Stock trading using computational intelligence - Free Essay Example

Sample details Pages: 30 Words: 8973 Downloads: 3 Date added: 2017/06/26 Category Statistics Essay Did you like this example? Stock Trading using Computational Intelligence t Computational Intelligence has been widely used in recent years in many areas, such as speech recognition, image analysis, adaptive control and time series prediction. This research attempts to explore the usefulness of neural network and support vector machine in financial market. Two popular stock market indexes have been studied: Hong Kong Hang Seng Stock Index and Dow Jones Transportation Index. The performance of neural network and support vector machine are evaluated in two dimensions: error in forecasting and trading profits. Popular technical indicator, percentage price oscillator (PPO), has been selected as training input and output. Predictive models use previous 8 days PPO to forecast future 5 days PPO. Empirical results on Hong Kong Hang Seng Index show that multilayer perceptron optimized with GA (MLP-GA) trading system obtain 6.71 times of original capital from 1997-1-29 to 2007-3-8, totally 2500 trading days. While support vector regression optimized by genetic algorithms (SVR-GA) trading system generates 5.705 times of original capital during the same time horizon. In contrast, conventional non-predictive trading system only produces 2.064 times of starting equity. Buy and Hold strategy gives 1.605 times return to investors. A recent published fuzzy trading system provides 5.781 dollars as final equity for 1 dollar initial investment. Don’t waste time! Our writers will create an original "Stock trading using computational intelligence" essay for you Create order Further evaluations of two intelligent trading systems have been made. A back test using the same parameters and same assumptions on Dow Jones Transportation Index have further proved the robustness of the proposed trading systems. MLP-GA trading system provides 4.87 times of initial capital and SVR-GA trading system obtains 5.168 as final equity. These two intelligent trading systems again outperform conventional trading system, which generate 2.805 dollars for 1 dollar investment. Acknowledgements I am very grateful to my final year project supervisor, Associate professor Wang Lipo, and would like to take this opportunity to thank him for his patient and insightful guidance throughout the project. Professor Wang always offers me detailed and valuable explanations and suggestions in our discussion, and provides me useful knowledge about doing research. Not only professor Wang enlightens me in academic area, he also arranges meeting with industrial professionals for me to discuss this project. Again, I would like to express my sincere appreciation to professor Wang. Zhu Ming April, 2010. Stock Trading using Computational Intelligence List of Figures Fig 21 A multi layer neural network with L layers 13 Fig 22 Maximum-margin hyperplane and margins for a SVM trained with samples from two classes. 16 Fig 23 Genetic Algorithm flowchart, with maximum 100 generation 18 Fig 24 One point crossover 19 Fig 25 roulette-wheel selection 20 Fig 31 Dow Jones Industrial Average price, with EMA plotted. 23 Fig 32 Using single EMA 23 Fig 33 Using two EMA to make decision 24 Fig 34 A predictive trading system. 26 Fig 35 Structure of GA optimized MLP 28 Fig 41 Training performance of MLP 33 Fig 42 MSE for out of sample data 34 Fig 43 Linear regression for trained neural network 35 Fig 44 Linear regression for out of sample data 36 Fig 45 Equity curve for intelligent and conventional trading systems 37 Fig 46 Trading signal of NN+GA trading system 38 Fig 47 Trading signal of conventional trading system 39 Fig 48 MSE for GA+SVR model 41 Fig 49 Equity curve for GA+SVR trading system and conventional trading system 42 Fig 410 Comparison of 4 trading systems 43 Fig 411 Equity curves of different trading system on DJT 44 List of Tables Table 31 Settings for GA and NN 26 Table 32 Settings for GA and SVR 29 Table 41 Data distributions for training and testing neural network 32 Table 42 Total return for different prediction time horizon 34 Table 43 Trading performance comparison 42 Stock Trading using Computational Intelligence Chapter 1 Introduction 1.1 Background Analyzing stock market is one of the most important and fascinating issue as it is highly related with the profitability of investment. There are two main types of analysis in financial market: technical analysis and fundamental analysis. Fundamental analysis is based on the premise that a stock, bond, fund, commodity, or a market as a whole has an underlying intrinsic value. By analyzing the fundamental characteristics, such as assets, liabilities, income, supply or demand, values can be determined [11]. Normally fundamental analysts use a trading strategy called Buy and Hold, since they tend to buy the stocks of undervalued companies or the companies with great growth potentials. They believe that the share price would rise eventually since the company they buy is growing. Hence, they would like to keep the stocks for a relative long time. On the other hand, technical analysis believes that the markets price reflects all the relevant information, such as news and events. Thus, pric e is the only information they need to analyze. In their perspective, history will repeat itself such that we could trade for profits. Therefore, technical analysis only employs historical data to build the model for future investment. Over the past decade, Computational Intelligence has been widely used in stock trading, such as using neural networks (NN) [10]). Using computational intelligence could provide opportunities for investors to combine the information gathered from fundamental analysis and technical analysis to make trading decision. Mainly, two types of input data have been used in computational intelligence. One type, price or technical indicators, is considered as technical analysis. The other type includes macroeconomic indices and information related to a specific company, such as the interest rate and P/E ratio. Many pioneer scholars have focused on minimizing the mean square error (MSE) in price direction prediction as well as providing paper profits in trading financial market. Patel et al [10] uses hierarchical coevolutionary fuzzy system (HiCEFS) to predict a technical indicator and hence build a prudent trading strategy. Furthermore, by testing this model with real world data of Hong Kong Hang Seng Index and NOL stock in Singapore Exchange, they achieved a final return of 14.251 times of original capital on NOL stock in 2329 trading days and 5.781 times of original capital on Hang Seng Index in 2461 trading days. 1.2 Objectives and Scope The objective of this project is to explore and examine the usefulness of computational intelligence in stock trading on Hong Kong Hang Seng Index and Dow Jones Transportation Index. The intelligent trading system built on matlab could analyze the historical data and generate buy or sell signals for any given time series. The main objectives are as follows: 1. Apply intelligent trading system on Hong Kong Hang Seng index to generate buy and sell signals. The intelligent trading system could be constructed with neural networks optimized by genetic algorithm or support vector machine optimized by genetic algorithm. 