Decision Sciences › Management Science and Operations Research

Stock Market Forecasting Methods

Works Count: 47603

Wikipedia

Description

This cluster of papers focuses on predicting stock market trends and movements using various techniques such as time series forecasting, neural networks, deep learning, support vector machines, sentiment analysis, and Twitter data. The research explores the application of these methods to financial time series data for stock market prediction.

Keywords

Stock Market Prediction; Time Series Forecasting; Neural Networks; Financial Time Series; Deep Learning; Support Vector Machines; LSTM Networks; Forecasting Models; Sentiment Analysis; Twitter Data

Sentiment analysis on social media for stock movement prediction

2015-08-07

Thien Hai Nguyen , Kiyoaki Shirai , Julien Velcin

Forecasting with artificial neural networks:

1998-03-01

Guoqiang Zhang , B. Eddy Patuwo , Michael Y. Hu

Technical analysis of the financial markets: a comprehensive guide to trading methods and applications

1999-03-01

John J. Murphy

This outstanding reference has already taught thousands of traders the concepts of technical analysis and their application in the futures and stock markets. Covering the latest developments in computer technology, … This outstanding reference has already taught thousands of traders the concepts of technical analysis and their application in the futures and stock markets. Covering the latest developments in computer technology, technical tools, and indicators, the second edition features new material on candlestick charting, intermarket relationships, stocks and stock rotation, plus state-of-the-art examples and figures. From how to read charts to understanding indicators and the crucial role technical analysis plays in investing, readers gain a thorough and accessible overview of the field of technical analysis, with a special emphasis on futures markets. Revised and expanded for the demands of today's financial world, this book is essential reading for anyone interested in tracking and analyzing market behavior.

Stock Price Prediction Using the ARIMA Model

2014-03-01

Ayodele A. Adebiyi , Aderemi O. Adewumi , C. K. Ayo

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 … 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.

View PDF Chat

Predicting direction of stock price index movement using artificial neural networks and support vector machines: The sample of the Istanbul Stock Exchange

2010-11-01

Yakup Kara , Melek Acar Boyacıoğlu , Ömer Kaan Baykan

A hybrid ARIMA and support vector machines model in stock price forecasting

2004-09-18

Ping‐Feng Pai , Chih-Sheng Lin

Application of support vector machines in financial time series forecasting

2001-08-01

Francis E. H. Tay , Lijuan Cao

Effective lengths of intervals to improve forecasting in fuzzy time series

2001-11-01

Kun‐Huang Huarng

Irrational Behavior and Economic Theory

1962-02-01

Gary S. Becker

Surveying stock market forecasting techniques – Part II: Soft computing methods

2008-07-19

George S. Atsalakis , Kimon P. Valavanis

Neural network forecasting for seasonal and trend time series

2003-11-20

G.Peter Zhang , Min Qi

Financial time series forecasting using support vector machines

2003-05-19

Kyoung-jae Kim

Textual analysis of stock market prediction using breaking financial news

2009-02-01

Robert P. Schumaker , Hsinchun Chen

Our research examines a predictive machine learning approach for financial news articles analysis using several different textual representations: bag of words, noun phrases, and named entities. Through this approach, we … Our research examines a predictive machine learning approach for financial news articles analysis using several different textual representations: bag of words, noun phrases, and named entities. Through this approach, we investigated 9,211 financial news articles and 10,259,042 stock quotes covering the S&P 500 stocks during a five week period. We applied our analysis to estimate a discrete stock price twenty minutes after a news article was released. Using a support vector machine (SVM) derivative specially tailored for discrete numeric prediction and models containing different stock-specific variables, we show that the model containing both article terms and stock price at the time of article release had the best performance in closeness to the actual future stock price (MSE 0.04261), the same direction of price movement as the future price (57.1% directional accuracy) and the highest return using a simulated trading engine (2.06% return). We further investigated the different textual representations and found that a Proper Noun scheme performs better than the de facto standard of Bag of Words in all three metrics.

