1st Edition
Algorithmic Trading and Quantitative Strategies
Algorithmic Trading and Quantitative Strategies provides an in-depth overview of this growing field with a unique mix of quantitative rigor and practitioner’s hands-on experience. The focus on empirical modeling and practical know-how makes this book a valuable resource for students and professionals.
The book starts with the often overlooked context of why and how we trade via a detailed introduction to market structure and quantitative microstructure models. The authors then present the necessary quantitative toolbox including more advanced machine learning models needed to successfully operate in the field. They next discuss the subject of quantitative trading, alpha generation, active portfolio management and more recent topics like news and sentiment analytics. The last main topic of execution algorithms is covered in detail with emphasis on the state of the field and critical topics including the elusive concept of market impact. The book concludes with a discussion of the technology infrastructure necessary to implement algorithmic strategies in large-scale production settings.
A GitHub repository includes data sets and explanatory/exercise Jupyter notebooks. The exercises involve adding the correct code to solve the particular analysis/problem.
I Introduction to Trading
1. Trading Fundamentals
A Brief History of Stock Trading
Market Structure and Trading Venues: A Review
Equity Markets Participants
Watering Holes of Equity Markets
The Mechanics of Trading
How Double Auction Markets Work
The Open Auction
Continuous Trading
The Closing Auction
Taxonomy of Data Used in Algorithmic Trading
Reference Data
Market Data
Market Data Derived Statistics
Fundamental Data and Other Datasets
Market Microstructure: Economic Fundamentals of Trading
Liquidity and Market Making
II Foundations: Basic Models and Empirics
2. Univariate Time Series Models
Trades and Quotes Data and their Aggregation: From Point Processes to Discrete Time Series
Trading Decisions as Short-Term Forecast Decisions
Stochastic Processes: Some Properties
Some Descriptive Tools and their Properties
Time Series Models for Aggregated Data: Modeling the Mean
Key Steps for Model Building
Testing for Nonstationary (Unit Root) in ARIMA Models: To Difference or Not To
Forecasting for ARIMA Processes
Stylized Models for Asset Returns
Time Series Models for Aggregated Data: Modeling the Variance
Stylized Models for Variance of Asset Returns
Exercises
3. Multivariate Time Series Models
Multivariate Regression
Dimension-Reduction Methods
Multiple Time Series Modeling
Co-integration, Co-movement and Commonality in Multiple Time Series
Applications in Finance
Multivariate GARCH Models
Illustrative Examples
Exercises
4. Advanced Topics
State-Space Modeling
Regime Switching and Change-Point Models
A Model for Volume-Volatility Relationship
Models for Point Processes
Stylized Models for High Frequency Financial Data
Models for Multiple Assets: High Frequency Context
Analysis of Time Aggregated Data
Realized Volatility and Econometric Models
Volatility and Price Bar Data
Analytics from Machine Learning Literature
Neural Networks
Reinforcement Learning
Multiple Indicators and Boosting Methods
Exercises
III Trading Algorithms
5. Statistical Trading Strategies and Back-Testing
Introduction to Trading Strategies: Origin and History
Evaluation of Strategies: Various Measures
Trading Rules for Time Aggregated Data
Filter Rules
Moving Average Variants and Oscillators
Patterns Discovery via Non-Parametric Smoothing Methods
A Decomposition Algorithm
Fair Value Models
Back-Testing and Data Snooping: In-Sample and Out-of-Sample Performance
Evaluation
Pairs Trading
Distance-Based Algorithms
Co-Integration
Some General Comments
Practical Considerations
Cross-Sectional Momentum Strategies
Extraneous Signals: Trading Volume, Volatility, etc
Filter Rules Based on Return and Volume
An Illustrative Example
Trading in Multiple Markets
Other Topics: Trade Size, etc
Machine Learning Methods in Trading
Exercises
6. Dynamic Portfolio Management and Trading Strategies
Introduction to Modern Portfolio Theory
Mean-Variance Portfolio Theory
Multifactor Models
Tests Related to CAPM and APT
An Illustrative Example
Implications for Investing
Statistical Underpinnings
Portfolio Allocation Using Regularization
Portfolio Strategies: Some General Findings
Dynamic Portfolio Selection
Portfolio Tracking and Rebalancing
Transaction Costs, Shorting and Liquidity Constraints
Portfolio Trading Strategies
Exercises
7. News Analytics: From Market Attention and Sentiment to Trading
