tracks
Three phases, beginner to advanced. Work top to bottom — each track builds on the one before it. Open a track to see its skills in the sidebar.
Beginner
7 tracksStocks & Prices
What a stock is, how its price is recorded (open, high, low, close, volume), and how to load and plot one over time.
Returns
Turn prices into returns — daily, multi-day, and annual — and learn the difference between simple and log returns.
Risk & Volatility
Measure how much a stock swings: variance, standard deviation, volatility, drawdown, and the worst peak-to-trough losses.
Distributions
The shapes returns and prices actually take: the normal baseline every risk model assumes, the lognormal behind prices and options, the fat tails and skew that break the tidy models, and the tools (Student's t, the central limit theorem, bootstrapping) for handling them honestly.
Portfolios
Combine stocks into a portfolio: covariance and correlation, weights, weighted returns, concentration, and the free lunch of diversification.
Risk-Adjusted Returns
Compare strategies fairly: the risk-free rate, excess return, and the Sharpe, Sortino, Calmar and Information ratios.
Vectors & Linear Algebra
Scale portfolio math to many assets with vectors, NumPy, the dot product, and the covariance matrix.
Core Quant
7 tracksFactors: Alpha & Beta
Split a return into market risk (beta) and skill (alpha), then meet the classic factors: value, quality, growth, and low-vol.
Signal Engineering
Turn raw prices into tradable signals: align data without lookahead, standardize with z-scores, rank a universe, measure predictive power with the information coefficient, and neutralize unwanted exposures.
Time Series
The time dimension of returns: stationarity, autocorrelation, and simple autoregressive forecasting, the statistical backbone of mean-reversion and momentum strategies.
Strategies
Build and backtest the classic alpha phenomena: momentum, mean reversion, value, quality, volatility, growth, and sentiment.
Risk Management
Measure and model risk: factor models, VaR and expected shortfall, stress testing, tail risk, and hedging.
Monte Carlo Simulation
Turn one history into thousands of possible futures: simulate price paths, estimate VaR and expected shortfall by simulation, and bootstrap a strategy's Sharpe to tell skill from luck.
Portfolio Construction
Turn signals into positions: optimization, risk parity, rebalancing, concentration limits, and controlling style drift.
Advanced
7 tracksExecution & Trading Costs
The gap between paper and live returns: market impact, order splitting, VWAP/TWAP, and implementation shortfall.
Research & Validation
Tell real edges from statistical mirages: backtesting pitfalls, overfitting, multiple testing, and clean data.
Frontier Topics
Beyond the core toolkit: machine learning for alpha, alternative data, derivatives, rates, and regime detection.
Machine Learning for Alpha
Apply machine learning to generate alpha, built from scratch in NumPy: linear regression as a return predictor, feature matrices, PCA denoising, walk-forward validation, regularization, and a full ML long/short backtest.
Bayesian Methods
Reason about markets with probability distributions, not point estimates: Bayesian updating, shrinkage of noisy estimates, Bayesian regression, and credible intervals for judging whether an edge is real.
Non-linear Transformations
Reveal signal that linear methods miss: where correlation fails but mutual information sees, the rank transform, log and power transforms for skewed features, and quantile buckets and interactions that lift a backtest.
Alternative Data
Turn non-price data into alpha: using fundamentals point-in-time without lookahead, building value and quality factors, and testing a news-sentiment signal, each judged by its information coefficient and a backtest.