A non-random walk down Wall Street

1999-12-01

Andrew W. Lo , A. Craig MacKinlay

For over half a century, financial experts have regarded the movements of markets as a random walk--unpredictable meanderings akin to a drunkard's unsteady gait--and this hypothesis has become a cornerstone … For over half a century, financial experts have regarded the movements of markets as a random walk--unpredictable meanderings akin to a drunkard's unsteady gait--and this hypothesis has become a cornerstone of modern financial economics and many investment strategies. Here Andrew W. Lo and A. Craig MacKinlay put the Random Walk Hypothesis to the test. In this volume, which elegantly integrates their most important articles, Lo and MacKinlay find that markets are not completely random after all, and that predictable components do exist in recent stock and bond returns. Their book provides a state-of-the-art account of the techniques for detecting predictabilities and evaluating their statistical and economic significance, and offers a tantalizing glimpse into the financial technologies of the future. The articles track the exciting course of Lo and MacKinlay's research on the predictability of stock prices from their early work on rejecting random walks in short-horizon returns to their analysis of long-term memory in stock market prices. A particular highlight is their now-famous inquiry into the pitfalls of data-snooping biases that have arisen from the widespread use of the same historical databases for discovering anomalies and developing seemingly profitable investment strategies. This book invites scholars to reconsider the Random Walk Hypothesis, and, by carefully documenting the presence of predictable components in the stock market, also directs investment professionals toward superior long-term investment returns through disciplined active investment management.

Predicting stock and stock price index movement using Trend Deterministic Data Preparation and machine learning techniques

2014-08-06

Jigar Patel , Sahil Shah , Priyank Thakkar +1 more

Forecasting stock market movement direction with support vector machine

2004-05-30

Wei Huang , Yoshiteru Nakamori , Shouyang Wang

Yahoo! for Amazon: Sentiment Extraction from Small Talk on the Web

2007-09-01

Sanjiv Ranjan Das , Mike Y. Chen

Extracting sentiment from text is a hard semantic problem. We develop a methodology for extracting small investor sentiment from stock message boards. The algorithm comprises different classifier algorithms coupled together … Extracting sentiment from text is a hard semantic problem. We develop a methodology for extracting small investor sentiment from stock message boards. The algorithm comprises different classifier algorithms coupled together by a voting scheme. Accuracy levels are similar to widely used Bayes classifiers, but false positives are lower and sentiment accuracy higher. Time series and cross-sectional aggregation of message information improves the quality of the resultant sentiment index, particularly in the presence of slang and ambiguity. Empirical applications evidence a relationship with stock values—tech-sector postings are related to stock index levels, and to volumes and volatility. The algorithms may be used to assess the impact on investor opinion of management announcements, press releases, third-party news, and regulatory changes.

A novel hybridization of artificial neural networks and ARIMA models for time series forecasting

2010-11-05

Mehdi Khashei , Mehdi Bijari

Using artificial neural network models in stock market index prediction

2011-02-10

Erkam Güreşen , Gülgün Kayakutlu , Tuğrul Daim

Genetic algorithms approach to feature discretization in artificial neural networks for the prediction of stock price index

2000-08-01

Kyoung-jae Kim , Ingoo Han

Is Technical Analysis in the Foreign Exchange Market Profitable? A Genetic Programming Approach

1997-12-01

Christopher J. Neely , Paul Weller , Rob Dittmar

Using genetic programming techniques to find technical trading rules, we find strong evidence of economically significant out-of-sample excess returns to those rules for each of six exchange rates over the … Using genetic programming techniques to find technical trading rules, we find strong evidence of economically significant out-of-sample excess returns to those rules for each of six exchange rates over the period 1981‐1995. Further, when the dollar/Deutsche mark rules are allowed to determine trades in the other markets, there is significant improvement in performance in all cases, except for the Deutsche mark/yen. Betas calculated for the returns according to various benchmark portfolios provide no evidence that the returns to these rules are compensation for bearing systematic risk. Bootstrapping results on the dollar/Deutsche mark indicate that the trading rules detect patterns in the data that are not captured by standard statistical models.