2. Examine entry and exit signals generated by intelligent trading system and non-intelligent trading system. Compare the empirical trading profits between them. 3. Compare the trading performance of intelligent trading system with other researchers work, using the same data and trading rules. 4. Further validate the trading systems performance by applying the proposed system on Dow Jones Transportation Average Index, and compare the trading profits with non-intelligent trading system. 1.3 Organisations This report is organized into 5 chapters: Chapter 1 provides some background knowledge of financial market and other researchers accomplishment on using computational intelligence in financial market. It also gives a detailed project objectives and scope. Chapter 2 introduces the background knowledge for this project, such as neural network, support vector machine and genetic algorithm. Chapter 3 describes the proposed methodology of this project. It introduces the technical indicators and inputs to the intelligent trading system, the architectures of the trading system. In addition, it also provides the settings for each intelligent prediction model, as well as the data preparation for these prediction models. Chapter 4 presents the empirical results of trading Hong Kong Hang Seng Index and Dow Jones Transportation Average Index. Furthermore, it compares the results with non-intelligent trading system as well as buy and hold strategy. Chapter 5 summarizes the project and provides the future work for the project. Chapter 2 Literature Review 2.1 Artificial Neural Networks An artificial neural network (ANN) is inspired by the structure and functions of biological neural networks, and expressed using mathematical models. It consists of an interconnected group of artificial neurons and processes information using a connectionist approach to computation. In most cases an ANN is an adaptive system that changes its structure based on external or internal information that flows through the network during the learning phase. Modern neural networks are non-linear statistical data modeling tools. They are usually used to model complex relationships between inputs and outputs or to find patterns in data. Neural networks are considered as highly parallel system which could learn from the past data and would be able to apply the knowledge learned to new data. 2.1.1 Multilayer Perceptron Neural Networks There are varies of ANN structures, multilayer perceptron neural networks (MLP) is one of them. It is a feed-forward network has a layered structure. Each layer consists of units which receive their input from units from a layer directly below and send their output to units in a layer directly above the unit. There are no connections within a layer Fig 21. The inputs are fed into the first layer and each input is associated with a weight. The first layer outputs are considered as second layers input and eventually calculated the final output. The activation function for each layer is described as: in which Information in MLP networks only move in the forward direction, from the input nodes through the hidden layers and to the output layer. There are also no loops in a MLP network. Fig 21 A multi layer neural network with L layers 2.1.2 Back Propagation Back propagation is a common method of teaching artificial neural networks how to perform a given task. It was first described by Arthur E. Bryson and Yu-Chi Ho in 1969,[14]. Back propagation is a supervised learning method, and is an implementation of the Delta rule. It requires a teacher that knows, or can calculate, the desired output for any given input. In another word, it has to be provided with desired output in order to calculate the errors. The errors propagate backwards from the output nodes to the inner nodes and from the inner nodes to input nodes. Hence back propagation is a method to calculate the gradient of the error for the network with respect to the networks modifiable weights, either in input layer or in hidden layer. In short, back propagation algorithm could be describe as below. Summary of the backpropagation technique: 1. Present a training sample to the neural network. 2. Compare the networks output to the desired output from that sample. Calculate the error in each output neuron. 3. For each neuron, calculate what the output should have been, and a scaling factor, how much lower or higher the output must be adjusted to match the desired output. This is the local error. 4. Adjust the weights of each neuron to lower the local error. 5. Assign blame for the local error to neurons at the previous level, giving greater responsibility to neurons connected by stronger weights. 6. Repeat from step 3 on the neurons at the previous level, using each ones blame as its error. 2.1.3 Levenberg-Marquardt Algorithm Levenberg-Marquardt Algorithm is used for training the neural network. It could be used to modify the ANNs weights of each layer. The Levenberg-Marquardt Algorithm interpolates between the Gauss-Newton algorithm and the method of gradient descent. It is more robust than the Gauss-Newton algorithm, which means that in many cases it finds a solution even if it starts very far off the final minimum. On the other hand, for well-behaved functions and reasonable starting parameters, the Levenberg-Marquardt Algorithm tends to be a bit slower than the Gauss-Newton algorithm. Levenberg-Marquardt Algorithm could be expressed as [15] 2.2 Support Vector Machine Support Vector Machine (SVM) is a relatively new learning method developed from statistical learning theory. Compared with traditional statistics, statistical learning theory does not assume infinite samples, but rather focused on estimations utilizing small samples. The basic idea of support vector machine is to find a hyperplane which separates the d-dimensional data perfectly into its two classes. Support Vector Machine is a supervised learning method which could map the input space to output space Fig 22. Given that a training set (), i = 1, the support vector machine requires the minimum value of following formula [17]. Fig 22 Maximum-margin hyperplane and margins for a SVM trained with samples from two classes. 2.2.1 Support Vector Regression Support Vector Machine used in regression was proposed in 1996 by Vladimir Vapnik, Harris Drucker, Chris Burges, Linda Kaufman and Alex Smola [18], which is called support vector regression (SVR). The model produced by support vector machine used in solving classification problems depends only on a subset of the training data or called support vectors, because the cost function for building the model does not care about training points that lie beyond the margin. Similarly, the model produced by SVR depends only on a subset of the training data, because the cost function for building the model ignores any training data close to the model prediction. Given a training set (), i = 1, the target of SVR is to find a linear function that could minimize the discrepancy between the desired output and predicted output. The optimal regression function is the same with SVM. There are several kernel functions commonly used in SVR, which includes liner, polynomial, radial basis function and sigmoid kernel function. Their respective formula is as below [23]: n Linear: n Polynomial: n Radial Basis Function (RBF): n Sigmoid: Here, are kernel parameters Support Vector Machine or SVR has some advantages when comparing to Neural Networks. For instance, it does not over fit the training data since it uses only several training data as support vectors. However, parameters in SVR would affect the final results in spite that SVR has much fewer parameters compared to NN. The main parameters in SVR are error insensitive tube around the regression function [19] and the balance of training errors with model complexity. 