Cascading logistic regression onto gradient boosted decision trees for forecasting and trading stock indices

Forecasting the direction of the daily changes of stock indices is an important yet difficult task for market participants. Advances on data mining and machine learning make it possible to develop more accurate predictions to assist investment decision making. This paper attempts to develop a learning architecture LR2GBDT for forecasting and trading stock indices, mainly by cascading the logistic regression (LR) model onto the gradient boosted decision trees (GBDT) model. Without any assumption on the underlying data generating process, raw price data and twelve technical indicators are employed for extracting the information contained in the stock indices. The proposed architecture is evaluated by comparing the experimental results with the LR, GBDT, SVM (support vector machine), NN (neural network) and TPOT (tree-based pipeline optimization tool) models on three stock indices data of two different stock markets, which are an emerging market (Shanghai Stock Exchange Composite Index) and a mature stock market (Nasdaq Composite Index and S&P 500 Composite Stock Price Index). Given the same test conditions, the cascaded model not only outperforms the other models, but also shows statistically and economically significant improvements for exploiting simple trading strategies, even when transaction cost is taken into account.     

A Takagi–Sugeno fuzzy model combined with a support vector regression for stock trading forecasting

The turning points prediction scheme for future time series analysis based on past and present information is widely employed in the field of financial applications. In this research, a novel approach to identify turning points of the trading signal using a fuzzy rule-based model is presented. The Takagi–Sugeno fuzzy rule-based model (the TS model) can accurately identify daily stock trading from sets of technical indicators according to the trading signals learned by a support vector regression (SVR) technique. In addition, when new trading points are created, the structure and parameters of the TS model are constantly inherited and updated. To verify the effectiveness of the proposed TS fuzzy rule-based modeling approach, we have acquired the stock trading data in the US stock market. The TS fuzzy approach with dynamic threshold control is compared with a conventional linear regression model and artificial neural networks. Our result indicates that the TS fuzzy model not only yields more profit than other approaches but also enables stable dynamic identification of the complexities of the stock forecasting system.     

Intelligent stock trading system with price trend prediction and reversal recognition using dual-module neural networks

This article presents an intelligent stock trading system that can generate timely stock trading suggestions according to the prediction of short-term trends of price movement using dual-module neural networks(dual net). Retrospective technical indicators extracted from raw price and volume time series data gathered from the market are used as independent variables for neural modeling. Both neural network modules of thedual net learn the correlation between the trends of price movement and the retrospective technical indicators by use of a modified back-propagation learning algorithm. Reinforcing the temporary correlation between the neural weights and the training patterns, dual modules of neural networks are respectively trained on a short-term and a long-term moving-window of training patterns. An adaptive reversal recognition mechanism that can self-tune thresholds for identification of the timing for buying or selling stocks has also been developed in our system. It is shown that the proposeddual net architecture generalizes better than one single-module neural network. According to the features of acceptable rate of returns and consistent quality of trading suggestions shown in the performance evaluation, an intelligent stock trading system with price trend prediction and reversal recognition can be realized using the proposed dual-module neural networks.     

An event based approach for quantifying the effects of securities fraud in the IT industry

Detecting the incidence and impact of illegal insider trading is a difficult process since access to the actual trading records of insiders that overlap precisely with fraudulent events is difficult. This paper provides a case study of a specific IT stock in Canada that was successfully prosecuted in the Canadian court system for market manipulation and illegal insider trading violations. The study provides a quantification of the impact of insider trading activities by the President directly through his own account or through accounts under his control, and illustrates the impact of some off-exchange transactions by the impugned parties. Overall, the costs of the insider trading violations are quite high, given the significant wealth effects produced by the events surrounding this case.     