View PDF Chat

Applied Time Series Econometrics

2004-08-02

Helmut Lütkepohl , Markus Krätzig

Time series econometrics is a rapidly evolving field. Particularly, the cointegration revolution has had a substantial impact on applied analysis. Hence, no textbook has managed to cover the full range … Time series econometrics is a rapidly evolving field. Particularly, the cointegration revolution has had a substantial impact on applied analysis. Hence, no textbook has managed to cover the full range of methods in current use and explain how to proceed in applied domains. This gap in the literature motivates the present volume. The methods are sketched out, reminding the reader of the ideas underlying them and giving sufficient background for empirical work. The treatment can also be used as a textbook for a course on applied time series econometrics. Topics include: unit root and cointegration analysis, structural vector autoregressions, conditional heteroskedasticity and nonlinear and nonparametric time series models. Crucial to empirical work is the software that is available for analysis. New methodology is typically only gradually incorporated into existing software packages. Therefore a flexible Java interface has been created, allowing readers to replicate the applications and conduct their own analyses.

Time series forecasting using a hybrid ARIMA and neural network model

2003-01-01

G.Peter Zhang

An artificial neural network (p,d,q) model for timeseries forecasting

2009-05-23

Mehdi Khashei , Mehdi Bijari

Nonlinear Time Series: Nonparametric and Parametric Methods

2003-01-01

Jianqing Fan , Qiwei Yao

Recurrent Neural Networks

1999-12-20

Lakhmi C. Jain , L. R. Medsker

From the Publisher: With applications ranging from motion detection to financial forecasting, recurrent neural networks (RNNs) have emerged as an interesting and important part of neural network research. Recurrent Neural … From the Publisher: With applications ranging from motion detection to financial forecasting, recurrent neural networks (RNNs) have emerged as an interesting and important part of neural network research. Recurrent Neural Networks: Design and Applications reflects the tremendous, worldwide interest in and virtually unlimited potential of RNNs - providing a summary of the design, applications, current research, and challenges of this dynamic and promising field.

Using genetic algorithms to find technical trading rules

1999-02-01

Franklin Allen , Risto Karjalainen

We use a genetic algorithm to learn technical trading rules for the S&P 500 index using daily prices from 1928 to 1995. After transaction costs, the rules do not earn … We use a genetic algorithm to learn technical trading rules for the S&P 500 index using daily prices from 1928 to 1995. After transaction costs, the rules do not earn consistent excess returns over a simple buy-and-hold strategy in the out-of-sample test periods. The rules are able to identify periods to be in the index when daily returns are positive and volatility is low and out when the reverse is true. These latter results can largely be explained by low-order serial correlation in stock index returns.

A LSTM-based method for stock returns prediction: A case study of China stock market

2015-10-01

Kai Chen , Yi Zhou , Fangyan Dai

The presented paper modeled and predicted China stock returns using LSTM. The historical data of China stock market were transformed into 30-days-long sequences with 10 learning features and 3-day earning … The presented paper modeled and predicted China stock returns using LSTM. The historical data of China stock market were transformed into 30-days-long sequences with 10 learning features and 3-day earning rate labeling. The model was fitted by training on 900000 sequences and tested using the other 311361 sequences. Compared with random prediction method, our LSTM model improved the accuracy of stock returns prediction from 14.3% to 27.2%. The efforts demonstrated the power of LSTM in stock market prediction in China, which is mechanical yet much more unpredictable.

Deep learning for event-driven stock prediction

2015-07-25

Xiao Ding , Yue Zhang , Ting Liu +1 more

We propose a deep learning method for event-driven stock market prediction. First, events are extracted from news text, and represented as dense vectors, trained using a novel neural tensor network. … We propose a deep learning method for event-driven stock market prediction. First, events are extracted from news text, and represented as dense vectors, trained using a novel neural tensor network. Second, a deep convolutional neural network is used to model both short-term and long-term influences of events on stock price movements. Experimental results show that our model can achieve nearly 6% improvements on S&P 500 index prediction and individual stock prediction, respectively, compared to state-of-the-art baseline methods. In addition, market simulation results show that our system is more capable of making profits than previously reported systems trained on S&P 500 stock historical data.

