Conference Paper

A General Framework for Risk Controlled Trading Based on Machine Learning and Statistical Arbitrage

Authors:
To read the full-text of this research, you can request a copy directly from the authors.

Abstract

Nowadays, machine learning usage has gained significant interest in financial time series prediction, hence being a promise land for financial applications such as algorithmic trading. In this setting, this paper proposes a general framework based on an ensemble of regression algorithms and dynamic asset selection applied to the well known statistical arbitrage trading strategy. Several extremely heterogeneous state-of-the-art machine learning algorithms, exploiting different feature selection processes in input, are used as base components of the ensemble, which is in charge to forecast the return of each of the considered stocks. Before being used as an input to the arbitrage mechanism, the final ranking of the assets takes also into account a quality assurance mechanism that prunes the stocks with poor forecasting accuracy in the previous periods. The framework has a general application for any risk balanced trading strategy aiming to exploit different financial assets. It was evaluated implementing an intra-day trading statistical arbitrage on the stocks of the S&P500 index. Our approach outperforms each single base regressor we adopted, which we considered as baselines. More important, it also outperforms Buy-and-hold of S&P500 Index, both during financial turmoil such as the global financial crisis, and also during the massive market growth in the recent years.

No full-text available

Request Full-text Paper PDF

To read the full-text of this research,
you can request a copy directly from the authors.

... They are extensively used within the academic literature to either generate trading signals (Pimenta et al., 2018) or as input features for ML models (Kara et al., 2011;Patel et al., 2015a,b). To construct the feature vector x t , similarly to (Carta et al., 2020;Carta et al., 2021b), we use the same set of nine technical indicators: Exponential Moving Average (EMA (10) (10)). To compute the feature vector of technical indicators for a day t, we use financial information over the past days to t (each indicator needs a different number of previous days to t, according to their formula). ...
... This fact makes LGB quite effective in processing large-scale and high-dimensional data, with the downside of being more prone to over-fitting. In their work, authors of (Carta et al., 2020;Carta et al., 2021b) use LGB to predict daily price returns. ...
Article
Full-text available
Machine learning techniques have recently become the norm for detecting patterns in financial markets. However, relying solely on machine learning algorithms for decision-making can have negative consequences, especially in a critical domain such as the financial one. On the other hand, it is well-known that transforming data into actionable insights can pose a challenge even for seasoned practitioners, particularly in the financial world. Given these compelling reasons, this work proposes a machine learning approach powered by eXplainable Artificial Intelligence techniques integrated into a statistical arbitrage trading pipeline. Specifically, we propose three methods to discard irrelevant features for the prediction task. We evaluate the approaches on historical data of component stocks of the S&P500 index and aim at improving not only the prediction performance at the stock level but also overall at the stock set level. Our analysis shows that our trading strategies that include such feature selection methods improve the portfolio performances by providing predictive signals whose information content suffices and is less noisy than the one embedded in the whole feature set. By performing an in-depth risk-return analysis, we show that the proposed trading strategies powered by explainable AI outperform highly competitive trading strategies considered as baselines.
... Several literature works have shown that Deep Learning approaches can be extremely powerful in order to tackle classification problems in various domains, including Natural Language Processing [6,24] , Computer Vision [16,31], Sentiment Analysis [2,7], Human-Robot Interaction [1,5]. Among these, Financial Technology (fintech) is a further field where these approaches have recently begun to be applied [4,8,9,[11][12][13]21]. However, financial markets are influenced by several factors, many of which are difficult to predict: news, investor mood, wars and many other events that rarely happen as the COVID-19 pandemic [25,37]. ...
