Paul F. Mende MIT Sloan School of Management Fall 2014 Course Syllabus 15.496 Data Technologies for Quantitative Finance Course Description. This course introduces students to financial market data and to data architecture & design, with applications to financial asset pricing, quantitative investment strategies, algorithmic trading, portfolio management, and risk management. Students will use MATLAB, SQL, Bloomberg, Access and other tools to develop data-driven analysis and applications. To understand modern financial markets, it is essential to understand market data: where it comes from, what it conveys (as well as what it hides), and how to analyze it in diverse contexts such as trading, market regulation, market microstructure, portfolio management, risk management, corporate finance, efficient markets, and adaptive markets. Financial data sets are often enormous in size and complexity, and many problems require efficiently merging, filtering, and slicing multiple sources at once. In this course we ll look at this data up close. Students will learn how data relationships are structured and how to use modern tools and technologies to manipulate, manage, and analyze it. Projects will be able to draw from extensive commercial and research data sources via a dedicated high-performance computing platform. At the end of the course students will have practical knowledge and experience that will be of value in other courses at Sloan and in industry as well. The course will use real-world data, applications, and cases to illustrate the principles and provide hands-on experience. Finance is an empirical subject, and possessing a practical mastery of financial market data will give you an edge. Pre-requisites and complements. This course is geared toward M.Fin. students. Others may enroll with the pre-requisite of 15.401, or its equivalent, and the educational background of students in the M.Fin. program. As this course provides tools and techniques for real-world applications of finance theory, it is highly complementary to 15.433, 15.437, and 15.450; and it is recommended for 15.460, Applied Quantitative Finance (formerly Analytics of Finance II ). Some programming skills will be necessary along with familiarity with Excel. Projects will make use of Matlab and SQL, although prior experience is not required. Students who have not yet had experience with Matlab can complete a tutorial at the start of the course. Course Requirements and Grading. Course requirements include: regular attendance and class preparation/participation in lectures and recitations (10 percent) and project/problem sets (90 percent). The major assignments are tentatively planned as follows: Project Topic Date posted Date due A Data preliminaries September 9 September 23 B Equity long/short trading September 23 October 7 C Portfolio and risk management October 7 November 4 D Options and volatility November 4 November 25 E High-frequency trading November 25 December 9 15.496 2014 by Paul F. Mende Page 1 of 5
Course Materials. Lecture notes, handouts, projects, and solution sets will be posted on Stellar. Data sets for problems, projects, and research will be available on the class database server, obelix.mit.edu. There is no textbook assigned for the course. References to readings or to other texts will be posted to Stellar during the course, as will pointers to software, tutorials, etc. Class Preparation and Participation. Class preparation and participation are important components of this course. This course is hands on, so students should be ready to reproduce the material and cases presented in lecture on their own. Students are expected to come to each class well prepared to discuss their results from previous assignments and projects. Computing Resources. Computing and data resources will be introduced in lecture and in recitation. Students can make use of the new virtual lab (which has replaced the former Sloan trading room), the Bloomberg terminals, and resources on the MIT network, along with MIT s access to multiple data vendors. A dedicated set of data servers and a data warehouse are available for class projects, and students are encouraged to familiarize themselves with these tools: Excel, Access, Matlab (all available from IS&T). The more adventurous are encouraged to download Microsoft s free SQL Server Express for its query and data management tools. Recitation. A weekly recitation session will be held in E51-151 on Fridays 9-10:30am. 15.496 List of Topics The course will draw from the list of topics below. Note that not all will be covered. 1. Mathematical, Financial, and Technical Preliminaries a. Course overview, and the role of data in quantitative finance b. Mathematics and statistics in financial analysis c. Brief tour of stochastic processes, financial time series, and portfolio theory d. Data technologies past, present, and future e. Data and computational architecture of a quantitative strategy f. From risk measurement to risk management 2. Sources of Financial Data a. Markets and exchanges b. Corporate data c. Financial analysts d. Time series data e. Asset attributes and corporate actions f. Economic data g. Data vendors and data delivery 15.496 2014 by Paul F. Mende Page 2 of 5