An adaptive modeling and asset allocation approach to financial markets based on discrete microstructure model

In previous works, it was verified that the discrete-time microstructure (DTMS) model, which is estimated by training dataset of a financial time series, may be effectively applied to asset allocation control on the following test data. However, if the length of test dataset is too long, prediction capability of the estimated DTMS model may gradually decline due to behavior change of financial market, so that the asset allocation result may become worse on the latter part of test data. To overcome the drawback, this paper presents a semi-on-line adaptive modeling and trading approach to financial time series based on the DTMS model and using a receding horizon optimization procedure. First, a long-interval identification window is selected, and the dataset on the identification window is used to estimate a DTMS model, which will be used to do asset allocation on the following short-term trading interval that is referred to as the trading window. After asset allocation is over on the trading window, the length-fixed identification window is then moved to a new window that includes the previous trading window, and a new DTMS model is estimated by using the dataset on the new identification window. Next, asset allocation continues on the next trading window that follows the previous trading window, and then the modeling and asset allocation process will go on according to the above steps. In order to enhance the flexibility and adaptability of the DTMS model, a comprehensive parameter optimization method is proposed, which incorporates particle swarm optimization (PSO) with Kalman filter and maximum likelihood method for estimating the states and parameters of DTMS model. Based on the adaptive DTMS model estimated on each identification window, an adaptive asset allocation control strategy is designed to achieve optimal control of financial assets. The parameters of the asset allocation controller are optimized by the PSO algorithm on each identification window. Case studies on Hang Seng Index (HSI) of Hong Kong stock exchange and S&P 500 index show that the proposed adaptive modeling and trading strategy can obtain much better asset allocation control performance compared with the parameters-fixed DTMS model.     

A Multiagent Approach to Q-Learning for Daily Stock Trading

The portfolio management for trading in the stock market poses a challenging stochastic control problem of significant commercial interests to finance industry. To date, many researchers have proposed various methods to build an intelligent portfolio management system that can recommend financial decisions for daily stock trading. Many promising results have been reported from the supervised learning community on the possibility of building a profitable trading system. More recently, several studies have shown that even the problem of integrating stock price prediction results with trading strategies can be successfully addressed by applying reinforcement learning algorithms. Motivated by this, we present a new stock trading framework that attempts to further enhance the performance of reinforcement learning-based systems. The proposed approach incorporates multiple Q-learning agents, allowing them to effectively divide and conquer the stock trading problem by defining necessary roles for cooperatively carrying out stock pricing and selection decisions. Furthermore, in an attempt to address the complexity issue when considering a large amount of data to obtain long-term dependence among the stock prices, we present a representation scheme that can succinctly summarize the history of price changes. Experimental results on a Korean stock market show that the proposed trading framework outperforms those trained by other alternative approaches both in terms of profit and risk management.     

基于矩阵画像的金融时序数据预测方法

针对金融市场中机构交易对股票市场中的散户投资行为具有较强的误导性的现象,提出了一种基于机构交易行为影响的趋势预测方法。首先,利用时间序列的矩阵画像（MP）方法,以股票换手率数据为切入点,构建不同兴趣模式长度下的基于机构交易行为影响的换手率波动知识库;其次,确定待预测股票在兴趣模式长度取何值时的预测结果精确度高;最后,根据该兴趣模式长度下的知识库,预测在机构交易行为影响下的单支股票的波动趋势。为验证趋势预测新方法的可行性和准确性,将其与自回归滑动平均（ARMA）模型和长短时记忆（LSTM）网络这两种预测方法进行对比分析,运用均方根误差（RMSE）与平均绝对百分误差（MAPE）评价指标综合比较3种方法对70支股票的预测结果。实验结果分析表明,与ARMA模型和LSTM网络相比,在70支的股票价格趋势预测上,所提方法有80%以上的股票预测结果更准确。     

A hybrid evolutionary dynamic neural network for stock market trend analysis and prediction using unscented Kalman filter

Stock market prediction is of great interest to stock traders and investors due to high profit in trading the stocks. A successful stock buying/selling generally occurs near price trend turning point. Thus the prediction of stock market indices and its analysis are important to ascertain whether the next day's closing price would increase or decrease. This paper, therefore, presents a simple IIR filter based dynamic neural network (DNN) and an innovative optimized adaptive unscented Kalman filter for forecasting stock price indices of four different Indian stocks, namely the Bombay stock exchange (BSE), the IBM stock market, RIL stock market, and Oracle stock market. The weights of the dynamic neural information system are adjusted by four different learning strategies that include gradient calculation, unscented Kalman filter (UKF), differential evolution (DE), and a hybrid technique (DEUKF) by alternately executing the DE and UKF for a few generations. To improve the performance of both the UKF and DE algorithms, adaptation of certain parameters in both these algorithms has been presented in this paper. After predicting the stock price indices one day to one week ahead time horizon, the stock market trend has been analyzed using several important technical indicators like the moving average (MA), stochastic oscillators like K and D parameters, WMS%R (William indicator), etc. Extensive computer simulations are carried out with the four learning strategies for prediction of stock indices and the up or down trends of the indices. From the results it is observed that significant accuracy is achieved using the hybrid DEUKF algorithm in comparison to others that include only DE, UKF, and gradient descent technique in chronological order. Comparisons with some of the widely used neural networks (NNs) are also presented in the paper.     