Conference Paper
Financial forecasting represents a challenging task, mainly due to the irregularity of the market, high fluctuations and noise of the involved data, as well as collateral phenomena including investor mood and mass psychology. In recent years, many researchers focused their work on predicting the performance of the market by exploiting novel Machine Learning and Deep Learning tools and techniques. However, many of the approaches proposed in the literature do not take adequately into account some important specific domain issues in the analysis of the results. Among these, it is worth to mention the bias introduced by the choice of model weights initialization and the considered observation periods, as well as the narrow separation between significant results and noise, typical of the financial domain. A thorough analysis of these peculiar issues lead to a substantial increase of the experiments and results to analyze, making the discovery of meaningful hidden patterns very difficult and time consuming to perform. To cope with these concerns and accompanying the current Machine Learning Interpretability trend, in this paper we propose a visual framework for in-depth analysis of results obtained from Deep Learning approaches, tackling classification tasks within the financial domain and aiming at a better interpretation and explanation of the trained Deep Learning models. Our framework offers a modular view, both general and targeted, of results data, providing several financial specific metrics, including Sharpe and Sortino ratios, Equity curves and Maximum Drawdown, as well as custom period analysis and reports, experiment comparison tools, and evaluation features for different algorithms.
Article
Full-text available
Financial markets forecasting represents a challenging task for a series of reasons, such as the irregularity, high fluctuation, noise of the involved data, and the peculiar high unpredictability of the financial domain. Moreover, literature does not offer a proper methodology to systematically identify intrinsic and hyper-parameters, input features, and base algorithms of a forecasting strategy in order to automatically adapt itself to the chosen market. To tackle these issues, this paper introduces a fully automated optimized ensemble approach, where an optimized feature selection process has been combined with an automatic ensemble machine learning strategy, created by a set of classifiers with intrinsic and hyper-parameters learned in each marked under consideration. A series of experiments performed on different real-world futures markets demonstrate the effectiveness of such an approach with regard to both to the Buy and Hold baseline strategy and to several canonical state-of-the-art solutions.
Article
Full-text available
Lenders, such as credit card companies and banks, use credit scores to evaluate the potential risk posed by lending money to consumers and, therefore, mitigating losses due to bad debt. Within the financial technology domain, an ideal approach should be able to operate proactively, without the need of knowing the behavior of non-reliable users. Actually, this does not happen because the most used techniques need to train their models with both reliable and non-reliable data in order to classify new samples. Such a scenario might be affected by the cold-start problem in datasets, where there is a scarcity or total absence of non-reliable examples, which is further worsened by the potential unbalanced distribution of the data that reduces the classification performances. In this paper, we overcome the aforementioned issues by proposing a proactive approach, composed of a combined entropy-based method that is trained considering only reliable cases and the sample under investigation. Experiments done in different real-world datasets show competitive performances with several state-of-art approaches that use the entire dataset of reliable and unreliable cases.
Article
Full-text available
We study model-driven statistical arbitrage strategies in U.S. equities. Trading signals are generated in two ways: using Principal Component Analysis and using sector ETFs. In both cases, we consider the residuals, or idiosyncratic components of stock returns, and model them as a mean-reverting process, which leads naturally to "contrarian" trading signals. The main contribution of the paper is the back-testing and comparison of market-neutral PCA-and ETF-based strategies over the broad universe of U.S. equities. Back-testing shows that, after accounting for transaction costs, PCA-based strategies have an average annual Sharpe ratio of 1.44 over the period 1997 to 2007, with a much stronger performances prior to 2003: during 2003-2007, the average Sharpe ratio of PCA-based strategies was only 0.9. On the other hand, strategies based on ETFs achieved a Sharpe ratio of 1.1 from 1997 to 2007, but experience a similar degradation of performance after 2002. We introduce a method to take into account daily trading volume information in the signals (using "trading time" as opposed to calendar time), and observe significant improvements in performance in the case of ETF-based signals. ETF strategies which use volume information achieve a Sharpe ratio of 1.51 from 2003 to 2007. The paper also relates the performance of mean-reversion statistical arbitrage strategies with the stock market cycle. In particular, we study in some detail the performance of the strategies during the liquidity cri-sis of the summer of 2007. We obtain results which are consistent with Khandani and Lo (2007) and validate their "unwinding" theory for the quant fund drawndown of August 2007.
Article
Full-text available
This paper comes together with growing evidence for power law-type scaling of correlation with time, a concept rooted in the original study by Benoit Mandelbrot on concentration of risk. We complete the cointelation model recently introduced via a statistical test, which uses measured correlation in different time gaps. We also provide an approximation of the expectation in the change in measured correlation via these various time steps and use our findings in order to introduce the concept of inferred correlation and the term structure of correlation. We finally illustrate our findings through the example of the relation between oil and BP, and present a few potential applications in the financial industry.