Deep Direct Reinforcement Learning for Financial Signal Representation and Trading

2016-02-15

Yue Deng , Feng Bao , Youyong Kong +2 more

Can we train the computer to beat experienced traders for financial assert trading? In this paper, we try to address this challenge by introducing a recurrent deep neural network (NN) … Can we train the computer to beat experienced traders for financial assert trading? In this paper, we try to address this challenge by introducing a recurrent deep neural network (NN) for real-time financial signal representation and trading. Our model is inspired by two biological-related learning concepts of deep learning (DL) and reinforcement learning (RL). In the framework, the DL part automatically senses the dynamic market condition for informative feature learning. Then, the RL module interacts with deep representations and makes trading decisions to accumulate the ultimate rewards in an unknown environment. The learning system is implemented in a complex NN that exhibits both the deep and recurrent structures. Hence, we propose a task-aware backpropagation through time method to cope with the gradient vanishing issue in deep training. The robustness of the neural system is verified on both the stock and the commodity future markets under broad testing conditions.

Analysis of Financial Time Series

2010-08-02

Ruey S. Tsay

Analysis of Financial Time Series

2005-08-19

Ruey S. Tsay

Predicting the Price of Bitcoin Using Machine Learning

2018-03-01

Sean Mcnally , Jason T. Roche , Simon Caton

The goal of this paper is to ascertain with what accuracy the direction of Bitcoin price in USD can be predicted. The price data is sourced from the Bitcoin Price … The goal of this paper is to ascertain with what accuracy the direction of Bitcoin price in USD can be predicted. The price data is sourced from the Bitcoin Price Index. The task is achieved with varying degrees of success through the implementation of a Bayesian optimised recurrent neural network (RNN) and a Long Short Term Memory (LSTM) network. The LSTM achieves the highest classification accuracy of 52% and a RMSE of 8%. The popular ARIMA model for time series forecasting is implemented as a comparison to the deep learning models. As expected, the non-linear deep learning methods outperform the ARIMA forecast which performs poorly. Finally, both deep learning models are benchmarked on both a GPU and a CPU with the training time on the GPU outperforming the CPU implementation by 67.7%.

View PDF Chat

Deep learning for finance: deep portfolios

2016-10-07

J.B. Heaton , Nick Polson , J. H. Witte

We explore the use of deep learning hierarchical models for problems in financial prediction and classification. Financial prediction problems – such as those presented in designing and pricing securities, constructing … We explore the use of deep learning hierarchical models for problems in financial prediction and classification. Financial prediction problems – such as those presented in designing and pricing securities, constructing portfolios, and risk management – often involve large data sets with complex data interactions that currently are difficult or impossible to specify in a full economic model. Applying deep learning methods to these problems can produce more useful results than standard methods in finance. In particular, deep learning can detect and exploit interactions in the data that are, at least currently, invisible to any existing financial economic theory. Copyright © 2016 John Wiley & Sons, Ltd.

Deep learning networks for stock market analysis and prediction: Methodology, data representations, and case studies