2.3 Genetic Algorithm Genetic algorithm (GA) is a searching technique to look for exact or approximate solutions for optimization and searching problems. It is considered as global search heuristics.GA uses techniques inspired by evolutionary biology such as inheritance, mutation, selection, and crossover. A typical genetic algorithm requires: 1. a genetic representation of the solution domain 2. fitness function to evaluate the solution domain In GA, an abstract representation of candidate solutions is called chromosomes, and it could be used in an optimization problem evolves toward better solutions. Solutions are represented in some encoding method, such as binary encoding. A fitness function is a particular type of objective function that prescribes the optimality of a solution so that a particular chromosome may be ranked against all the other chromosomes. The evolution usually starts from a population of randomly generated individuals. In each generation, the fitness of every individual in the population is evaluated. Based on their fitness, the fittest group of individuals are selected and through reproduction, crossover or mutation to form a new population. The new population is then used in the next iteration of the algorithm. Commonly, the algorithm terminates when either a maximum number of generations has been produced, or a satisfactory fitness level has been reached for the population. A common genetic algorit hm is shown Fig 23. Fig 23 Genetic Algorithm flowchart, with maximum 100 generation 2.3.1 Operators of Genetic Algorithm When generating the next generation population of solutions, GA would use genetic operators: crossover, and/or mutation. For each new solution to be produced, a pair of parent solutions is selected for breeding from the pool selected previously. By producing a child solution using the above methods of crossover and mutation, a new solution is created which typically shares many of the characteristics of its parents. Crossover selects genes from parent chromosomes and creates a new offspring. One common way is using single crossover point on both parents organism strings. All data beyond that point in either organism string is swapped between the two parent organisms. An illustration on one point crossover is shown in Fig 24 Fig 24 One point crossover There are other ways for crossover, for example two crossover points could be chosen. Crossover can be rather complicated and very depends on encoding of chromosome. In some cases, GA performance could be enhanced by trying out other crossover techniques. After a crossover is performed, mutation takes place. The purpose of mutation in GA is to preserve and introduce diversity. Local minima could be prevented because of mutation, and the population of chromosomes would not be too similar to each other so that the evolution could continue. Mutation changes the new offspring randomly. For binary encoding, a common way is switching a few randomly chosen bits from 1 to 0 or from 0 to 1. 2.3.2 Selection in Genetic Algorithm Selection would choose individual genomes from a population for breeding next generation. There are varies of selection algorithms, such as roulette-wheel selection, rank selection or Tournament selection. Roulette-wheel selection chooses parents according to their fitness. The chromosome has high fitness possesses the higher chances to be selected. The fitness level is used to associate a probability of selection with each individual chromosome. This algorithm could be imagined as roulette wheel in casino, where the larger piece has higher probability to be chosen, as shown in Fig 25. If is the fitness of individual i in the population, its probability of being selected is, where N is the number of individuals in the population. Fig 25 roulette-wheel selection Tournament selection involves running several tournaments among a few individuals chosen at random from the population. The winner of each tournament (the one with the best fitness) is selected. Selection pressure is easily adjusted by changing the tournament size. If the tournament size is larger, weak individuals have a smaller chance to be selected. Chapter 3 Intelligent Trading System Design 3.1 Technical Analysis Technical analysts seek to identify price patterns and trends in financial markets and attempt to exploit those patterns.[20] People who are using technical analysis would search for archetypal patterns, such as the well-known head and shoulders or double top reversal patterns, study indicators such as moving averages, and look for forms such as lines of support, resistance, channels, and more obscure formations such as flags, pennants or balance days. In this project, only indicators have been studied since they are quantitative and do not require ambiguous identifications. Among all the technical indicators, moving average is considered as the simplest and most useful one. It is popular because moving average could discover the trends by smoothing the prices. Most importantly, moving average could be a useful tool since investors can make profits through trends. Exponential moving average (EMA), being one of the moving average indicators, is considered as more adaptive since it puts more weights on recent prices, e.g., todays close price, while putting less weights on earlier days. Equation below shows the calculation of EMA: The plot of long term EMA of 45 days and short term EMA of 15 days are plotted with close price for Dow Jones Industrial Average Index in Fig 31, all data and figures are provided by yahoo finance. Dow Jones Industrial Average price, with EMA plotted. There are many ways of using EMA, and two common uses are introduced here. First, investors could take a long position, or buy the stock index when close price is above the EMA, and take a short position when close price is under EMA. An example is shown in Fig 32, using 30 days of EMA on Dow Jones Industrial Average. Although there are some whipsaw in the middle, using single EMA is helpful to investor when making buy or sell decisions. Fig 32 Using single EMA Another way of using EMA is taking a long position (buy) when short term EMA is above long term EMA, and taking a short position when short term EMA is under long term EMA. An example of how to buy or sell is illustrated in Fig 33, using 15 days EMA and 45 days of EMA. As we could see on the chart, this method is effective by taking large profits and suffering small losses. Fig 33 Using two EMA to make decision It is clear that EMA could help investors to identify the trend. However, being able to discover the trend is not good enough, the trading rule should be established to take profits through the trend. However, using chart and technical indicators are not sufficient since there are some serious disadvantages. For example, we do not know whether this technical indicator could bring