Conference Paper
Full-text available
Stock price prediction is an important topic in finance and economics which has spurred the interest of researchers over the years to develop better predictive models. The autoregressive integrated moving average (ARIMA) models have been explored in literature for time series prediction. This paper presents extensive process of building stock price predictive model using the ARIMA model. Published stock data obtained from New York Stock Exchange (NYSE) and Nigeria Stock Exchange (NSE) are used with stock price predictive model developed. Results obtained revealed that the ARIMA model has a strong potential for short-term prediction and can compete favourably with existing techniques for stock price prediction.
Article
Full-text available
The purpose of this study is to compare the forecasting performance of a neural network (NN) model and a time-series (SARIMA) model in Korean Stock Exchange. In particular, we investigate whether the back-propagation neural network (BPNN)model outperforms the seasonal autoregressive integrated moving average (SARIMA) model in forecasting the Korea Composite Stock Price Index (KOSPI) and its return. Forecasting performance is ev1auated by the forecasting accuracy of each model.
Article
Full-text available
Prediction of stock price index movement is regarded as a challenging task of financial time series prediction. An accurate prediction of stock price movement may yield profits for investors. Due to the complexity of stock market data, development of efficient models for predicting is very difficult. This study attempted to develop two efficient models and compared their performances in predicting the direction of movement in the daily Istanbul Stock Exchange (ISE) National 100 Index. The models are based on two classification techniques, artificial neural networks (ANN) and support vector machines (SVM). Ten technical indicators were selected as inputs of the proposed models. Two comprehensive parameter setting experiments for both models were performed to improve their prediction performances. Experimental results showed that average performance of ANN model (75.74%) was found significantly better than that of SVM model (71.52%).
Article
Full-text available
Ensembles are a widely used and effective technique in machine learning – their success is commonly attributed to the degree of disagreement, or ‘diversity’, within the ensemble. For ensembles where the individual estimators output crisp class labels, this ‘diversity’ is not well understood and remains an open research issue. For ensembles of regression estimators, the diversity can be exactly formulated in terms of the covariance between individual estimator outputs, and the optimum level is expressed in terms of a bias-variance-covariance trade-off. Despite this, most approaches to learning ensembles use heuristics to encourage the right degree of diversity. In this work we show how to explicitly control diversity through the error function. The first contribution of this paper is to show that by taking the combination mechanism for the ensemble into account we can derive an error function for each individual that balances ensemble diversity with individual accuracy. We show the relationship between this error function and an existing algorithm called negative correlation learning, which uses a heuristic penalty term added to the mean squared error function. It is demonstrated that these methods control the bias-variance-covariance trade-off systematically, and can be utilised with any estimator capable of minimising a quadratic error function, for example MLPs, or RBF networks. As a second contribution, we derive a strict upper bound on the coefficient of the penalty term, which holds for any estimator that can be cast in a generalised linear regression framework, with mild assumptions on the basis functions. Finally we present the results of an empirical study, showing significant improvements over simple ensemble learning, and finding that this technique is competitive with a variety of methods, including boosting, bagging, mixtures of experts, and Gaussian processes, on a number of tasks.
Article
Full-text available
We test a Wall Street investment strategy, “pairs trading,” with daily data over 1962–2002. Stocks are matched into pairs with minimum distance between normalized historical prices. A simple trading rule yields average annualized excess returns of up to 11% for self-financing portfolios of pairs. The profits typically exceed conservative transaction-cost estimates. Bootstrap results suggest that the “pairs” effect differs from previously documented reversal profits. Robustness of the excess returns indicates that pairs trading profits from temporary mispricing of close substitutes. We link the profitability to the presence of a common factor in the returns, different from conventional risk measures.