A hybrid genetic-neural architecture for stock indexes forecasting

In this paper, a new approach for time series forecasting is presented. The forecasting activity results from the interaction of a population of experts, each integrating genetic and neural technologies. An expert of this kind embodies a genetic classifier designed to control the activation of a feedforward artificial neural network for performing a locally scoped forecasting activity. Genetic and neural components are supplied with different information: The former deal with inputs encoding information retrieved from technical analysis, whereas the latter process other relevant inputs, in particular past stock prices. To investigate the performance of the proposed approach in response to real data, a stock market forecasting system has been implemented and tested on two stock market indexes, allowing for account realistic trading commissions. The results pointed to the good forecasting capability of the approach, which repeatedly outperformed the “Buy and Hold” strategy.     

Trend following deep Q-Learning strategy for stock trading

Computers and algorithms are widely used to help in stock market decision making. A few questions with regards to the profitability of algorithms for stock trading are can computers be trained to beat the markets? Can an algorithm take decisions for optimal profits? And so forth. In this research work, our objective is to answer some of these questions. We propose an algorithm using deep Q-Reinforcement Learning techniques to make trading decisions. Trading in stock markets involves potential risk because the price is affected by various uncertain events ranging from political influences to economic constraints. Models that trade using predictions may not always be profitable mainly due to the influence of various unknown factors in predicting the future stock price. Trend Following is a trading idea in which, trading decisions, like buying and selling, are taken purely according to the observed market trend. A stock trend can be up, down, or sideways. Trend Following does not predict the stock price but follows the reversals in the trend direction. A trend reversal can be used to trigger a buy or a sell of a certain stock. In this research paper, we describe a deep Q-Reinforcement Learning agent able to learn the Trend Following trading by getting rewarded for its trading decisions. Our results are based on experiments performed on the actual stock market data of the American and the Indian stock markets. The results indicate that the proposed model outperforms forecasting-based methods in terms of profitability. We also limit risk by confirming trading actions with the trend before actual trading.     

A Forex trading expert system based on a new approach to the rule-base evidential reasoning

Currently FOREX (foreign exchange market) is the largest financial market over the world. Usually the Forex market analysis is based on the Forex time series prediction. Nevertheless, trading expert systems based on such predictions do not usually provide satisfactory results. On the other hand, stock trading expert systems called also “mechanical trading systems”, which are based on the technical analysis, are very popular and may provide good profits. Therefore, in this paper we propose a Forex trading expert system based on some new technical analysis indicators and a new approach to the rule-base evidential reasoning (RBER) (the synthesis of fuzzy logic and the Dempster–Shafer theory of evidence). We have found that the traditional fuzzy logic rules lose an important information, when dealing with the intersecting fuzzy classes, e.g., such as Low and Medium and we have shown that this property may lead to the controversial results in practice. In the framework of the proposed in the current paper new approach, an information of the values of all membership functions representing the intersecting (competing) fuzzy classes is preserved and used in the fuzzy logic rules. The advantages of the proposed approach are demonstrated using the developed expert system optimized and tested on the real data from the Forex market for the four currency pairs and the time frames 15 m, 30 m, 1 h and 4 h.     