2017-04-22

Eunsuk Chong , Chulwoo Han , Frank C. Park

View PDF Chat

A Dual-Stage Attention-Based Recurrent Neural Network for Time Series Prediction

2017-07-28

Yao Qin , Dongjin Song , Haifeng Chen +3 more

The Nonlinear autoregressive exogenous (NARX) model, which predicts the current value of a time series based upon its previous values as well as the current and past values of multiple … The Nonlinear autoregressive exogenous (NARX) model, which predicts the current value of a time series based upon its previous values as well as the current and past values of multiple driving (exogenous) series, has been studied for decades. Despite the fact that various NARX models have been developed, few of them can capture the long-term temporal dependencies appropriately and select the relevant driving series to make predictions. In this paper, we propose a dual-stage attention-based recurrent neural network (DA-RNN) to address these two issues. In the first stage, we introduce an input attention mechanism to adaptively extract relevant driving series (a.k.a., input features) at each time step by referring to the previous encoder hidden state. In the second stage, we use a temporal attention mechanism to select relevant encoder hidden states across all time steps. With this dual-stage attention scheme, our model can not only make predictions effectively, but can also be easily interpreted. Thorough empirical studies based upon the SML 2010 dataset and the NASDAQ 100 Stock dataset demonstrate that the DA-RNN can outperform state-of-the-art methods for time series prediction.

View PDF Chat

Deep learning with long short-term memory networks for financial market predictions

2017-12-06

Thomas Fischer , Christopher Krauß

View PDF Chat

A deep learning framework for financial time series using stacked autoencoders and long-short term memory

2017-07-14

Wei Bao , Jun Yue , Yulei Rao

The application of deep learning approaches to finance has received a great deal of attention from both investors and researchers. This study presents a novel deep learning framework where wavelet … The application of deep learning approaches to finance has received a great deal of attention from both investors and researchers. This study presents a novel deep learning framework where wavelet transforms (WT), stacked autoencoders (SAEs) and long-short term memory (LSTM) are combined for stock price forecasting. The SAEs for hierarchically extracted deep features is introduced into stock price forecasting for the first time. The deep learning framework comprises three stages. First, the stock price time series is decomposed by WT to eliminate noise. Second, SAEs is applied to generate deep high-level features for predicting the stock price. Third, high-level denoising features are fed into LSTM to forecast the next day's closing price. Six market indices and their corresponding index futures are chosen to examine the performance of the proposed model. Results show that the proposed model outperforms other similar models in both predictive accuracy and profitability performance.

View PDF Chat

Stock market's price movement prediction with LSTM neural networks

2017-05-01

David M. Q. Nelson , Adriano C. M. Pereira , Renato Arantes de Oliveira

Predictions on stock market prices are a great challenge due to the fact that it is an immensely complex, chaotic and dynamic environment. There are many studies from various areas … Predictions on stock market prices are a great challenge due to the fact that it is an immensely complex, chaotic and dynamic environment. There are many studies from various areas aiming to take on that challenge and Machine Learning approaches have been the focus of many of them. There are many examples of Machine Learning algorithms been able to reach satisfactory results when doing that type of prediction. This article studies the usage of LSTM networks on that scenario, to predict future trends of stock prices based on the price history, alongside with technical analysis indicators. For that goal, a prediction model was built, and a series of experiments were executed and theirs results analyzed against a number of metrics to assess if this type of algorithm presents and improvements when compared to other Machine Learning methods and investment strategies. The results that were obtained are promising, getting up to an average of 55.9% of accuracy when predicting if the price of a particular stock is going to go up or not in the near future.

Financial time series forecasting model based on CEEMDAN and LSTM

2018-12-13

Jian Cao , Zhi Li , Jian Li

A Comparison of ARIMA and LSTM in Forecasting Time Series

2018-12-01

Sima Siami‐Namini , Neda Tavakoli , Akbar Siami Namin

Forecasting time series data is an important subject in economics, business, and finance. Traditionally, there are several techniques to effectively forecast the next lag of time series data such as … Forecasting time series data is an important subject in economics, business, and finance. Traditionally, there are several techniques to effectively forecast the next lag of time series data such as univariate Autoregressive (AR), univariate Moving Average (MA), Simple Exponential Smoothing (SES), and more notably Autoregressive Integrated Moving Average (ARIMA) with its many variations. In particular, ARIMA model has demonstrated its outperformance in precision and accuracy of predicting the next lags of time series. With the recent advancement in computational power of computers and more importantly development of more advanced machine learning algorithms and approaches such as deep learning, new algorithms are developed to analyze and forecast time series data. The research question investigated in this article is that whether and how the newly developed deep learning-based algorithms for forecasting time series data, such as "Long Short-Term Memory (LSTM)", are superior to the traditional algorithms. The empirical studies conducted and reported in this article show that deep learning-based algorithms such as LSTM outperform traditional-based algorithms such as ARIMA model. More specifically, the average reduction in error rates obtained by LSTM was between 84 - 87 percent when compared to ARIMA indicating the superiority of LSTM to ARIMA. Furthermore, it was noticed that the number of training times, known as "epoch" in deep learning, had no effect on the performance of the trained forecast model and it exhibited a truly random behavior.