investors consistent long term profits. Also, we do not know how many shares should we buy or sell. Without providing more information on these topics, investors may not dare to trade with real money. However, a quantitative trading system based on these indicators could concur the shortcomings. A well established trading system would be able to tell when to buy and when to sell, as well as how many shares to buy and sell. In addition, a trading system could provide back testing results, which could present the trading performance to investors, such as the equity curve or maximum drawdown. Therefore, in this project, a quantitative trading system is built and tested. This trading system uses a technical indicator named Percentage Price Oscillator (PPO), PPO is calculated as formula below: A buy signal is triggered if PPO is greater than 0, in other words, when short term EMA crosses over with long term EMA. A sell signal is triggered if PPO is less than 0, which means long term EMA is above short term EMA. This trading system is a typical trend following system which could catch every major trend to make promising profit, while suffering minuscule losses when significant trends are absent in the market. 3.2 Computational Intelligence in Trading When using PPO trading system, there would be a lag between the time when the trend starts and the time when the trading system detects it. Failing to compensate the lag has been a dominant disadvantage of traditional trading systems (without prediction). An intelligent trading system attempts to predict PPO in the near future, so as to enter the market before the trend while closing the position before the market falls. The input for our intelligence trading system studied in this paper is PPO of the last 8 days and the output is PPO in the future 5 days. The intelligent model is either an MLP optimized by GA or an SVM optimized by GA. 0.2% of transaction cost and slippage are counted in the process of calculating profits, as indicated in Fig 34. Fig 34 A predictive trading system. 3.3 Experimental Settings 3.3.1 GA optimized neural network In this project, a feed-forward MLP with one hidden layer is used. The number of hidden neurons is determined to be 30 by the trial and error. The Levenberg-Marquardt algorithm is used to train the MLP. Initial weights of the neural network are determined by GA. The settings for the NN+GA model are selected as Table 31. GA settings the population size of GA 300 Maximum Generation 800 Stop criteria maximum generation reached the probability of mutation 0.02 Neural Network Settings layers Single hidden layer with 30 neurons Transfer function Transig, purelin Training Levenberg-Marquardt performance Mse (mean square error) Table 31 Settings for GA and NN Using genetic algorithm to determine the initial weight and bias is essential since they have great impact on the generalization ability of the neural network. If the weights and bias are initialized with some random number and they happen to be far way from a good solution, or near local optimum, the neural network may not be trained to achieve good performance. Being trapped in local extremes is normally happened. On the other hand, appropriate initialization would put the weights and bias near a good solution, and hence provide a high chance for neural network to reach better outcome. In this project, genetic algorithm is chosen to provide the initial weights and bias for neural network. The structure of using GA to optimize MLP is shown in Fig 35. The fitness in GA is based on the error of predicted output and desired output, shown as below Where is the desired output and is predicted output. 3.3.2 GA optimized SVR Main parameters in SVR are error insensitive tube around the regression function [14] and the balance of training errors with model complexity. In this paper, GA is used to determine the best SVR parameters. The structure of GA optimized SVR is the same as using GA to optimize MLP, where GA is trying to minimize the difference between desired output and predicted output. The settings for GA optimized SVR model are listed in Table 32 Fig 35 Structure of GA optimized MLP GA settings the population size of GA 30 Maximum Generation 200 Stop criteria maximum generation reached the probability of mutation 0.05 the probability of crossover 0.4 SVR Settings Kernel function radial basis function Table 32 Settings for GA and SVR 3.4 Preprocessing Input Data Once the appropriate raw input data has been selected (in this case, they are previous 8 days PPO) , it must be preprocessed; otherwise, the neural network will not produce accurate forecasts. The decisions made in this phase of development are critical to the performance of a network. Normalization is commonly used to distribute the input data evenly and scale it into an acceptable range for the network. Knowledge of the domain is important in choosing preprocessing methods to highlight underlying features in the data, which can increase the networks ability to learn the association between inputs and outputs. In normalizing data, the goal is to ensure that the statistical distribution of values for each net input and output is roughly uniform. In addition, the values should be scaled to match the range of the input neurons. This means that along with any other transformations performed on network inputs, each input should be normalized as well. In this project, mapping the training input minimum and maximum values between -1 and 1 is adopted as normalizing method. In this method, it is assumed that the input has only finite real values, and that the elements are not all equal, as indicated below. Where in this case is 1, is -1. is the largest number of training input, while is the smallest number of training input. stands for each individual training data, and is the normalized training data. For the testing set, data should also be scaled to a certain range, as training set does. However, the largest number and smallest number of testing set are not available since we assume these data are unknown for trading simulation. Therefore, the testing data set are scaled using the parameters in training input data. In specific, and are still the largest number and smallest number in training data set. Chapter 4 Results and Evaluation This chapter illustrates the experiment results for 2 intelligent trading models, which are using GA optimized MLP and using GA optimized SVR. In addition, it introduces some evaluation criteria, and evaluates the prediction models according to these criteria. Furthermore, it analyzes and compares the return of capital and maximum drawdown with other publication as well as conventional trading method. 