This course will survey the field of quantitative investment strategies from a "buy side" perspective, through the eyes of portfolio managers, analysts, and investors. Financial modeling often involves avoiding complexity in favor of simplicity and practical compromise. The "buy side" of the marketplace is dominated less by highly rigorous mathematics or miraculous discoveries, and more by a mix of analytical and financial understanding, computation, sensible risk management, and a general sense of humbleness in search for an "edge" in investing and performance at reasonable risk/reward levels.
The purpose of this course is to give students direct exposure to those problems facing "buy side" quantitative analysts on Wall Street. In this practically oriented course, we combine all necessary material scattered in finance, computer science and statistics into a project-based curriculum, which give students hands-on experience to solve real world problems in portfolio management.
Students will work with market and historical data to develop and test trading and risk management strategies with an eye towards the practical considerations of financial data modeling, as well as real-world considerations and limitations.
In this course, we will wear the hat of a buy-side quant (e.g. someone hired by a mutual fund or a hedge fund, etc…), whose responsibility is to develop two profitable quantitative investment strategies over the course of 3 months (length of the course). Programming projects are required to complete this course.
Discussion of appropriate topics in Data Aggregation, Data Analysis and Data Mining in search for the investment "edge". Emphasis on identification and mitigation of hidden biases (e.g. selection bias) in analyzing data. We will collect and analyze fair amounts of historical data.
We will employ a variety of tools, including back-tests, in-sample/out-sample comparisons and Monte Carlo analysis, to study a strategy’s robustness and sensitivity to a given choice of parameters.
Benchmarking, Attribution and Performance Analysis
Discussion of appropriate topics in style and performance analysis and general evaluation of strategy performance against objectives and benchmarks. We will also discuss appropriateness of various reward/risk measures (e.g. Sharpe Ratio and other statistics) as they pertain to specific strategies and objectives.
Transaction Cost Model
We will develop practical transaction cost models based on evaluation of the chosen markets. We will also discuss how to incorporate various practical limitations into a proper model.
Portfolio Construction and Optimization
Given a perceived edge and trading approach, we will study appropriate mechanisms for constructing and managing the portfolio over time.
Analysis of Leverage
With leverage (explicit or implicit) being a common source of actual (or hidden) risk, we will focus on analyzing hidden leverage due to potential non-linearities in portfolio relationships. This is especially relevant for any hedge-fund types of strategies or strategies employing futures or other "non-standard" instruments.
We will survey VAR-based and other approaches to risk management. As with the whole course, the emphasis will be from a "buy"-side perspective, i.e. we will not be concerned with risk from the point of view of fraud, business risks, etc... rather we will be concerned as investment managers to manage the risk of the portfolio on a daily basis.
Grading emphasis will be placed on the completeness and rigor of the student’s approach. During the length of the course, several industry speakers will be invited to speak on various issues in quantitative portfolio management.
Dr. Alex Greyserman is currently principal in a private hedge-fund investment partnership. Over the past 5 years, he was Chief Investment Officer a 500mil+ hedge fund, with responsibility for research, development, and implementation of all trading strategies. Dr. Greyserman holds a MS in Electrical Engineering from Columbia University and a Ph.D. in Statistics from Rutgers University.
Sufficient programming skills to carry out data-intensive analysis are required. Possible programming languages include C/C++/Perl/Matlab, and others. Given sufficient proficiency, it may be possible to conduct the analysis in Excel.
Samplings from the following books will be provided for valuable reading materials:
Schwager, J. Market Wizards (1990)
Schwager, J. New Market Wizards (1994)
Schwager, J. Stock Market Wizards (2000)
O’Shaughnessy, James P., What Works on Wall Street, (McGraw Hill, 1996)
Bernstein, Peter L. Against the Gods: The Remarkable Story of Risk, (Wiley, 1996)
Bernstein, Peter L. Capital Ideas (Free Press Paperback, 1993)
Fabozzi, Frank Investment Management (Prentice Hall, 1995)
Haugen, Robert A. The Inefficient Stock Market (Prentice Hall, 1998)
Haugen, Robert A. Beast on Wall Street (Prentice Hall, 1999)
Sharpe, William F., Gordon J. Alexander, Investments (Prentice Hall, 1999)
Taggart, Robert A., Quantitative Analysis for Investment Management (Prentice Hall, 1996)