Prediction machines: the simple economics of artificial intelligence

2019-10-01

Donald Heath

Our first key insight is that the new wave of artificial intelligence does not actually bring us intelligence but instead a critical component of intelligence – prediction. – Ajay Agrawal, … Our first key insight is that the new wave of artificial intelligence does not actually bring us intelligence but instead a critical component of intelligence – prediction. – Ajay Agrawal, Joshua G...

Financial time series forecasting with deep learning : A systematic literature review: 2005–2019

2020-02-19

Ömer Berat Sezer , Mehmet Ugur Gudelek , Ahmet Murat Özbayoğlu

View PDF Chat

The Performance of LSTM and BiLSTM in Forecasting Time Series

2019-12-01

Sima Siami‐Namini , Neda Tavakoli , Akbar Siami Namin

Machine and deep learning-based algorithms are the emerging approaches in addressing prediction problems in time series. These techniques have been shown to produce more accurate results than conventional regression-based modeling. … Machine and deep learning-based algorithms are the emerging approaches in addressing prediction problems in time series. These techniques have been shown to produce more accurate results than conventional regression-based modeling. It has been reported that artificial Recurrent Neural Networks (RNN) with memory, such as Long Short-Term Memory (LSTM), are superior compared to Autoregressive Integrated Moving Average (ARIMA) with a large margin. The LSTM-based models incorporate additional "gates" for the purpose of memorizing longer sequences of input data. The major question is that whether the gates incorporated in the LSTM architecture already offers a good prediction and whether additional training of data would be necessary to further improve the prediction. Bidirectional LSTMs (BiLSTMs) enable additional training by traversing the input data twice (i.e., 1) left-to-right, and 2) right-to-left). The research question of interest is then whether BiLSTM, with additional training capability, outperforms regular unidirectional LSTM. This paper reports a behavioral analysis and comparison of BiLSTM and LSTM models. The objective is to explore to what extend additional layers of training of data would be beneficial to tune the involved parameters. The results show that additional training of data and thus BiLSTM-based modeling offers better predictions than regular LSTM-based models. More specifically, it was observed that BiLSTM models provide better predictions compared to ARIMA and LSTM models. It was also observed that BiLSTM models reach the equilibrium much slower than LSTM-based models.

Modelling Financial Time Series.

1987-06-01

Gordon Anderson , Stephen J. Taylor

Journal Article Modelling Financial Time Series Get access Modelling Financial Time Series. By STEPHEN TAYLOR. (Chichester & New York: John Wiley, 1986. Pp. xvi + 268. ÃÂ£19.95 paperback.) Gordon Anderson … Journal Article Modelling Financial Time Series Get access Modelling Financial Time Series. By STEPHEN TAYLOR. (Chichester & New York: John Wiley, 1986. Pp. xvi + 268. ÃÂ£19.95 paperback.) Gordon Anderson Gordon Anderson McMaster University Search for other works by this author on: Oxford Academic Google Scholar The Economic Journal, Volume 97, Issue 386, 1 June 1987, Pages 512–513, https://doi.org/10.2307/2232901 Published: 01 June 1987

Foundations of Technical Analysis: Computational Algorithms, Statistical Inference, and Empirical Implementation

