Logo
← Back to archive

SWAT: Students Winning at Algo Trading

Archived

Competitive quant team building statistical and ML trading strategies for the CME, MIT, and Cornell challenges.

Finance
Ran
Fall 2022 – Spring 2025
Commitment
5-10 hrs/week

What you need coming in

Experience with Python and comfort with NumPy and Pandas. Trading APIs and underlying finance concepts were taught. STAT 350/511, AP Statistics, or similar strongly recommended; STAT/MA 416 and any AI/ML coursework a bonus.

Overview

The quant trading team competed in trading competitions run by universities and trading firms. Members learned to develop algorithmic trading strategies using statistics and machine learning. Existing knowledge of financial markets wasn’t required — that was covered before competitions. The goal was to send teams to the MIT, Cornell, and CME trading challenges, among others.

Member training

  1. Experience with Python, comfortable using NumPy and Pandas
  2. Relevant trading APIs were taught, starting with QuantConnect, along with underlying concepts in finance and trading
  3. Each competition differs slightly, so meeting focus was tailored to the upcoming competition

Difficulty — Intermediate

STAT 350/511, AP Statistics, or similar strongly recommended. STAT/MA 416 and any AI/ML coursework a bonus.

Competitions

  1. Onboarding (September 14–30)
  2. CME University Challenge (October 1–27)
  3. MIT Trading Challenge (October 28–29)
  4. Cornell Trading Competition (late October)

Time commitment

At least one meeting per week and 5+ hours per week, higher during competitions. Some competitions required travel. Each competition had a slightly different team, so members didn’t attend all of them. General strategy meetings ran throughout the semester for practice.