Introduction to News Analytics: Behavioral Finance and Investor
Cognitive Biases
Automated News Analysis and Market Sentiment
News Analytics and Applications to Trading
Discussion / Future of Social Media and News in Algorithmic Trading
IV Execution Algorithms
8. Modeling Trade Data
Normalizing Analytics
Order Size Normalization: ADV
Time-Scale Normalization: Characteristic Time
Intraday Return Normalization: Mid-Quote Volatility
Other Microstructure Normalization
Intraday Normalization: Profiles
Remainder (of the Day) Volume
Auctions Volume
Microstructure Signals
Limit Order Book (LOB): Studying Its Dynamics
LOB Construction and Key Descriptives
Modeling LOB Dynamics
Models Based on Hawkes Process
Models for Hidden Liquidity
Modeling LOB: Some Concluding Thoughts
9. Market Impact Models
Introduction
What is Market Impact
Modeling Transaction Costs
Historical Review of Market Impact Research
Some Stylized Models
Price Impact in the High Frequency Setting
Models Based on LOB
Empirical Estimation of Transaction Costs
Review of Select Empirical Studies
10. Execution Strategies
Execution Benchmarks: Practitioner’s View
Evolution of Execution Strategies
Layers of an Execution Strategy
Scheduling Layer
Order Placement
Order Routing
Formal Description of Some Execution Models
First Generation Algorithms
Second Generation Algorithms
Multiple Exchanges: Smart Order Routing Algorithm
Execution Algorithms for Multiple Assets
Extending the Algorithms to Other Asset Classes
V Technology Considerations
11. The Technology Stack
From Client Instruction to Trade Reconciliation
Algorithmic Trading Infrastructure
HFT Infrastructure
ATS Infrastructure
Regulatory Considerations
Matching Engine
Client Tiering and other Rules
12. The Research Stack
Data Infrastructure
Calibration Infrastructure
Simulation Environment
TCA Environment
Conclusion
Biography
Raja Velu is a professor of Finance and Analytics in Whitman School of Management at Syracuse University. He served as a Technical Architect at Yahoo! in the Sponsored Search Division and was a visiting scientist at IBM-Almaden, Microsoft Research, Google and JPMC. He has also held visiting positions at Stanford's Statistics department, Indian School of Business, the National University of Singapore, and Singapore Management University.
Maxence Hardy is a Managing Director and the Head of eTrading Quantitative Research for Equities and Futures at J.P.Morgan, based in New York. Mr. Hardy is responsible for the development of agency algorithmic trading strategies for the Equities and Futures divisions globally.
Daniel Nehren is a Managing Director and the Head of Statistical Modelling and Development for Equities at Barclays. Based in New York, Mr. Nehren is responsible for the development of algorithmic trading and analytics products. Mr. Nehren has more than 19 years of experience in equity trading working for some of the most prestigious financial firms including Citadel, J.P Morgan, and Goldman Sachs.
"This work does a marvelous job of emphasizing the dual significance of determining the fair value of an asset as well as designing the optimal way to interact with the markets. Optimizing valuation is equally important to optimizing order execution. Both skills must be mastered to avoid selection bias and capturing value. This book must be read!"
-Peter J. Layton, Principal, Blackthorne Capital Management, LLC"An outstanding and timely synthesis of the state of art algorithmic trading ideas. I will recommend it to all who is serious on the foundations."
-Guofu Zhou, Frederick Bierman and James E. Spears, Professor of Finance, Olin Business School, Washington University in St. Louis"This book is a rigorous yet practical introduction to the subject and takes the reader to some advanced concepts in quantitative algo trading. [...] What really makes the book stand out in terms of the learning experience for the interested reader is its no-nonsense attitude to hands-on-learning. The authors have the ethos that algo trading is not a spectator sport and that the best way to learn is to get their hands dirty and find out for themselves. In this spirit, each section contains a set of practical examples and exercises with data available in the corresponding GitHub repository. [...]."
-Gordon Lee, in Quantitative Finance, January 2023"I enjoyed reading this excellent book. I strongly recommend this book to mathematical and applied statisticians generally and to portfolio analysts in particular."
-Ramalingam Shanmugam, in the Journal of Statistical Computation and Simulation, June 2023