4.1 Experimental Data This intelligent trading system uses Hong Kong Hang Seng Stock Index (HSI) from 1986-12-31 to 1997-1-28, total 2500 daily close price as in sample training session, and uses HSI from 1997-1-29 to 2007-3-8, total 2500 daily price as out of sample testing data. All the HSI index data was obtained from Yahoo Finance (https://finance.yahoo.com/q/hp?s=^HSI). In sample data used to train the neural network have been separated into three sets: training, validation, and testing. In this project, we divide the input data randomly such that the first 60% of the samples are assigned to the training set, the next 20% to the validation set, and the last 20% to the test set. Table 41 is to summarize the distribution of experimental data. Data Set Distribution (%) Distribution(data) Training Data Training set 60% 1500 Validation set 20% 500 Test set 20% 500 Total 100% 2500 Testing Data 100% 2500 Table 41 Data distributions for training and testing neural network 4.2 GA Optimized MLP Trading System 4.2.1 Forecasting Performance The GA optimized MLP model is used to predict the future 5 days PPO. The performance of this predicative model could be evaluated by mean square error (MSE). MSE could be expressed as below Where is the target output and is the predicted output. The performance of forecasting in terms of MSE is 0.0087 for out of sample data, while 0.00213 for in sample data. In either case, we could see that the MSE is relatively small, which means the prediction is acceptable. In Fig 42, the difference between desired output and predicted output is plotted, as we could see, although there are some large errors in prediction, most of the forecasting is acceptable. Fig 41 Training performance of MLP The training results of neural networks could be further evaluated by linear regression. The best network is indicated by the correlation coefficient, r closed to unity (r à ¢Ã¢â‚¬ °Ã‹â€  1) Fig 44 shows the linear regression for out of sample data, which is 0.91227. Although it is nearly 8% lower compare with performance of in sample data, this model could still be considered as well trained neural network. Fig 42 MSE for out of sample data 4.2.2 Empirical Trading Results and Evaluation PPO of future 5 days is selected to be desired output after prudent consideration. As a matter of fact, forecasting larger time horizon would definitely produce more profit, which is made by early entry and early exit. On the other hand, the larger the time horizon, the harder it is to predict. This would increase the chance of wrong prediction, which decreases the profit. Table 42 is total return of investing 1 dollar, with different prediction time horizon. Prediction Time Horizon Total return No prediction 2.064 Predict future 3 days PPO 4.357 Predict future 5 days PPO 6.910 Predict future 7 days PPO 5.464 Table 42 Total return for different prediction time horizon In this experiment, reinvesting all capital is selected as the money management strategy, in which the trading system would re-invest all the profit and initial capital for next buy or sell decision. Fig 43 Linear regression for trained neural network Fig 44 Linear regression for out of sample data The proposed trading system assumes that it is possible to enter the market using the close price on the same day which triggers the trading signal. In addition, it assumes that the initial capital is 1 dollar and it is valid to buy or sell fraction number of the HSI. The PPO is calculated using parameters that short term of 15 days EMA and long term of 45 days EMA. The equity curves of proposed intelligent trading systems are shown in Fig 45 with equity curve of conventional trading system and equity curve for buy and hold trading strategy in contrast. The predictive MLP+GA model achieves 6.71 times of original capital from 1997-1-28 to 2007-3-8 while in the mean time, a non-predictive trading system only achieves 2.064 for 1 dollar investment, and buy and hold trading strategy generates 1.605 as final capital. In comparison, Huang and Quek et al. [10] use hierarchical coevolutionary fuzzy system (HiCEFS) to achieve 5.781 times of original capital on Hang Seng Index on the same trading days. Fig 45 Equity curve for intelligent and conventional trading systems Sample testing data is shown in Fig 47, it is obvious that prediction trading system would enter the market and exit the market earlier compared with trading system without prediction. However, using prediction has certain disadvantage. During non-trendy time, the proposed trading system may make wrong prediction and hence suffer some losses. For example, NN+GA trading system enters the market at day 61 at price 13030 and exit on the day 138 at price 15600, takes profit of 2570 points. On the other hand, for trading system without prediction, it enters the market at day 65 at price 13630 and exit at day 144 at price 13710, takes a profit of 80 points. That is the reason why the predictive model performs better than trading system without prediction. But during non-trendy market, such as around day 400, the trading system without prediction holds the position while the intelligent model made a wrong prediction. In this case, the investment incurred some losses. Fig 46 Trading signal of NN+GA trading system Fig 47 Trading signal of conventional trading system Moreover, another important criterion to evaluate the trading system is the maximum drawdown (MDD). MDD is defined as the maximum cumulative loss from a market peak to the following trough [22] The trading system using NN+GA suffers a MDD from 3.079 dollars to 2.443 dollars, which is 20.65% of the highest capital. In contrast, the trading system without prediction would have a MDD from 1.705 dollars to 1 dollar, which is 41.34% of the highest capital. Buy and Hold strategy suffers a MDD from 1 dollar to 0.466 dollar, which is 53.4% drop from the peak capital. Thus the NN+GA trading system reduced the risk involved. As it is shown in Fig 45 regarding the conventional trading system without prediction, the capital is back to original 1 dollar after 1276 trading days. This may shake peoples will to follow this system. On the other hand, the MDD happened in NN+GA trading system is from day 903 to day 930, which is easier for investors to follow the trading system. All the trading records are listed in appendix A. 4.3 GA Optimized SVR Trading System 4.3.1 Forecasting Performance The performance of forecasting future 5 days PPO using GA optimized SVR is evaluated in terms of MSE. MSE is 0.0058 for out of sample data, in contrast, MSE is 0.0087 in using GA optimized NN model for the same data. In another word, GA optimized SVR has smaller MSE, or better forecasting. In Fig 48, the difference between desired output and predicted output is plotted. However, better forecasting does not guarantee better profitability. Some wrong prediction at the top or at the bottom would bring larger losses comparing with wrong prediction at other situations. 