Curriculum
Nine missions in quantitative equity research: six core missions, each building on the last — from the basic overfitting problem through to real data and execution — plus three advanced electives on market microstructure and trading costs. All missions run in Google Colab. No local installation required.
Every mission comes in Python, R, and Julia — the same hidden-holdout engine scores all three, so the same idea earns the same out-of-sample result in any language.
6 + 3
Missions
10–15 hrs
Total time
Colab notebooks
Format
OOS Sharpe ratio
Assessment
The overfitting trap
Why in-sample performance is not evidence
Learning objectives
- 1.Explain the difference between in-sample and out-of-sample performance
- 2.Construct a cross-sectional signal from synthetic factor data
- 3.Measure information coefficient (IC) and its decay
- 4.Submit a strategy and interpret your OOS Sharpe score
Key concepts
Prerequisites
Basic Python and NumPy. No finance background required.
Reading & references
The survey is the “why” behind this mission: most published edges are in-sample artefacts that vanish out of sample.
The limit-order book
Market microstructure and adversarial trading
Learning objectives
- 1.Describe how a limit-order book operates
- 2.Implement a simple market-making agent
- 3.Measure adverse selection and inventory risk
- 4.Compete against other agents in the Arena
Key concepts
Prerequisites
Mission 1 or familiarity with financial returns.
Reading & references
These explain the matching engine and the spread/inventory trade-off before you put an agent on the live book.
Alpha discovery
Systematic search under the multiple-testing burden
Learning objectives
- 1.Apply walk-forward validation to avoid look-ahead bias
- 2.Control for the multiple-testing problem when scanning features
- 3.Measure signal decay and distinguish structural from incidental alpha
- 4.Build a composite signal from uncorrelated sub-signals
Key concepts
Prerequisites
Mission 1. Basic statistics (t-tests, correlation).
Reading & references
The replication-crisis survey covers the multiple-testing burden head-on: with hundreds of factors tested, a high t-stat is not enough.
Strategy library
Replication, combination, and the factor zoo
Learning objectives
- 1.Replicate canonical factor strategies (momentum, value, quality)
- 2.Measure pairwise correlation and diversification benefit
- 3.Combine strategies using equal weighting and minimum-variance
- 4.Diagnose over-fitting vs. genuine factor exposure
Key concepts
Prerequisites
Mission 3. Familiarity with the research library.
Reading & references
Replicate & run
Read the survey for each factor, then run its verified replication — these are the building blocks you combine in this mission.
Real data
From synthetic markets to live equity panels
Learning objectives
- 1.Fetch and clean a real equity panel using yfinance
- 2.Identify data quality issues (survivorship bias, stale prices)
- 3.Adapt synthetic strategies to real market conditions
- 4.Compare real vs. synthetic factor behaviour
Key concepts
Prerequisites
Missions 1–3. Recommended: review the anomaly tracker.
Reading & references
The size survey is a cautionary tale of how data choices (delisting, microcaps, survivorship) make or break a “real” result.
Advanced agents
Reinforcement learning meets market microstructure
Learning objectives
- 1.Implement a basic RL agent for order execution
- 2.Measure and minimise market impact
- 3.Combine a alpha signal with an execution layer
- 4.Evaluate end-to-end strategy performance
Key concepts
Prerequisites
Missions 1–5. Familiarity with basic RL concepts helpful.
Reading & references
Execution lives in the microstructure detail — read the L3 exchange page, then layer an execution policy on top of an alpha signal.
Advanced electives
Standalone deep-dives that extend the core arc. Take them in any order once you have the fundamentals — they go further into market microstructure and the economics of trading.
Queue dynamics
Trading the realistic exchange (L3)
Learning objectives
- 1.Explain FIFO queue priority and why queue position is the maker’s core asset
- 2.Simulate a resting limit order order-by-order: drain the queue, then fill
- 3.Model the cancel race against latency, and adverse selection
- 4.Observe queue dynamics on a real Bitstamp BTC/USD L3 feed
Key concepts
Prerequisites
Mission 2 (the limit-order book). Connects to the live L3 arena.
Reading & references
The L3 page explains queue position and the cancel race in plain English; the competition is where you trade a real order-by-order book.
The cost of trading
Turnover, transaction costs & capacity
Learning objectives
- 1.Quantify how transaction costs scale with turnover and erase paper alpha
- 2.Use rebalance frequency and no-trade bands to control turnover
- 3.Find a strategy’s break-even cost — where its edge disappears
- 4.Reason about capacity: why size itself moves the price against you
Key concepts
Prerequisites
Mission 3 or any strategy you want to pressure-test for real-world costs.
Reading & references
Replicate & run
The replication library reports gross vs net-of-cost OOS Sharpe; the high-turnover reversal strategy is the cleanest example of an edge that costs erase.
Pairs trading
Statistical arbitrage & the spurious-cointegration trap
Learning objectives
- 1.Distinguish correlation from cointegration — and why only the latter is tradeable
- 2.Test for cointegration (OLS hedge ratio + ADF / Engle–Granger) and form a spread
- 3.Trade the spread with a z-score entry/exit rule and evaluate it
- 4.See spurious cointegration: false equilibria that break out of sample
Key concepts
Prerequisites
Mission 1 (out-of-sample thinking). A first time-series strategy.
Reading & references
Read
Replicate & run
Run the canonical Gatev et al. distance pairs replication, then read the reversal survey for the mean-reversion economics behind the spread.
Using this as a course
Each mission is designed to stand alone as a 1–2 hour lab session. For a semester course, Missions 1–3 work well as the first half with Missions 4–6 as the research project phase. Classroom cohorts give students a private leaderboard graded on OOS Sharpe — not in-sample performance. Each mission above carries its own Reading & references — the topic surveys and replications it draws on — to assign alongside the notebook.