Article
Over the past 15 years, there have been a number of studies using text mining for predicting stock market data. Two recent publications employed support vector machines and second-order Factorization Machines, respectively, to this end. However, these approaches either completely neglect interactions between the features extracted from the text, or they only account for second-order interactions. In this paper, we apply higher-order Factorization Machines, for which efficient training algorithms have only been available since 2016. As Factorization Machines require hyperparameters to be specified, we also introduce a novel adaptive-order algorithm for automatically determining them. Our study is the first one to make use of social media data for predicting minute-by-minute stock returns, namely the ones of the S&P 500 stock constituents. We show that, unlike a trading strategy employing support vector machines, Factorization-Machine-based strategies attain positive returns after transactions costs for the years 2014 and 2015. Especially the approach applying the adaptive-order algorithm outperforms classical approaches with respect to a multitude of criteria, and it features very favorable characteristics.
Article
This paper presents a state-of-the-art approach for sentiment polarity classification. Our approach relies on an ensemble of Bidirectional Long Short-Term Memory networks equipped with a neural attention mechanism. The system makes use of pre-trained word embeddings, and is capable of predicting new vectors for out-of-vocabulary words, by learning distributional representations based on word spellings. Also, during the training process the recurrent neural network is used to perform a fine-tuning of the original word embeddings, taking into account information about sentiment polarity. This step can be particularly helpful for sentiment analysis, as word embeddings are usually built based on context information, while words with opposite sentiment polarity often occur in similar contexts. The system described in this paper is an improved version of an approach that competed in a recent challenge on semantic sentiment analysis. We evaluate the performance of the system on the same multi-domain test set used by the organizers of the challenge, showing that our approach allows reaching better results with respect to the previous top-scoring system. Last but not least, we embedded the proposed sentiment polarity approach on top of a humanoid robot to lively identify the sentiment of the speaking user.
Article
Machine learning algorithms and big data are transforming all industries including the finance and portfolio management sectors. While these techniques, such as Deep Belief Networks or Random Forests, are becoming more and more popular on the market, the academic literature is relatively sparse. Through a series of applications involving hundreds of variables/predictors and stocks, this article presents some of the state-of-the-art techniques and how they can be implemented to manage a long-short portfolio. Numerous practical and empirical issues are developed. One of the main questions beyond big data use is the value of information. Does an increase in the number of predictors improve the portfolio performance? Which features are the most important? A large number of predictors means, potentially, a high level of noise. How do the algorithms manage this? This article develops an application using a 22-year trading period, up to 300 U.S. large caps and around 600 predictors. The empirical results underline the ability of these techniques to generate useful trading signals for portfolios with important turnovers and short holding periods (one or five days). Positive excess returns are reported between 1993 and 2008. They are strongly reduced after accounting for transaction costs and traditional risk factors. When these machine learning tools were readily available in the market, excess returns turned into the negative in most recent times. Results also show that adding features is far from being a guarantee to boost the alpha of the portfolio.
Article
The search for models to predict the prices of financial markets is still a highly researched topic, despite major related challenges. The prices of financial assets are non-linear, dynamic, and chaotic; thus, they are financial time series that are difficult to predict. Among the latest techniques, machine learning models are some of the most researched, given their capabilities for recognizing complex patterns in various applications. With the high productivity in the machine learning area applied to the prediction of financial market prices, objective methods are required for a consistent analysis of the most relevant bibliography on the subject. This article proposes the use of bibliographic survey techniques that highlight the most important texts for an area of research. Specifically, these techniques are applied to the literature about machine learning for predicting financial market values, resulting in a bibliographical review of the most important studies about this topic. Fifty-seven texts were reviewed, and a classification was proposed for markets, assets, methods, and variables. Among the main results, of particular note is the greater number of studies that use data from the North American market. The most commonly used models for prediction involve support vector machines (SVMs) and neural networks. It was concluded that the research theme is still relevant and that the use of data from developing markets is a research opportunity.