h. Time scales, scaling, and irregularly spaced data i. Metadata and meta-metadata j. Slowly changing data k. Data integrity; error detection and correction; audit trails 3. Data Architecture and Design a. Logical and physical architectures b. Transactions, operations, decision support c. Entity-relationship modeling d. Dimensional design e. Data design for financial analytics f. Integrating data from multiple sources; security masters g. Data domains: market, model, portfolio 4. Data Management Tools and Technologies a. Scale, performance, and dealing with massive data sets b. Single-user systems: Excel, Access, c. Client-server systems: RDBMS d. Multi-dimensional databases, OLAP, business intelligence e. Big Data tools, technologies, and approaches f. Data extraction, transformation, and loading g. Query design and optimization h. Data visualization tools 5. Varieties of Time Series Data a. Regularly and irregularly spaced data b. Analytics and scaling c. Prices d. Returns e. Trades f. Quotes g. Volumes h. Derived and calculated data i. Dealing with redundancy and correction j. Time series of windowed data (e.g., volatility) 6. Applications to Backtest Simulations a. Definitions of backtesting and forward testing b. Look-ahead, overfitting, and other data-snooping biases c. Bayesian decision theory and sequential analysis d. Monte Carlo simulation e. Data design and infrastructure for backtesting 15.496 2014 by Paul F. Mende Page 3 of 5
7. Applications a. Portfolio management b. Algorithmic trading c. Quantitative equity strategies d. Option hedging e. Index arbitrage f. Convertible arbitrage g. Risk management and risk aggregation h. Performance attribution and factor analysis i. Volatility surfaces j. Volatility modeling and trading k. Dispersion 8. Digital Dashboards a. Data visualization b. Taxonomy of graphs for quantitative finance c. Static and batch reporting d. Dynamic graphs and tables e. Pivot tables and pivot charts f. Systems for real-time trading 9. Data Aggregation and Data Slicing a. Factor exposures b. Risk management c. Performance attribution d. Risk factors e. Industry membership f. Asset class 15.496 2014 by Paul F. Mende Page 4 of 5
Course Schedule Note: This is an approximate schedule for the course and meant to serve as a rough guide to the pace of the course; some material may take more or less than the allotted time. Session Date Topics Applications & Cases Assignments 1 September 9 Introduction and overview; Mathematical and financial preliminaries; Role of market data in financial theory and practice. Law of one price. Sources of financial data. 2 September 16 Data architecture and design, Part I; Introduction to algorithmic trading and backtesting. 3 September 23 Data architecture and design, Part II; Data management tools and technologies 4 September 30 Varieties of time series data; Applications to backtest simulations and strategy development 5 October 7 Communicating with data; Designs for data visualization; Designs for decision support. 6 October 15 Data aggregation; Multidimensional data design Market regimes, jumps and volatility; rational and irrational investors; equity trading. Mean-reversion, inefficient markets; equity long/short investment strategies. Index arbitrage, correlation and hedging Option hedging; highfrequency trading Risk management; portfolio management; performance attribution Hedge funds; funds of funds. Project A posted Project A due; Project B posted Project B due; Project C posted October 21 No class SIP 7 October 28 Options; Data design for derivatives and related instruments 8 November 4 Volatility surfaces; Data design for complex portfolios and nonlinear risks; Designs for stress scenarios Delta hedging in theory and practice Relative value, carry, and correlation trades. Project C due; Project D posted November 11 No class 9 November 18 Market microstructure Limit order book; FX 10 November 25 Introduction to high-frequency finance; trade messaging; real-time architectures Market making; market taking; liquidity metrics Project D due; Project E posted 11 December 2 High-frequency trading High-frequency strategies; flash crash. 12 December 9 Review, synthesis, future directions Project E due 15.496 2014 by Paul F. Mende Page 5 of 5