larsson larsson aberg statistical arbitrage final version 2002

Download Larsson Larsson Aberg Statistical Arbitrage Final Version 2002

Post on 08-Apr-2015

332 views

Category:

Documents

0 download

Embed Size (px)

TRANSCRIPT

A Market Neutral Statistical Arbitrage Trading Model01/01/95 07/02/97 01/01/00 07/02/020.500.511.522.533.544.5x 107DateSkrDevelopment of Our Default ModelAFGXAFGX 3m Moving AverageOur Model 3m Moving AverageOur ModelNet Position ShortErik LarssonLars LarssonJohan berg13th March 2003Larsson, Larsson and berg are students at the Stockholm School of Economics and the Royal Institute ofTechnology. The authors thank Patrik Sfvenblad (Head of Portfolio Management at RPM), Andrei Shleifer(Professor of Economics at Harvard), and Johan Ahlberg (Datastream expert) for helpful comments. Email:primus@kth.se (Erik), d98-lla@nada.kth.se (Lars) and 17774@student.hhs.se (Johan).AbstractThe momentum eect is a systematic ineciency in the market that can be exploitedby a trading strategy. This conclusion is supported by theoretical and empirical evidence.But the academic research that tries to quantify the performance of this kind of strategyoften relies on a methodology that is too simplistic. The question arises what performancea trader realistically could achieve in relation to the results presented in academic journals.To answer this, we have written a computer program to run simulations with the addedrealism of transaction costs and more advanced trading rules based on a wider array of datathan classic methodology allows. This has been done on Swedish stocks between 1995 and2001. We then compare the simulation based on our own advanced model with a simulationthat emulates a simplistic methodology.It is found that the negative impact on return of including transaction costs is outweighedby the lower risk attributed to our more advanced trading rules, as indicated by e.g. Sharpeand standard measures of risk. We can thus conclude that the momentum eect might beeven more attractive as a basis for a trading strategy than have been suggested in prioracademic research.As an academic paper, we think that the methodology (our simulation platform) used toobtain the conclusion in our thesis is more important than the conclusion itself. It is evidentthat a good evaluation of any trading strategy requires more realistic simulations than iscommonplace in academia today.Contents1 Introduction 11.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.2 Purpose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.3 Delimitation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.4 Outline and Readers Guide . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 Previous Research 42.1 Theoretical Justication for Momentum Strategies . . . . . . . . . . . . . . . 42.1.1 Ecient Markets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42.1.2 Data mining and Spurious Data . . . . . . . . . . . . . . . . . . . . . 62.2 A Model of Over- and Underreaction . . . . . . . . . . . . . . . . . . . . . . . 72.2.1 Conclusions and Discussion . . . . . . . . . . . . . . . . . . . . . . . . 122.3 Theoretical Motivation for the Stop-Loss Rule . . . . . . . . . . . . . . . . . . 142.4 Empirical Evidence Using CAR . . . . . . . . . . . . . . . . . . . . . . . . . . 152.4.1 The Momentum Eect . . . . . . . . . . . . . . . . . . . . . . . . . . . 152.4.2 The Role of Volume, Fundamental and Sentiment Variables . . . . . . 172.4.3 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183 An Overview of Our Model and Its Context 193.1 Dierent Hedge Fund Strategies . . . . . . . . . . . . . . . . . . . . . . . . . . 193.2 Our Model an Optimization Problem . . . . . . . . . . . . . . . . . . . . . . 203.3 Technical Specication of the Computer Program . . . . . . . . . . . . . . . . 224 Data 244.1 Description of the Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245 Methodology Signal Generation and Risk Control 255.1 Signal Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255.2 Risk Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265.3 Assumptions in Our Simulations . . . . . . . . . . . . . . . . . . . . . . . . . 275.3.1 Basic Assumptions of Our Trading Model . . . . . . . . . . . . . . . . 276 Results - Portfolio Simulations 296.1 Results of Our Trading Model . . . . . . . . . . . . . . . . . . . . . . . . . . . 296.1.1 Comments on the Results of Our Model . . . . . . . . . . . . . . . . . 326.1.2 Comparison with the Simplistic Momentum Strategy . . . . . . . . . . 337 Conclusions 35A Appendices 36A.1 Robustness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36A.1.1 Four Criteria of Robustness . . . . . . . . . . . . . . . . . . . . . . . . 36A.1.2 Tables of Results for Robustness . . . . . . . . . . . . . . . . . . . . . 37A.1.3 Discussion on the Robustness of Our Model . . . . . . . . . . . . . . . 43A.2 Statistical Comparison of Our Model and the Simplistic Model . . . . . . . . 45A.2.1 Volatility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45A.2.2 Return . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45A.3 Screenshot of the Program . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47A.4 A Day of Trading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48iA.5 A List of the Swedish Stocks Used . . . . . . . . . . . . . . . . . . . . . . . . 49A.6 Some Basic Financial and Statistical Concepts . . . . . . . . . . . . . . . . . 50iiList of Figures1 Realizations of earnings over 600 periods . . . . . . . . . . . . . . . . . . . . . 102 The essence of the momentum eect. . . . . . . . . . . . . . . . . . . . . . . . 113 Histogram of our simulation of the overreaction eect . . . . . . . . . . . . . 134 Histogram of our simulation of the underreaction eect . . . . . . . . . . . . . 135 Conceptual view of the computer system. . . . . . . . . . . . . . . . . . . . . 236 A diagram over the development of our model, compared with the same in-vestment in AFGX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297 Histogram describing the distribution of the daily returns of our model . . . . 318 Histogram describing the distribution of the monthly returns of our model . . 319 The program in action . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47List of Tables1 Illustration of how the investor changes her beliefs . . . . . . . . . . . . . . . 92 A brief summary of empirical research on the momentum eect. . . . . . . . . 163 The momentum eect in Germany, compared with Jegadeesh-Titmans resultsfor USA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174 The momentum eect in Sweden 1980-1999 . . . . . . . . . . . . . . . . . . . 185 Hedge fund history at a glance . . . . . . . . . . . . . . . . . . . . . . . . . . 196 Categorization of hedge funds . . . . . . . . . . . . . . . . . . . . . . . . . . . 207 Statistical Arbitrage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208 Performance Jan. 1990 - Dec. 1999 of various hedge fund categories . . . . . 219 Our strategy . . . . . . . . . . . . . . . . . . . . . . . . .

Recommended

View more >