Book
Effectively Access, Transform, Manipulate, Visualize, and Reason about Data and Computation Data Science in R: A Case Studies Approach to Computational Reasoning and Problem Solving illustrates the details involved in solving real computational problems encountered in data analysis. It reveals the dynamic and iterative process by which data analysts approach a problem and reason about different ways of implementing solutions. The book’s collection of projects, comprehensive sample solutions, and follow-up exercises encompass practical topics pertaining to data processing, including: • Non-standard, complex data formats, such as robot logs and email messages • Text processing and regular expressions • Newer technologies, such as Web scraping, Web services, Keyhole Markup Language (KML), and Google Earth • Statistical methods, such as classification trees, k-nearest neighbors, and naïve Bayes • Visualization and exploratory data analysis • Relational databases and Structured Query Language (SQL) • Simulation • Algorithm implementation • Large data and efficiency Suitable for self-study or as supplementary reading in a statistical computing course, the book enables instructors to incorporate interesting problems into their courses so that students gain valuable experience and data science skills. Students learn how to acquire and work with unstructured or semistructured data as well as how to narrow down and carefully frame the questions of interest about the data. Blending computational details with statistical and data analysis concepts, this book provides readers with an understanding of how professional data scientists think about daily computational tasks. It will improve readers’ computational reasoning of real-world data analyses.
Article
Long short-term memory (LSTM) networks are a state-of-the-art technique for sequence learning. They are less commonly applied to financial time series predictions, yet inherently suitable for this domain. We deploy LSTM networks for predicting out-of-sample directional movements for the constituent stocks of the S&P 500 from 1992 until 2015. With daily returns of 0.46 percent and a Sharpe ratio of 5.8 prior to transaction costs, we find LSTM networks to outperform memory-free classification methods, i.e., a random forest (RAF), a deep neural net (DNN), and a logistic regression classifier (LOG). The outperformance relative to the general market is very clear from 1992 to 2009, but as of 2010, excess returns seem to have been arbitraged away with LSTM profitability fluctuating around zero after transaction costs. We further unveil sources of profitability, thereby shedding light into the black box of artificial neural networks. Specifically, we find one common pattern among the stocks selected for trading - they exhibit high volatility and a short-term reversal return profile. Leveraging these findings, we are able to formalize a rules-based short-term reversal strategy that yields 0.23 percent prior to transaction costs. Further regression analysis unveils low exposure of the LSTM returns to common sources of systematic risk - also compared to the three benchmark models.
Article
Deep neural networks (DNNs) are powerful types of artificial neural networks (ANNs) that use several hidden layers. They have recently gained considerable attention in the speech transcription and image recognition community (Krizhevsky et al., 2012) for their superior predictive properties including robustness to overfitting. However their application to algorithmic trading has not been previously researched, partly because of their computational complexity. This paper describes the application of DNNs to predicting financial market movement directions. In particular we describe the configuration and training approach and then demonstrate their application to backtesting a simple trading strategy over 43 different Commodity and FX future mid-prices at 5-minute intervals. All results in this paper are generated using a C implementation on the Intel Xeon Phi co-processor which is 11.4x faster than the serial version and a Python strategy backtesting environment both of which are available as open source code written by the authors.
Article
Building classification models is an intrinsically practical exercise that requires many design decisions prior to deployment. We aim to provide some guidance in this decision making process. Specifically, given a classification problem with real valued attributes, we consider which classifier or family of classifiers should one use. Strong contenders are tree based homogeneous ensembles, support vector machines or deep neural networks. All three families of model could claim to be state-of-the-art, and yet it is not clear when one is preferable to the others. Our extensive experiments with over 200 data sets from two distinct archives demonstrate that, rather than choose a single family and expend computing resources on optimising that model, it is significantly better to build simpler versions of classifiers from each family and ensemble. We show that the Heterogeneous Ensembles of Standard Classification Algorithms (HESCA), which ensembles based on error estimates formed on the train data, is significantly better (in terms of error, balanced error, negative log likelihood and area under the ROC curve) than its individual components, picking the component that is best on train data, and a support vector machine tuned over 1089 different parameter configurations. We demonstrate HESCA+, which contains a deep neural network, a support vector machine and two decision tree forests, is significantly better than its components, picking the best component, and HESCA. We analyse the results further and find that HESCA and HESCA+ are of particular value when the train set size is relatively small and the problem has multiple classes. HESCA is a fast approach that is, on average, as good as state-of-the-art classifiers, whereas HESCA+ is significantly better than average and represents a strong benchmark for future research.
