Finance Project — Asset Management

Institute of Strategic Management and Finance, Ulm University

Authors
Affiliation

Prof. Dr. Andre Guettler

Institute of Strategic Management and Finance, Ulm University

Oliver Padmaperuma

Institute of Strategic Management and Finance, Ulm University

About this course

A hands-on Master’s-level project course on building empirical asset-management research pipelines in R and applying them to a non-traditional asset class — prediction markets, using the Polymarket Quant Bench dataset (curated OHLCV bars over Jon Becker’s polymarket-data dump). Across five lectures and a final-presentations session, students master the core ML toolbox (linear models, Ridge, Lasso, Elastic Net, cross-validation), then design their own indicator library, derive trading signals, back-test a strategy, and present results in groups of three.

Course at a glance

Term

Summer 2026
Level

Master
ECTS

6
Language

English
When

Wednesdays 12:15–13:45
Where

Helmholtzstraße 18, room E60, Ulm
Format 5 lectures + final-presentation session, group project (3 students)
Assessment Project (50%) + Final presentation (50%)
Sign-up Course Moodle page by 15 April 2026

Learning outcomes

By the end of the course, students will be able to:

  1. Design and execute a full empirical workflow in R: load, clean, model, back-test, report.
  2. Apply Ridge, Lasso, and Elastic Net regularisation, and pick hyper-parameters via cross-validation.
  3. Engineer technical / statistical / external-data indicators for prediction-market price time-series.
  4. Implement a simple back-test honestly (with transaction costs) and interpret performance metrics.
  5. Communicate empirical findings in writing (R Markdown → PDF) and orally (20-minute presentation).

Required materials & setup

Bring a laptop to every session — Lecture 2 onwards is hands-on. Install the following before Lecture 2:

  • R (≥ 4.3) — https://cran.r-project.org/

  • RStudio Desktop — https://posit.co/download/rstudio-desktop

  • Reference text — An Introduction to Statistical Learning with Applications in R [@james2021introduction]; free PDF and exercises at https://www.statlearning.com.

  • R packages (install once, load per project):

    Bucket Packages
    Data access arrow (after the one-time hf download … — see Lecture 5)
    Data wrangling tidyverse, lubridate, data.table
    Time-series core xts, zoo, tsibble, slider
    Indicators TTR, quantmod, tidyquant
    ML / regularisation glmnet, caret or tidymodels
    Back-test / metrics PerformanceAnalytics
    Reporting rmarkdown, knitr, kableExtra, ggplot2, patchwork

The Polymarket Quant Bench dataset is public on HuggingFace (CC-BY-4.0). One-time setup, in a terminal:

pip install huggingface_hub
hf download smf-ulm/polymarket-quant-bench \
    --repo-type dataset --local-dir data/   # repo's polymarket/ lands here

~603 MB on disk inside your project’s data/ folder. Pin --revision <sha> to a specific HF commit before you submit (so the marker re-runs against the same snapshot). Add data/ to .gitignore. After the download, read in R with arrow::open_dataset() (see Lecture 5).

Schedule

Assessment & deadlines

  • Project (Code + Report) — 50% of the final grade. Rmd code + knitr-rendered PDF report (10–15 pages). Due 30 June 2026, 18:00. See the project brief.
  • Final Presentation — 50%. 20-minute group presentation in class on 1 July 2026; submit slides as PDF inside the same project zip. See the presentation brief.
  • Group size: 3 students (we allocate if you don’t form one).
  • Submission pattern: zip and email subject Asset2026_surname1_surname2_surname3, sent to oliver.padmaperuma@uni-ulm.de with andre.guettler@uni-ulm.de and your team-mates in CC.

Instructors

Access & contact