is actually done: with code, with data, and with intellectual honesty about what we can and cannot conclude.
Faculty mentor: Prof. Lei Gao, Associate Professor of Finance, Costello College of Business, George Mason University. 2026 ASSIP slot: REMOTE only. Program director: Dr. Amanda Haymond Still · cosassip@gmu.edu.
8 Weeks
40 Chapters
40 Notebooks
34 Data cards
5 Capstones
1 Instructor manual
Two interlinked sites
This project ships as two sites on a single GitHub Pages domain. Use the wiki for cohort logistics (calendars, mentors, templates, FAQ, symposium info); use the textbook for the chapters, problem sets, notebooks, capstones, and instructor’s manual.
Cohort wiki — /
Operational dashboard: week-by-week schedule, project templates (PAP, poster), mentor roster, code of conduct, FAQ, symposium logistics, reading list, data access, setup guide.
Textbook — /textbook/ (you are here)
The 8-week curriculum: probability and inference, OLS, causal inference I & II, paper-reading guides, text/AI, two project weeks, plus appendices A–E, the capstone gallery, and the instructor’s manual.
The 8-week arc — ASSIP 2026 calendar
Orientation is Thursday Jun 18, 2026. The cohort works Mon–Fri 9 am – 5 pm from Mon Jun 22 through the ASSIP Research Symposium on Wed Aug 12. Holidays off: Fri Jun 19 (Juneteenth) and Fri Jul 3 (Independence Day observance).
Week 1
Probability, Inference, and the Logic of Estimation
The two laws, expectations as geometry, sampling distributions, LLN/CLT, the mechanics of a t-test. Lab: simulating the coin-flip universe. Jun 22 – Jun 26, 2026.
Week 2
OLS, Geometry, and Inference
OLS in matrix form, Gauss-Markov, Frisch-Waugh-Lovell, robust and clustered standard errors, omitted-variable bias. Lab: Fama-MacBeth on CRSP. Jun 29 – Jul 2, 2026 (Jul 3 off).
Week 3
Causal Inference Foundations
Potential outcomes and SUTVA, matching and propensity scores, entropy balancing, IV and weak-IV pathology. Lab: making weak instruments lie. Jul 6 – Jul 10, 2026.
Week 4
Quasi-Experiments in Finance
Event studies, staggered DiD (TWFE vs. Callaway-Sant’Anna), regression discontinuity, synthetic control, Bartik shift-share. Lab: HMDA DiD. Jul 13 – Jul 17, 2026.
Week 5
Reading the Empirical Finance Canon
Reader’s guides to Fama-French 1992 & 1993, Jegadeesh-Titman 1993, Petersen 2009, Bertrand-Duflo-Mullainathan 2004. What does a JF paper actually do? Jul 20 – Jul 24, 2026.
Week 6
Text, AI, and Alternative Data
Patent value, 10-K similarity, Loughran-McDonald sentiment, fair-lending decomposition, and an AI copilot lab — without fooling yourself. Jul 27 – Jul 31, 2026.
Week 7
Your Project, Part I
Idea generation, data acquisition, pre-analysis plan, dataset construction, identification memo. The week you stop being a student. Aug 3 – Aug 7, 2026.
Week 8
Your Project, Part II — Symposium
Specification curve, robustness battery, poster build, peer review, defense. Poster session at the ASSIP Research Symposium, Wed Aug 12. Aug 10 – Aug 12, 2026.
Who this is for
This site is built for the ASSIP 2026 Empirical Finance cohort — a group of 14 advanced high-school students (15+) and pre-graduation undergraduates selected by GMU’s College of Science to work full-time, Monday through Friday 9 am – 5 pm, from June 22 through the ASSIP Research Symposium on August 12, 2026.
The group is at the high end of typical ASSIP group size (2–15), so we run a layered mentorship model: Prof. Gao (faculty mentor, remote slot) over the whole cohort, with PhD / RA sub-leaders working closely with sub-pods on data engineering, identification, and writing. See Cohort & Mentorship Model for how that scales for 14 students.
The methodological bar is real. You should be comfortable with strong precalculus or calculus, ready to learn Python quickly if you have not yet, and willing to read math that begins:
\hat{\boldsymbol{\beta}}_{\text{OLS}} = (\mathbf{X}'\mathbf{X})^{-1}\mathbf{X}'\mathbf{y}.
If that equation is intimidating but interesting, you are exactly the intended reader. Appendix A is a math toolkit; Appendix B is a setup guide; the prerequisite self-test will tell you which gaps to close before orientation on Jun 18.
What you’ll build — and what ASSIP requires
The required ASSIP 2026 deliverable is a research poster presented at the ASSIP Research Symposium on Wed Aug 12, 2026 (in-person poster session for on-site interns + virtual breakouts for remote interns — which is all of us, since our 2026 slot is remote). Beyond that, by the end of Week 8 you will have:
- A research poster for the Aug 12 symposium, audience includes Mason faculty, mentors, families, College of Science leadership, and Mason News press.
- A research abstract submitted to the Journal of Student-Scientists’ Research (JSSR, GMU Mason Publishing).
- An optional full paper if your project supports it.
- A public replication packet: code, data references, README, environment file, one-command rebuild.
- 3 GMU College of Science credits for completion, plus participation in the ASSIP scientific-writing and career forum.
- Reading fluency in five canonical empirical-finance papers and five more from the text/AI tradition.
How to use the book
For students
Read each week’s chapters in order. Work the problem sets before opening the solutions. Run every notebook. The labs are the part where you actually learn. Anchor every analysis decision to the identification memo you will write in Week 7.
For instructors & mentors
The instructor manual has the ASSIP-specific pacing guide, grading rubrics, common pitfalls, and answer keys. The capstone gallery gives five fully-specified projects with synthetic illustrative results.
For independent readers
You can run every notebook on Colab or Binder directly from each notebook page — look for the buttons at the top. The environment.yml pins the conda environment.
For everyone
Every chapter that makes a quantitative claim flags the claim with a [CHECK] tag if it has not yet been verified against a primary source. This is on purpose. Research is a process, not a posture.
Acknowledgements
This curriculum is built for the ASSIP 2026 (Aspiring Scientists Summer Internship Program) at George Mason University’s College of Science — director Dr. Amanda Haymond Still, program contact cosassip@gmu.edu. ASSIP makes research-grade methods accessible to high-school students and pre-graduation undergraduates who are ready for them, regardless of GPA, and offers full tuition waivers for Pell-eligible and reduced-lunch students. The articulation matrix maps every learning objective to the ASSIP symposium deliverable and to downstream undergraduate / PhD-level competencies, with sources cited.
The textbook was drafted in a multi-agent workflow that pairs planning, writing, and review agents under human supervision. See About for the methodology and the editorial standard.