基于深度迁移学习的多尺度股票预测

股票市场不仅是上市公司的重要融资渠道,也是重要的投资市场,股票预测一直受到人们的关注。为了充分利用来自不同股票价格的信息,提高股票的预测效果,提出一种多尺度股票价格预测模型TL-EMD-LSTM-MA（TELM）。TELM模型通过经验模态分解将收盘价分解为多个时间尺度分量,不同时间尺度分量震荡频率不同,反映了不同的周期性信息;根据分量的震荡频率选择不同方法进行预测,高频分量利用深度迁移学习的方法训练堆叠LSTM,低频分量利用移动平均法进行预测;将所有分量的预测值相加作为收盘价的最终预测输出。通过深度迁移学习训练的堆叠LSTM,包含来自不同股票的信息,具备更多行业或市场的知识,能有效降低预测误差。利用移动平均法预测低频分量,更有效捕获股票的总体趋势。对中国A股市场内500支股票以及上证指数、深证成指等指数进行预测,结果表明,与其他模型相比,TELM预测误差最低,拟合优度最高。根据TELM预测的股票收盘价模拟股票交易过程,结果表明TELM投资风险低、收益高。     

An improved differential evolution trained neural network scheme for nonlinear system identification

This paper presents an improved nonlinear system identification scheme using di？erential evolution （DE）, neural network （NN） and Levenberg Marquardt algorithm （LM）. With a view to achieve better convergence of NN weights optimization during the training, the DE and LM are used in a combined framework to train the NN. We present the convergence analysis of the DE and demonstrate the efficacy of the proposed improved system identification algorithm by exploiting the combined DE and LM training of the NN and suitably implementing it together with other system identification methods, namely NN and DE＋NN on a number of examples including a practical case study. The identification results obtained through a series of simulation studies of these methods on different nonlinear systems demonstrate that the proposed DE and LM trained NN approach to nonlinear system identification can yield better identification results in terms of time of convergence and less identification error.     

An intelligent hybrid trading system for discovering trading rules for the futures market using rough sets and genetic algorithms

Discovering intelligent technical trading rules from nonlinear and complex stock market data, and then developing decision support trading systems, is an important challenge. The objective of this study is to develop an intelligent hybrid trading system for discovering technical trading rules using rough set analysis and a genetic algorithm (GA). In order to obtain better trading decisions, a novel rule discovery mechanism using a GA approach is proposed for solving optimization problems (i.e., data discretization and reducts) of rough set analysis when discovering technical trading rules for the futures market. Experiments are designed to test the proposed model against comparable approaches (i.e., random, correlation, and GA approaches). In addition, these comprehensive experiments cover most of the current trading system topics, including the use of a sliding window method (with or without validation dataset), the number of trading rules, and the size of training period. To evaluate an intelligent hybrid trading system, experiments were carried out on the historical data of the Korea Composite Stock Price Index 200 (KOSPI 200) futures market. In particular, trading performance is analyzed according to the number of sets of decision rules and the size of the training period for discovering trading rules for the testing period. The results show that the proposed model significantly outperforms the benchmark model in terms of the average return and as a risk-adjusted measure.     

Discovery of trading points based on Bayesian modeling of trading rules

Mining hidden patterns with different technical indicators from the historical financial data has been regarded as an efficient way to determine the trading decisions in the financial market. Technical analysis has shown that a number of specific combinations of technical indicators could be treated as trading patterns for forecasting efficient trading directions. However, it is a challenging assignment to discover those combinations. In this paper, we innovatively propose to use a biclustering algorithm to detect the trading patterns. The discovered trading patterns are then utilized to forecast the market movement based on the Naive Bayesian algorithm. Finally, the Adaboost algorithm is applied to improve the accuracy of the forecasts. The proposed method was implemented on seven historical stock datasets and the average performance was compared with that of four existing algorithms. Experimental results demonstrated that the proposed algorithm outperforms the other four algorithms and can provide a valuable reference in the financial investments.     

Differential Evolution and Levenberg Marquardt Trained Neural Network Scheme for Nonlinear System Identification

This paper proposes a new nonlinear system identification scheme using differential evolution (DE), neural network and Levenberg Marquardt algorithm (LM). Here, DE and LM in a combined framework are used to train a neural network for achieving better convergence of neural network weight optimization. A number of examples including a practical case-study have been considered for implementation of different system identification methods namely, only NN, DE+NN and DE+LM+NN. After, a series of simulation studies of these methods on the different nonlinear systems it has been confirmed that the proposed DE and LM trained NN approach to nonlinear system identification has yielded better identification results in terms of time of convergence and less identification error.     