4.3.2 Empirical Trading Results and Evaluation The same assumptions are made as using GA+NN trading system. In addition, 15 days EMA and 45 days EMA are used to form PPO. The equity curve of GA+SVR trading system is shown in Fig 49 with equity curve of conventional trading system in contrast. This GA+SVR trading system achieves 5.705 times of original capital. Fig 48 MSE for GA+SVR model Although this predictive model does not achieve profit as much as GA+NN model, it has its own advantage. First, this SVR model would provide consistent performance after each training session. Second, in term of prediction accuracy, GA+SVR model offers smaller prediction errors while GA+NN mode has larger errors. Last, it trades less frequently compared with GA+NN model, this would give investors different options to choose which type of trading systems are fitting to them. For active traders, GA+NN model could be more suitable for them, while for less active investors, GA+SVR model could be adopted since it trades less frequently. The comparison of GA+NN trading system, GA+SVR trading system, conventional trading system and buy and hold strategy is shown in Fig 410, the equity curves for 4 trading system mentioned above are plotted together for comparison. Fig 49 Equity curve for GA+SVR trading system and conventional trading system Trading System Final Equity MDD Win ratio Trading times Long position times Short position times GA+NN trading system 6.71 20.65% 49.6% 127 63 64 GA+SVR trading system 5.705 28.5% 44.8% 67 30 37 Conventional trading system 2.064 41.34% 48.7% 41 20 21 Buy and hold strategy 1.605 53.4% 100% 1 1 0 Table 43 Trading performance comparison Fig 410 Comparison of 4 trading systems 4.4 Further Evaluation In designing trading system, one of the most important issues is to avoid over curve fitting the system to back testing data. The more you bend your system around to improve performance on past data, the less likely it is your system will trade profitably in the future. Past performance will only approximate future performance to the extent the system is not over curve fitted. There are many ways to examine the over curve fitting trap. One way is to do back testing long enough. The longer the historical time period a system can trade profitably, the more robust it is. Another way to guard effectively against over-curve-fitting is to make sure your system works in many markets using the same parameters. Hence, the trading system is further evaluated by applying to Dow Jones Transportation Index (DJT). The data used as in sample training data is from 1968-9-20 to 1978-9-5, totally 2500 trading days, and data used as out of sample testing data is from 1978-9-6 to 1988-7-26, which is 2500 trading days. All data is from yahoo finance (https://finance.yahoo.com/q?s=^DJT). All the same assumptions are the same as trading HSI using intelligent trading systems. The equity curves of GA+NN trading system and GA+SVR trading system are shown in Fig 411 with equity curve of conventional trading system in contrast. This GA+SVR trading system achieves 5.168 times of original capital, while the predictive GA+NN model achieves 4.87 times of original capital while in the mean time, a non-predictive trading system only achieves 2.805 for 1 dollar investment. Fig 411 Equity curves of different trading system on DJT GA+NN trading system and GA+SVR trading system outperform the conventional trading system again on DJT. This further proves that using computational intelligence would enhance the performance of conventional trading system. In addition, the proposed intelligent trading systems, using GA+NN or using GA+SVR, would survive in different market, such as DJT and HSI, and be able to generate profits consistently. Chapter 5 Conclusion In this project, a predictive trading system is proposed to trade on real market data of Hong Kong Hang Seng Index, and trade on Dow Jones Transportation Index as cross market validation. Neural network optimized by GA and support vector regression optimized by GA are implemented as predictive model in the trading system. The trading system mainly uses technical indicator price percentage oscillator (PPO) as trading rules. Hence the predictive model uses last 8 days PPO as input to predict future 5 days PPO, and based on predicted PPO to make trading decisions. The testing period is 10 years, which is long enough to reduce the possibility of curve fitting. The proposed predictive trading system produces around 3 times more profits on HSI compared with conventional trading system without prediction, and around 2 times more profits on DJT compared with non predictive trading system. Despite promising profits generated by the trading system, further improvements such as applying the system to other new immerging markets, such as China Stock market, or applying a better money management strategy can be considered as future research area. Furthermore, due to the randomness introduced by GA, neural network may not always be trained well enough every time. We shall study effective ways to assure reasonable performance for each training session. References [1] E. F. Fama, The Behavior of Stock Market Prices, Business, vol. 38, pp. 34-105, 1965. [2] A. P. N. Refenes, A. N. Burgess, and Y. Bentz, Neural networks in financial engineering: A study in methodology, IEEE Transactions on Neural Networks, vol. 8, no. 6, pp. 1222 1267, 1997. [3] CHEN, Kuan-Yu and Chia-Hui HO, An Improved Support Vector Regression Modeling for Taiwan Stock Exchange Market Weighted Index Forecasting, ICNNB 05: International Conference on Neural Networks and Brain, Volume 3, , 2005 [4] L. Cao and F. Tay, Support Vector Machine with adaptive parameters in financial time series forecasting, IEEE Transactions on Neural Networks, vol. 14, no. 6, pp. 1506-1518, 2003. [5] P.B. Patel and T. Marwala, forecasting closing price indices using neural networks. In International Conference on Systems, Man and Cybernetics, pp. 2351-2356, Oct 8-11, 2006, Taipei, Taiwan. [6] S.H. Lee, H.J. Kim and J.S. Lim, forecasting short term KOSPI time series based on NEWFM, in Advance Language Processing and Web Information Technology (ALPIT), pp. 303-307, July, 2007. [7] B. Doeksen, A. Abraham, J. Thomas, and M. Paprzycki, Real stock trading using soft computing models, in Information Technology: Coding and Computing (ITCC), 2005, vol. 2, pp. 162-167. [8] A.S. Chen, M. T. Leung, and H. Daouk, Application of neural networks to an emerging financial market: Forecasting and trading the Taiwan Stock Index, Computers and Operations Research, vol. 30, no. 6, pp. 901-923, May 2003. [9] K.K. Ang and C. Quek, Stock Trading Using RSPOP: A Novel Rough Set-Based Neuro-Fuzzy Approach, IEEE Transactions on Neural Networks, vol. 17, no.5, pp. 1301 1315, 2006. [10] H.M. Huang, M. Pasquier, and C. Quek, Financial Market Trading System With a Hierarchical Coevolutionary Fuzzy Predictive Model, IEEE Transactions on Evolutionary Computation, vol. 13, no.1, pp. 56 70, 2009. [11] H. Bandy, Quantitative Trading Systems, Blue Owl Press, 2007. [12] B. Krose and P.V.D Smagt, Introduction to Neural Network. The University of Amsterdam, 1996. [13] S. Russell and P. Norvig. Artificial Intelligence A Modern Approach. p. 578. [14] A.E.Bryson and Yu-Chi Ho. Applied optimal control: optimization, estimation, and control. Xerox College Publishing. pp. 481. [15] P.N. Bahrun and M.N. Taib, Selected Malaysia Stock Predictions using Artificial Neural Network, in International Colloquium on Signal Processing Its Applications (CSPA), 2009, pp. 428 431. [16] Lipo Wang (ed.), Support Vector Machines: Theory and Applications. Berlin, Springer, 2005. [17] C.W. Hsu, C.C. Chang, and C.J. Lin, A practical guide to support vector classification, Department of Computer Science and Information Engineering, National Taiwan University, Taipei, Taiwan, 2003. [Online]. Available: https://www.csie.ntu.edu.tw/~cjlin/papers/guide/guide.pdf [18] H. Drucker, C. J.C. Burges, L. Kaufman, A. Smola and V. Vapnik. Support Vector Regression Machines. Advances in Neural Information Processing Systems 9, NIPS 1996, 155-161, MIT Press. [19] A. J. Smola and B. Scholkopf, A tutorial on support vector regression, NeuroCOLT2 Technical Report NC2-TR-1998-030, 2003. [20] John J. Murphy, Technical Analysis of the Financial Markets ,New York Institute of Finance, 1999, pages 1-5,24-31. [21] M. Magdon-Ismail, A. Atiya, Maximum