Article
During the week of August 6, 2007, a number of quantitative long/short equity hedge funds experienced unprecedented losses. It has been hypothesized that a coordinated deleveraging of similarly constructed portfolios caused this temporary dislocation in the market. Using the simulated returns of long/short equity portfolios based on five specific valuation factors, we find evidence that the unwinding of these portfolios began in July 2007 and continued until the end of 2007. Using transactions data, we find that the simulated returns of a simple marketmaking strategy were significantly negative during the week of August 6, 2007, but positive before and after, suggesting that the Quant Meltdown of August 2007 was the combined effects of portfolio deleveraging throughout July and the first week of August, and a temporary withdrawal of marketmaking risk capital starting August 8th. Our simulations point to two unwinds---a mini-unwind on August 1st starting at 10:45am and ending at 11:30am, and a more sustained unwind starting at the open on August 6th and ending at 1:00pm---that began with stocks in the financial sector and long Book-to-Market and short Earnings Momentum. These conjectures have significant implications for the systemic risks posed by the hedge-fund industry.
Article
In recent years, machine learning research has gained momentum: New developments in the field of deep learning allow for multiple levels of abstraction and are starting to supersede well-known and powerful tree-based techniques mainly operating on the original feature space. All these methods can be applied to various fields, including finance. This article implements and analyses the effectiveness of deep neural networks (DNN), gradient-boosted-trees (GBT), random forests (RAF), and several ensembles of these methods in the context of statistical arbitrage. Each model is trained on lagged returns of all stocks in the S&P 500, after elimination of survivor bias. From 1992 to 2015, daily one-day-ahead trading signals are generated based on the probability forecast of a stock to outperform the general market. The highest k probabilities are converted into long and the lowest k probabilities into short positions, thus censoring the less certain middle part of the ranking. Empirical findings are promising. A simple, equal-weighted ensemble (ENS1) consisting of one deep neural network, one gradient-boosted tree, and one random forest produces out-of-sample returns exceeding 0.45 percent per day for prior to transaction costs. Irrespective of the fact that profits are declining in recent years, our findings pose a severe challenge to the semi-strong form of market efficiency.
Article
The prequential approach is founded on the premises that the purpose of statistical inference is to make sequential probability forecasts for future observations, rather than to express information about parameters. Many traditional parametric concepts, such as consistency and efficiency, prove to have natural counterparts in this formulation, which sheds new light on these and suggests fruitful extensions.
Article
It has been widely accepted by many studies that non-linearity exists in the financial markets and that neural networks can be effectively used to uncover this relationship. Unfortunately, many of these studies fail to consider alternative forecasting techniques, the relevance of input variables, or the performance of the models when using different trading strategies. This paper introduces an information gain technique used in machine learning for data mining to evaluate the predictive relationships of numerous financial and economic variables. Neural network models for level estimation and classification are then examined for their ability to provide an effective forecast of future values. A cross-validation technique is also employed to improve the generalization ability of several models. The results show that the trading strategies guided by the classification models generate higher risk-adjusted profits than the buy-and-hold strategy, as well as those guided by the level-estimation based forecasts of the neural network and linear regression models.
Article
Pairs trading is a popular speculation strategy. Several implementation methods are proposed in the literature: they can be based on a distance criterion or on co-integration. This article extends previous research in another direction: the combination of forecasting techniques (Neural Networks) and multi-criteria decision making methods (Electre III). The key contribution of this paper is the introduction of multi-step-ahead forecasts. It leads to major changes in the trading system and raises new empirical and methodological questions. The results of an application based on S&P 100 Index stocks are promising: this methodology could be a powerful tool for pairs selection in a highly non-linear environment.
Article
The key to successful stock market forecasting is achieving best results with minimum required input data. Given stock market model uncertainty, soft computing techniques are viable candidates to capture stock market nonlinear relations returning significant forecasting results with not necessarily prior knowledge of input data statistical distributions. This paper surveys more than 100 related published articles that focus on neural and neuro-fuzzy techniques derived and applied to forecast stock markets. Classifications are made in terms of input data, forecasting methodology, performance evaluation and performance measures used. Through the surveyed papers, it is shown that soft computing techniques are widely accepted to studying and evaluating stock market behavior.
