Dates and Locations

26 & 27 Oct '10	New York, USA	Millenium Hilton Hotel >>
25 & 26 Nov '10	London, UK	Apex City of London Hotel >>
18 & 19 Jan '11	Hong Kong	Sheraton Hong Kong Hotel and Towers >>

Please visit our Course Diary to see dates and locations for all of our UK and overseas courses.

Course Outline: Pair Trading & Related Strategies

This two day workshop introduces delegates to pairs trading and related strategies, with particular reference to research and implementation using MATLAB.

Overview
- Stationarity, cointegration, and mean reversion
- Pair trading in stocks, ETF's, futures, and currencies
- Beyond trading pairs: trading stock baskets against index and other statistical arbitrage strategies

Tutorial to MATLAB
- Overview of capabilities as research, backtesting, and automated trading platform
- Exercises: building some utilities useful for trading and plotting simple graphs
- Using toolboxes: a) Datafeed toolbox; b) Statistics toolbox; c) Financial, financial derivatives, fixed-income, econometrics, optimization, neural network, wavelet toolboxes
- Freeware: Spatial-econometrics.com

Basic concepts in pair trading
Stationarity of a time series
a) Concept of stationarity, and why is it useful
b) Statistical test for stationarity
c) Exercise: Using MATLAB and spatial-econometrics toolbox to find out if CADAUD is stationary
d) Exercise: if CADAUD is stationary, backtest a possible trading strategy
e) Exercise: Find out if EURGBP is stationary

Cointegration of 2 time series
a) Concept of cointegration, and why is it useful
b) Statistical test for cointegration
c) Exercise: Find out if GLD-GDX is cointegrating
d) Exercise: Find out if IGE-EWC is cointegrating
e) Exercise: Building and backtesting a simple Bollinger-band pair trading strategy on IGE-EWC

Calculating the half-life for mean-reversion
a) How long is the expected holding period for a mean-reversion trade?
b) Why is computing half-life better than computing average holding period?
c) Using Ornstein-Uhlenbeck formula to determine half-life
d) Exercise: Compute the half-life of mean-reversion for IGE-EWC

Practical aspects of pair trading
- The pitfall(s) of cointegration analysis
- Is cointegration necessary for pair trading?: a) There are cases where 2 instruments are not cointegrated over long time-scale, but nevertheless mean-revert under certain situations; b) Mean reversion may depend on time of day, or day of week, or day of month -- seasonal effects; c) Mean-reversion may depend on trading horizon
- "Parameterless" pair trading strategies: a way to "pyramid" or "layer" your pairs positions
- What about stop loss?
- Pair trading stocks, ETFs, and futures: a) Which is the best market?; b) The pitfalls of each market and some ways to avoid them
- High-frequency pair-trading: a) Divergence between backtest and reality; b) Exercise: backtesting a high frequency calendar spread strategy; c) Exercise: building a automated trading program for calendar spread strategy;

Beyond trading pairs
- Trading an ETF against a subset of its component stocks
- Exercise: Construct a trading model of XLE against its components
- Statistical arbitrage: picking two subsets of stocks out of an index and trading them against each other
- Exercise: Constructing a market neutral statistical arbitrage strategy

Conclusion
- Mean-reversion vs. momentum strategies: why focus on the former?
- Cointegration and mean-reversion: related, but neither necessary nor sufficient
- Different ways to construct mean-reverting portfolios
- Software for download, books and articles for references