Drawdown, Risk Magazine, Volume 17, Number 10, pp. 99-102, October, 2004. [22] M. Magdon-Ismail, A. Atiya, A. Pratap, Y. Abu-Mostafa, On the Maximum Drawdown of a Brownian Motion, Journal of Applied Probability, Vol. 41, no. 1, PP. 147-161, March, 2004. [23] Appendix The trading details of NN+GA trading system on HSI are listed below. There would be price difference between exit and enter on the same day. This is due to consideration of slippage and commissions. enter short position at price 12414.3 at trading day 45 exit short position at price 13020.8 at trading day 61 enter long position at price 13033.8 at trading day 61 exit long position at price 15598.9 at trading day 138 enter short position at price 15583.3 at trading day 138 exit short position at price 15547.2 at trading day 139 enter long position at price 15562.7 at trading day 139 exit long position at price 15534 at trading day 140 enter short position at price 15518.5 at trading day 140 exit short position at price 14776.8 at trading day 165 enter long position at price 14791.6 at trading day 165 exit long position at price 14810.8 at trading day 166 enter short position at price 14796 at trading day 166 exit short position at price 10525.5 at trading day 244 enter long position at price 10536 at trading day 244 exit long position at price 10232 at trading day 254 enter short position at price 10221.8 at trading day 254 exit short position at price 10671 at trading day 255 enter long position at price 10681.7 at trading day 255 exit long position at price 11151.6 at trading day 295 enter short position at price 11140.4 at trading day 295 exit short position at price 10968.3 at trading day 296 enter long position at price 10979.3 at trading day 296 exit long position at price 10977.5 at trading day 297 enter short position at price 10966.5 at trading day 297 exit short position at price 8189.25 at trading day 395 enter long position at price 8197.44 at trading day 395 exit long position at price 7849.96 at trading day 397 enter short position at price 7842.11 at trading day 397 exit short position at price 7701.61 at trading day 408 enter long position at price 7709.31 at trading day 408 exit long position at price 7946.04 at trading day 409 enter short position at price 7938.09 at trading day 409 exit short position at price 7837.61 at trading day 410 enter long position at price 7845.45 at trading day 410 exit long position at price 7883.46 at trading day 411 enter short position at price 7875.58 at trading day 411 exit short position at price 7564.54 at trading day 412 enter long position at price 7572.1 at trading day 412 exit long position at price 7744.72 at trading day 413 enter short position at price 7736.98 at trading day 413 exit short position at price 8506.79 at trading day 415 enter long position at price 8515.3 at trading day 415 exit long position at price 9499.5 at trading day 488 enter short position at price 9490 at trading day 488 exit short position at price 9913.58 at trading day 511 enter long position at price 9923.49 at trading day 511 exit long position at price 12436.9 at trading day 567 enter short position at price 12424.4 at trading day 567 exit short position at price 12346.9 at trading day 568 enter long position at price 12359.3 at trading day 568 exit long position at price 12409.2 at trading day 569 enter short position at price 12396.8 at trading day 569 exit short position at price 12308.5 at trading day 570 enter long position at price 12320.8 at trading day 570 exit long position at price 12059.3 at trading day 571 enter short position at price 12047.2 at trading day 571 exit short position at price 12471.6 at trading day 575 enter long position at price 12484.1 at trading day 575 exit long position at price 13093.7 at trading day 609 enter short position at price 13080.6 at trading day 609 exit short position at price 13473.8 at trading day 616 enter long position at price 13487.3 at trading day 616 exit long position at price 13591 at trading day 617 enter short position at price 13577.4 at trading day 617 exit short position at price 13254.3 at trading day 618 enter long position at price 13267.6 at trading day 618 exit long position at price 13167.1 at trading day 619 enter short position at price 13153.9 at trading day 619 exit short position at price 13566.7 at trading day 629 enter long position at price 13580.3 at trading day 629 exit long position at price 13214.4 at trading day 652 enter short position at price 13201.2 at trading day 652 exit short position at price 13322.1 at trading day 677 enter long position at price 13335.4 at trading day 677 exit long position at price 15574.6 at trading day 730 enter short position at price 15559 at trading day 730 exit short position at price 15275.3 at trading day 732 enter long position at price 15290.6 at trading day 732 exit long position at price 15167.5 at trading day 735 enter short position at price 15152.4 at trading day 735 exit short position at price 15917.8 at trading day 738 enter long position at price 15933.7 at trading day 738 exit long position at price 15653.9 at trading day 741 enter short position at price 15638.2 at trading day 741 exit short position at price 15789.8 at trading day 742 enter long position at price 15805.6 at trading day 742 exit long position at price 16491.4 at trading day 785 enter short position at price 16474.9 at trading day 785 exit short position at price 16850.7 at trading day 787 enter long position at price 16867.6 at trading day 787 exit long position at price 16487.7 at trading day 788 enter short position at price 16471.2 at trading day 788 exit short position at price 15278.3 at trading day 793 enter long position at price 15293.6 at trading day 793 exit long position at price 15367.1 at trading day 795 enter short position at price 15351.8 at trading day 795 exit short position at price 15900.1 at trading day 824 enter long position at price 15916 at trading day 824 exit long position at price 16629.8 at trading day 893 enter short position at price 16613.2 at trading day 893 exit short position at price 15820.8 at trading day 930 enter long position at price 15836.6 at trading day 930 exit long position at price 15504.8 at trading day 932 enter short position at price 15489.3 at trading day 932 exit short position at price 15329.6 at trading day 955 enter long position at price 15344.9 at trading day 955 exit long position at price 15024.5 at trading day 959 enter short position at price 15009.5 at trading day 959 exit short position at price 15188 at trading day 960 enter long position at price 15203.2 at trading day 960 exit long position at price 14659.3 at trading day 962 enter short position at price 14644.7 at trading day 962 exit short position at price 15436.5 at trading day 971 enter long position at price 15452 at trading day 971 exit long position at price 15527.4 at trading day 999 enter short position at price 15511.8 at trading day 999 exit short position at price 13718.1 at trading day 1046 enter long position at price 13731.9 at trading day 1046 exit long position at price 13600.8 at trading day 1048 enter short position at price 13587.2 at trading day 1048 exit short position at price 13585.1 at trading day 1050 enter long position at price 13598.7 at trading day 1050 exit long position at price 13636.6 at trading day 1052 enter short position