Advanced Trading Rules (eBook)

Introduction

In presenting the second edition of this book, we have added three new chapters, in particular focusing on the area of technical analysis (chartism). We feel that this material should be included in any broad contemporary study on trading rules and we hope this inclusion will encourage further research on this area.

The past few years have seen an extraordinary explosion in the use of quantitative systems designed to trade in the foreign exchange and futures markets. This is witnessed by exponential growth of alternative investments, namely futures funds and hedge funds. Curiously, research on this area has been fragmented and sporadic. The purpose of this book is to bring together leading academics and practitioners who are working on systematic trading rules. It is well known that futures fund managers, among others, tend to rely on some sort of systematic trading rules. Available statistics suggest that systematic traders outnumber their discretionary counterparts by a ratio of two to one. As we will see in Chapter 13, the gap is even bigger for sectorized markets such as foreign exchange, interest rates and stock index futures.

This book does not present an exhaustive review of dynamic strategies applied by traders and fund managers, as this would be a hazardous task given the speed at which forecasting techniques and markets evolve. The purpose of this book is rather to introduce the reader to the theory of trading rules and their application. Numerous forecasting strategies are covered in this book, including technical indicators, chartism, neural networks and genetic algorithms.

There are two common factors linking all the strategies investigated in this book. First, all forecasting techniques attempt to predict the direction of price movements. Second, the criterion used to assess forecasting accuracy is economic significance. Trading rules are built out of forecasting strategies and their profitability subsequently measured.

Our primary concern is to specify trading rule-based tools which allow proper testing of the efficient market hypothesis. A market is said to be informationally efficient if prices in that market reflect all relevant information as fully as possible. This demanding requirement for an efficient market is often relaxed to a statement that trading systems cannot use information to outperform passive investment strategies when transaction costs and risk are considered. This book shows that many financial markets, especially foreign exchange and futures, may not be efficient according to this definition.

This book hopes to combine intellectual challenge and practical application, as reflected by the distinction and variety of the contributors: academics, traders, central bankers, tracking agencies and fund managers. Some readers will be interested in this book for what it says about the practical use of technical analysis and others for what it says about the distributional properties of dynamic strategies. The interaction between mathematical theory and financial practice has intensified since the development of Modern Portfolio Theory in the 1950s and the Black-Scholes analysis of the early 1970s, and this has reached a point where no firm can ignore it.

Any virtue can become a vice if taken to extremes, and just so with the application of mathematical models in finance practice. At times the mathematics of the models become too interesting and we lose sight of the models’ ultimate purpose: improving portfolio performance, risk management and trading book performance. Computer simulation of dynamic strategies using real data from foreign exchange, emerging and futures markets, will show that substantial risk-adjusted profits can be achieved. However, as with any computer simulation in financial markets, one cannot know how accurate the analysis is until one tries in real time with real money. Consequently, a complementary study of the usefulness of quantitative techniques must involve the review of fund managers’ performance using systematic trading rules.

This book includes three sections: the stochastic properties of trading rules, applications to the foreign exchange market and trading the futures markets. We shall next discuss the contributions of each of the fifteen papers.

The first section deals with the stochastic properties of trading rules (six chapters).

1 Blake LeBaron uses moving-average based rules as specification tests on the process for foreign exchange rates. Several models for regime shifts and persistent trends are simulated and compared with results from the actual series. The results show that these simple models cannot capture some aspects of the series studied. Finally, the economic significance of the trading results is tested. Returns distributions from the trading rules are compared with returns on risk-free assets and returns from the US stock market.

2 Andrew Lo, Harry Mamaysky and Jiang Wang propose a systematic and automatic approach to technical pattern recognition using non-parametric kernel regression, and apply this method to a large number of US stocks from 1962 to 1996 to evaluate the effectiveness of technical analysis. By comparing the unconditional empirical distribution of daily stock returns to the conditional distribution – conditioned on specific technical indicators such as head-and-shoulders or double-bottoms – they find that over the 31-year sample period, several technical indicators do provide incremental information and may have some practical value.

3 Daan Matheussen and Stephen Satchell assess the performance of various trading rules for TAA (tactical asset allocation) modelling across equity indices in the emerging markets. The authors find that rules based on mean and variance information and using a rolling window of information outperform all others absolutely and in a risk-adjusted sense, even when they take into account transaction costs.

4 Emmanuel Acar establishes the expected return and variance of linear forecasting strategies assuming that the underlying logarithmic returns follow some Gaussian process. The necessary and sufficient conditions to maximize profits are specified. This chapter shows that many technical forecasts can be formulated as linear predictors. The effect of conditional heteroskedasticity is investigated using Monte Carlo simulations.

5 George Kuo derives some exact results about the probabilistic characteristics of realized returns from two simple moving-average trading rules. The first rule needs only the information contained in the asset return at the present time to issue trading signals while the second rule requires the whole past history of the asset price to do so.

6 Emmanuel Acar and Stephen Satchell establish the distribution of returns generated by a portfolio including two active strategies assuming that underlying markets follow an elliptical distribution. The timing is triggered by linear forecasts for the sake of tractability. The most important finding is that conventional portfolio theory might not apply to active directional strategies even when the underlying assets follow a multivariate normal distribution.

The second section of this book demonstrates that the foreign exchange markets may be seen as inefficient given the number of profitable strategies which can be built out of varied forecasts (four chapters).

7 John Okunev and Derek White evaluate the performance of multiple classes of foreign exchange trading rules across eight base currencies. Specifically, they compare trading rules that focus on individual currencies with those that follow a long–short strategy across multiple currencies. The trading rules include pure momentum, buying/selling based upon relative interest rates, and moving-average rules. They find that a long–short strategy across multiple currencies outperforms trading rules that focus on individual currencies. In addition, they find that significant benefits may accrue by combining long–short moving-average rules across multiple currencies with long–short positions based upon relative interest rates.

8 Christian Dunis considers Artificial Neural Networks (ANNs), and discusses their application to economic and financial forecasting and their increasing success. This chapter investigates the application of ANNs to intraday foreign exchange forecasting and stresses some of the problems encountered with this modelling technique. As forecasting accuracy does not necessarily imply economic significance, the results are also evaluated by means of a trading strategy.

9 Kevin Chang and Carol Osler assess the incremental value of the head-and-shoulders pattern (H&S), consistently cited by technical analysts as particularly frequent and reliable, relative to filter rules. On an incremental basis, they show that the H&S trading rules add noise but no value. Thus, a trader would do no better, and possibly worse, by following both H&S and filter rules instead of filter rules only.

10 Pierre Lequeux investigates the assumption that the interest rates market leads the currency markets as money flows from one country to another. For a systematic trader the hypothesis is quite attractive; indeed if such a cross-correlation exists it will enable him to devise profitable trading strategies.

Finally, the third section analyses the application of stop-loss rules and other...