Article
Pairs trading is a popular quantitative speculation strategy. This article proposes a general and flexible framework for pairs selection. The method uses multiple return forecasts based on bivariate information sets and multi-criteria decision techniques. Our approach can be seen as a sort of forecast combination but the output of the method is a ranking. It helps to detect potentially under- and overvalued stocks. A first application with S&P 100 index stocks provides promising results in terms of excess return and directional forecasting.
Article
In Support Vector (SV) regression, a parameter nu controls the number of Support Vectors and the number of points that come to lie outside of the so-called epsilon-insensitive tube. For various noise models and SV parameter settings, we experimentally determine the values of nu that lead to the lowest generalization error. We find good agreement with the values that had previously been predicted by a theoretical argument based on the asymptotic efficiency of a simplified model of SV regression. As a side effect of the experiments, valuable information about the generalization behavior of the remaining SVM parameters and their dependencies is gained. The experimental findings are valid even for complex 'real-world' data sets. Based on our results on the role of the nu-SVM parameters, we discuss various model selection methods.
Article
Despite abundant research which focuses on estimating the level of return on stock market index, there is a lack of studies examining the predictability of the direction/sign of stock index movement. Given the notion that a prediction with little forecast error does not necessarily translate into capital gain, we evaluate the efficacy of several multivariate classification techniques relative to a group of level estimation approaches. Specifically, we conduct time series comparisons between the two types of models on the basis of forecast performance and investment return. The tested classification models, which predict direction based on probability, include linear discriminant analysis, logit, probit, and probabilistic neural network. On the other hand, the level estimation counterparts, which forecast the level, are exponential smoothing, multivariate transfer function, vector autoregression with Kalman filter, and multilayered feedforward neural network. Our comparative study also measures the relative strength of these models with respect to the trading profit generated by their forecasts. To facilitate more effective trading, we develop a set of threshold trading rules driven by the probabilities estimated by the classification models. Empirical experimentation suggests that the classification models outperform the level estimation models in terms of predicting the direction of the stock market movement and maximizing returns from investment trading. Further, investment returns are enhanced by the adoption of the threshold trading rules.
Lightgbm: A highly efficient gradient boosting decision tree
  • G Ke
  • Q Meng
  • T Finley
  • T Wang
  • W Chen
  • W Ma
  • Q Ye
  • T Y Liu
  • I Guyon
  • U V Luxburg
  • S Bengio
  • H Wallach
  • R Fergus
Ke, G., Meng, Q., Finley, T., Wang, T., Chen, W., Ma, W., Ye, Q., Liu, T.Y.: Lightgbm: A highly efficient gradient boosting decision tree. In: Guyon, I., Luxburg, U.V., Bengio, S., Wallach, H., Fergus, R., Vishwanathan, S., Garnett, R. (eds.) Advances in Neural Information Processing Systems 30. pp. 3146-3154. Curran Associates, Inc. (2017)
The Evolution of Technical Analysis: Financial Prediction from Babylonian Tablets to Bloomberg Terminals
  • A Lo
  • J Hasanhodzic
Lo, A., Hasanhodzic, J.: The Evolution of Technical Analysis: Financial Prediction from Babylonian Tablets to Bloomberg Terminals. Bloomberg, Wiley (2011)
Statistical arbitrage on the kospi 200: An exploratory analysis of classification and prediction machine learning algorithms for day trading
  • I Sutherland
  • Y Jung
  • G Lee
Sutherland, I., Jung, Y., Lee, G.: Statistical arbitrage on the kospi 200: An exploratory analysis of classification and prediction machine learning algorithms for day trading. Journal of Economics and International Business Management 6(1), 10-19 (2018)
Applying Deep Learning to Enhance Momentum Trading Strategies in Stocks
  • L Takeuchi
Takeuchi, L.: Applying Deep Learning to Enhance Momentum Trading Strategies in Stocks (2013)