Oil forecasts: which models work
Compare statistical and machine-learning forecasting models for oil-market instruments and find out which ones hold up out of sample.
Research topic: Comparative Analysis of Forecasting Models for the Oil Market
- Length
- 10 weeks
- Group size
- 3–6 people
- Level
- Intermediate
- Tooling
- Python
About the project
Part of the Financial markets, quantitative modelling and trading strategies track. The team states a hypothesis, builds a model or strategy and tests it on historical data.
Topics are limited to those for which the mentor can provide data or suitable public data exists. The outcome is reproducible code, a report and a presentation you can show to an employer or investor.
What you’ll do
Weeks 1–2
Literature review and data collection
Weeks 3–5
Building the model or strategy
Weeks 6–8
Backtesting and robustness checks
Weeks 9–10
Conclusions and final presentation
Open materials
- Reading listBooks & materials(opens on an external site)
- Intro lectureVideo(opens on an external site)
- Starter repoCode(opens on an external site)
Lectures, data and code for enrolled groups unlock in the student area once the project starts.
Instructor

Lidia B.
Financial markets & risk models · HSE University
An investment-bank practitioner who builds interest-rate, fair-value and credit-risk models. Teaches derivatives to master's and MBA students at HSE University.
Following this project
Enrol
The group has to fill by 5 January
Open group tariff
₽24,000per person
Your fee may be refunded by a partner investor if the results are strong.
₽24,000Starts 12 January
Enrol