Data Science in Finance Course

Lesson 1 • Probability and Statistics Refresher

Reviews distributions, moments, hypothesis testing, and correlation relevant to finance. Provides the statistical foundation required for risk and return modelling.

Lesson 2 • Data Sourcing and Financial APIs

Surveys market data providers, financial APIs, and alternative data sources. Students learn to programmatically retrieve and validate real financial data.

Lesson 3 • Financial Statements as Data Sources

Interprets balance sheets, income statements, and cash flow statements as structured datasets. Links accounting line items to quantitative signals used in analysis.

Lesson 4 • Core Financial Markets and Instruments

Covers equities, fixed income, derivatives, and FX markets as analytical objects. Establishes the financial vocabulary needed for all subsequent modelling work.

Lesson 1 • Financial Data Visualisation

Builds static and interactive charts for prices, returns, risk metrics, and portfolios. Effective visualisation communicates analytical findings to both technical and non-technical audiences.

Lesson 2 • Return Calculation and Normalisation

Derives simple, log, and excess returns and applies normalisation techniques. Standardised return series are the primary input for risk and factor models.

Lesson 3 • Exploratory Data Analysis for Finance

Applies EDA techniques to uncover distributional properties and relationships in financial data. Guides feature selection and model design decisions in later chapters.

Lesson 4 • Cleaning and Transforming Financial Data

Addresses missing values, outliers, survivorship bias, and look-ahead bias in financial datasets. Clean data pipelines prevent common errors in downstream modelling.

Lesson 1 • Time Series Models for Prices and Returns

Covers AR, MA, ARMA, and ARIMA models for financial time series forecasting. Students select and validate models using information criteria and residual diagnostics.

Lesson 2 • Linear Regression in Finance

Uses OLS regression to model asset returns against market and macro factors. Establishes the regression toolkit that underpins factor analysis and risk decomposition.

Lesson 3 • Factor Models and Risk Decomposition

Implements single-factor and multi-factor models to decompose return and risk. Factor exposures quantify systematic vs. idiosyncratic risk for portfolio construction.

Lesson 4 • Regularisation and Model Selection

Applies Ridge, Lasso, and Elastic Net to high-dimensional financial regression problems. Regularisation prevents overfitting in factor models with many candidate predictors.

Lesson 1 • Value at Risk and Expected Shortfall

Derives VaR and ES using historical, parametric, and Monte Carlo methods. These metrics are the primary regulatory and internal risk reporting tools in finance.

Lesson 2 • Regulatory Capital and Risk Frameworks

Surveys capital adequacy frameworks, internal model requirements, and risk reporting standards. Connects quantitative risk outputs to regulatory compliance and governance processes.

Lesson 3 • Stress Testing and Scenario Analysis

Designs historical and hypothetical stress scenarios to assess portfolio resilience. Scenario analysis complements VaR by capturing tail events beyond statistical models.

Lesson 4 • Credit Risk Quantification

Models probability of default, loss given default, and exposure at default. Credit risk metrics feed into loan pricing, capital allocation, and portfolio management.

Lesson 1 • Classification Models for Financial Events

Builds classifiers to predict credit default, earnings surprises, and market regimes. Addresses class imbalance and threshold selection specific to financial classification tasks.

Lesson 2 • Model Explainability and Validation

Uses SHAP values, permutation importance, and partial dependence plots to interpret ML models. Explainability is required for regulatory approval and stakeholder trust in finance.

Lesson 3 • Natural Language Processing for Finance

Extracts sentiment and signals from earnings calls, news, and filings using NLP techniques. Text-derived features augment quantitative models with qualitative information.

Lesson 4 • Supervised Learning for Return Prediction

Trains decision trees, random forests, and gradient boosting on financial features to predict returns. Covers feature engineering, hyperparameter tuning, and walk-forward validation.

Lesson 1 • Factor-Based Portfolio Construction

Constructs long-short factor portfolios using value, momentum, quality, and low-volatility signals. Factor tilts are combined and neutralised to isolate target exposures.

Lesson 2 • Mean-Variance Optimisation

Implements Markowitz mean-variance optimisation to construct efficient portfolios. Students derive the efficient frontier and understand the sensitivity of weights to input estimates.

Lesson 3 • Risk Parity and Alternative Weighting

Implements risk parity, equal risk contribution, and inverse volatility weighting schemes. These approaches diversify risk rather than capital, improving drawdown characteristics.

Lesson 4 • Portfolio Performance Attribution

Decomposes portfolio returns into allocation, selection, and factor contributions. Attribution analysis links portfolio outcomes to specific investment decisions and exposures.

Lesson 1 • Backtesting Framework Design

Builds an event-driven backtesting engine with realistic order handling and cost modelling. Proper framework design prevents look-ahead bias and overfitting to historical data.

Lesson 2 • Overfitting and Bias Detection

Identifies data snooping, multiple testing bias, and overfitting in backtested strategies. Deflated Sharpe ratio and combinatorial purged cross-validation are applied as remedies.

Lesson 3 • Execution and Market Microstructure

Covers order book dynamics, execution algorithms, and transaction cost analysis. Microstructure awareness is essential for translating backtested signals into live trading.

Lesson 4 • Trading Signal Generation

Constructs momentum, mean-reversion, and cross-sectional signals from price and fundamental data. Signal quality is evaluated using information coefficient and decay analysis.

Lesson 1 • Capstone Project and Deployment

Integrates all course skills into a full end-to-end financial data science project. Students present a deployed, documented solution demonstrating professional-grade competency.

Lesson 2 • Deep Learning for Financial Sequences

Applies LSTM, GRU, and Transformer architectures to financial time series and text data. Deep learning captures nonlinear temporal patterns beyond classical statistical models.

Lesson 3 • Cloud and Big Data Infrastructure

Deploys financial data pipelines on cloud platforms using distributed computing frameworks. Scalable infrastructure handles the volume and velocity of institutional financial data.

Lesson 4 • Alternative Data Integration

Processes satellite imagery, web scraping, credit card transactions, and social media signals. Alternative data provides informational edges when integrated with traditional financial data.