A portfolio optimization model using Genetic Network Programming with control nodes

Many evolutionary computation methods applied to the financial field have been reported. A new evolutionary method named “Genetic Network Programming” (GNP) has been developed and applied to the stock market recently. The efficient trading rules created by GNP has been confirmed in our previous research. In this paper a multi-brands portfolio optimization model based on Genetic Network Programming with control nodes is presented. This method makes use of the information from technical indices and candlestick chart. The proposed optimization model, consisting of technical analysis rules, are trained to generate trading advice. The experimental results on the Japanese stock market show that the proposed optimization system using GNP with control nodes method outperforms other traditional models in terms of both accuracy and efficiency. We also compared the experimental results of the proposed model with the conventional GNP based methods, GA and Buy&Hold method to confirm its effectiveness, and it is clarified that the proposed trading model can obtain much higher profits than these methods.     

Integrating metaheuristics and Artificial Neural Networks for improved stock price prediction

Stock market price is one of the most important indicators of a country's economic growth. That's why determining the exact movements of stock market price is considerably regarded. However, complex and uncertain behaviors of stock market make exact determination impossible and hence strong forecasting models are deeply desirable for investors' financial decision making process. This study aims at evaluating the effectiveness of using technical indicators, such as simple moving average of close price, momentum close price, etc. in Turkish stock market. To capture the relationship between the technical indicators and the stock market for the period under investigation, hybrid Artificial Neural Network (ANN) models, which consist in exploiting capabilities of Harmony Search (HS) and Genetic Algorithm (GA), are used for selecting the most relevant technical indicators. In addition, this study simultaneously searches the most appropriate number of hidden neurons in hidden layer and in this respect; proposed models mitigate well-known problem of overfitting/underfitting of ANN. The comparison for each proposed model is done in four viewpoints: loss functions, return from investment analysis, buy and hold analysis, and graphical analysis. According to the statistical and financial performance of these models, HS based ANN model is found as a dominant model for stock market forecasting.     

大数据分析的应用案例——投资模型的稳健性

交易模型的稳健性,指的是该模型的利润率曲线的波动性较小,没有大起大落。针对一个基于支持向量回归（SVR）技术的算法交易模型的稳健性问题,提出了使用若干导出指标训练统一的交易模型的策略,以及投资组合多样化的方法。首先,介绍基于支持向量回归技术的算法交易模型;然后,基于常用指标,构造了若干导出指标,用于股票价格的短期预测。这些指标,刻画了近期价格运动的典型模式、超买/超卖市场状态,以及背离市场状态。对这些指标进行了规范化,用于训练交易模型,使得模型可以泛化到不同的股票;最后,设计了投资组合多样化方法。在投资组合里,各个股票之间的相关性,有时会导致较大的投资损失;因为具有较强相关关系的股票,其价格朝相同方向变化。如果交易模型预测的价格走势不正确,引起止损操作,那么这些具有较强相关关系的股票,将引发雪崩式的止损,于是导致损失加剧。把股票根据相似性聚类到不同类别,通过从不同聚类类别中选择若干股票来构成多样化的投资组合,其中,股票的相似性,通过交易模型在不同股票上近期的利润曲线的相似度进行计算。在900只股票10年的价格大数据上进行了实验,实验结果显示,交易模型能够获得超过定期存款的超额利润率,年化利润率为8.06%。交易模型的最大回撤由13.23%降为5.32%,夏普指数由81.23%提高到88.79%,交易模型的利润率曲线波动性降低,说明交易模型的稳健性获得了提高。     

Grammatical Evolution-based ensembles for algorithmic trading

The literature on trading algorithms based on Grammatical Evolution commonly presents solutions that rely on static approaches. Given the prevalence of structural change in financial time series, that implies that the rules might have to be updated at predefined time intervals. We introduce an alternative solution based on an ensemble of models which are trained using a sliding window. The structure of the ensemble combines the flexibility required to adapt to structural changes with the need to control for the excessive transaction costs associated with over-trading. The performance of the algorithm is benchmarked against five different comparable strategies that include the traditional static approach, the generation of trading rules that are used for single time period and are subsequently discarded, and three alternatives based on ensembles with different voting schemes. The experimental results, based on market data, show that the suggested approach offers very competitive results against comparable solutions and highlight the importance of containing transaction costs.