at price 13623 at trading day 1052 exit short position at price 13459.2 at trading day 1057 enter long position at price 13472.6 at trading day 1057 exit long position at price 13721.3 at trading day 1058 enter short position at price 13707.5 at trading day 1058 exit short position at price 13878 at trading day 1059 enter long position at price 13891.8 at trading day 1059 exit long position at price 13174.4 at trading day 1066 enter short position at price 13161.2 at trading day 1066 exit short position at price 13703.4 at trading day 1071 enter long position at price 13717.1 at trading day 1071 exit long position at price 13523.3 at trading day 1075 enter short position at price 13509.8 at trading day 1075 exit short position at price 10609.3 at trading day 1176 enter long position at price 10619.9 at trading day 1176 exit long position at price 11209.4 at trading day 1219 enter short position at price 11198.2 at trading day 1219 exit short position at price 11013.6 at trading day 1220 enter long position at price 11024.6 at trading day 1220 exit long position at price 10964.1 at trading day 1221 enter short position at price 10953.1 at trading day 1221 exit short position at price 11003 at trading day 1253 enter long position at price 11014 at trading day 1253 exit long position at price 10863.1 at trading day 1265 enter short position at price 10852.2 at trading day 1265 exit short position at price 11032.9 at trading day 1269 enter long position at price 11044 at trading day 1269 exit long position at price 10878 at trading day 1270 enter short position at price 10867.2 at trading day 1270 exit short position at price 11217.2 at trading day 1281 enter long position at price 11228.4 at trading day 1281 exit long position at price 11359.8 at trading day 1311 enter short position at price 11348.4 at trading day 1311 exit short position at price 11312.5 at trading day 1312 enter long position at price 11323.9 at trading day 1312 exit long position at price 11402.4 at trading day 1313 enter short position at price 11391 at trading day 1313 exit short position at price 9787.49 at trading day 1411 enter long position at price 9797.28 at trading day 1411 exit long position at price 9560.46 at trading day 1415 enter short position at price 9550.9 at trading day 1415 exit short position at price 9655.36 at trading day 1419 enter long position at price 9665.02 at trading day 1419 exit long position at price 9613.84 at trading day 1424 enter short position at price 9604.23 at trading day 1424 exit short position at price 9865.65 at trading day 1427 enter long position at price 9875.52 at trading day 1427 exit long position at price 9656.46 at trading day 1448 enter short position at price 9646.8 at trading day 1448 exit short position at price 9834.08 at trading day 1465 enter long position at price 9843.91 at trading day 1465 exit long position at price 9552.02 at trading day 1470 enter short position at price 9542.47 at trading day 1470 exit short position at price 9155.57 at trading day 1544 enter long position at price 9164.73 at trading day 1544 exit long position at price 13024.1 at trading day 1753 enter short position at price 13011 at trading day 1753 exit short position at price 12326.9 at trading day 1811 enter long position at price 12339.2 at trading day 1811 exit long position at price 12050.7 at trading day 1817 enter short position at price 12038.6 at trading day 1817 exit short position at price 12185.5 at trading day 1824 enter long position at price 12197.7 at trading day 1824 exit long position at price 12285.8 at trading day 1827 enter short position at price 12273.5 at trading day 1827 exit short position at price 12220.1 at trading day 1828 enter long position at price 12232.4 at trading day 1828 exit long position at price 11939.4 at trading day 1837 enter short position at price 11927.5 at trading day 1837 exit short position at price 12123.6 at trading day 1840 enter long position at price 12135.8 at trading day 1840 exit long position at price 12395.1 at trading day 1841 enter short position at price 12382.7 at trading day 1841 exit short position at price 12320.2 at trading day 1842 enter long position at price 12332.5 at trading day 1842 exit long position at price 12852.4 at trading day 1907 enter short position at price 12839.5 at trading day 1907 exit short position at price 13054.7 at trading day 1910 enter long position at price 13067.7 at trading day 1910 exit long position at price 13712 at trading day 1958 enter short position at price 13698.3 at trading day 1958 exit short position at price 13578.3 at trading day 1976 enter long position at price 13591.8 at trading day 1976 exit long position at price 13555.8 at trading day 1977 enter short position at price 13542.2 at trading day 1977 exit short position at price 13845.6 at trading day 1981 enter long position at price 13859.5 at trading day 1981 exit long position at price 13772 at trading day 1997 enter short position at price 13758.2 at trading day 1997 exit short position at price 13941.5 at trading day 1999 enter long position at price 13955.4 at trading day 1999 exit long position at price 13890.9 at trading day 2001 enter short position at price 13877 at trading day 2001 exit short position at price 13906.9 at trading day 2002 enter long position at price 13920.8 at trading day 2002 exit long position at price 13832.5 at trading day 2004 enter short position at price 13818.7 at trading day 2004 exit short position at price 13776.5 at trading day 2008 enter long position at price 13790.2 at trading day 2008 exit long position at price 13603.6 at trading day 2009 enter short position at price 13590 at trading day 2009 exit short position at price 13750.2 at trading day 2029 enter long position at price 13764 at trading day 2029 exit long position at price 13627 at trading day 2044 enter short position at price 13613.4 at trading day 2044 exit short position at price 13867.1 at trading day 2053 enter long position at price 13880.9 at trading day 2053 exit long position at price 14847.8 at trading day 2144 enter short position at price 14832.9 at trading day 2144 exit short position at price 14629.5 at trading day 2169 enter long position at price 14644.1 at trading day 2169 exit long position at price 14627.4 at trading day 2170 enter short position at price 14612.8 at trading day 2170 exit short position at price 14788 at trading day 2172 enter long position at price 14802.8 at trading day 2172 exit long position at price 15542.1 at trading day 2249 enter short position at price 15526.5 at trading day 2249 exit short position at price 15519.8 at trading day 2250 enter long position at price 15535.3 at trading day 2250 exit long position at price 15720.4 at trading day 2251 enter short position at price 15704.6 at trading day 2251 exit short position at price 15729 at trading day 2252 enter long position at price 15744.8 at trading day 2252 exit long position at price 16313.4 at trading day 2294 enter short position at price 16297 at trading day 2294 exit short position at price 15805.5 at trading day 2295 enter long position at price 15821.3 at trading day 2295 exit long position at price 15864.6 at trading day 2296 enter short position at price 15848.7 at trading day 2296 exit short position at price 16326.7 at trading day 2324 enter long position at price 16343 at trading day 2324 27