2000-08-01

Andrew W. Lo , Harry Mamaysky , Jiang Wang

Technical analysis, also known as “charting,” has been a part of financial practice for many decades, but this discipline has not received the same level of academic scrutiny and acceptance … Technical analysis, also known as “charting,” has been a part of financial practice for many decades, but this discipline has not received the same level of academic scrutiny and acceptance as more traditional approaches such as fundamental analysis. One of the main obstacles is the highly subjective nature of technical analysis—the presence of geometric shapes in historical price charts is often in the eyes of the beholder. In this paper, we propose a systematic and automatic approach to technical pattern recognition using nonparametric kernel regression, and we apply this method to a large number of U.S. stocks from 1962 to 1996 to evaluate the effectiveness of technical analysis. By comparing the unconditional empirical distribution of daily stock returns to the conditional distribution—conditioned on specific technical indicators such as head‐and‐shoulders or double bottoms—we find that over the 31‐year sample period, several technical indicators do provide incremental information and may have some practical value.

View PDF Chat

Dividend yields and expected stock returns: alternative procedures for inference and measurement

1992-07-01

Robert J. Hodrick

Alternative ways of conducting inference and measurement for long-horizon forecasting are explored with an application to dividend yields as predictors of stock returns. Monte Carlo analysis indicates that the Hansen … Alternative ways of conducting inference and measurement for long-horizon forecasting are explored with an application to dividend yields as predictors of stock returns. Monte Carlo analysis indicates that the Hansen and Hodrick (1980) procedure is biased at long horizons, but the alternatives perform better. These include an estimator derived under the null hypothesis as in Richardson and Smith (1991), a reformulation of the regression as in Jegadeesh (1990), and a vector autoregression (VAR) as in Campbell and Shiller (1988), Kandel and Stambaugh (1988), and Campbell (1991). The statistical properties of long-horizon statistics generated from the VAR indicate interesting patterns in expected stock returns.

Analysis of Financial Time Series

2002-01-11

Ruey S. Tsay

Generative AI for financial document summarization and risk analysis

2025-06-24

Narangarav Batbaatar | World Journal of Advanced Research and Reviews

This paper discusses how generative AI can be applied to the field of financial document summarization and risk analysis to handle the issues of high volumes of complicated financial information. … This paper discusses how generative AI can be applied to the field of financial document summarization and risk analysis to handle the issues of high volumes of complicated financial information. The main goal consists of determining how effective AI-based models can be when it comes to summarizing financial documents and improving the process of risk assessment. The study provides a mixed-methods investigation of the case studies of AI-powered systems implemented in financial institutions, as well as a performance analysis according to the major metrics, including accuracy, efficiency, and risk prediction. Among the main insights, it is possible to mention that generative AI considerably enhances the quality and speed of summarization of financial documents, allowing institutions to analyze huge volumes of data in real-time and improving risk analysis. Additionally, machine learning models have a competitive advantage in eliminating people error and biasness in risk assessment. This study is important as it explains how generative AI has the potential to transform financial document processing and risk management to provide viable solutions to financial institutions to enhance the decision-making process, minimize operational expenses, and broaden the scopes of overall risk management. The paper has ended with suggestions on the future AI use in finance.

Long-term forecasting of stock prices using time series models: Evidence from solar industries India Ltd

2025-06-24

Shariq Mohammed , Imran Azad , Ataur Rahman Farooqi +2 more | International Journal of Innovative Research and Scientific Studies

Long-term stock price analysis is essential for understanding market dynamics and supporting investment decisions. Using data from the National Stock Exchange, the study focuses on Solar Industries India Limited's closing … Long-term stock price analysis is essential for understanding market dynamics and supporting investment decisions. Using data from the National Stock Exchange, the study focuses on Solar Industries India Limited's closing prices between January 1, 2000, and December 31, 2024. The primary objectives are to use time series analysis to model and predict closing prices, identify underlying trends, and investigate relationships with trading activity. An ARIMA model was used with Minitab software to account for data noise, trends, and seasonal fluctuations. Statistical criteria guided the model selection process, providing optimal fit and reliability. These charts helped determine trends, patterns, and seasonal elements. A total of 729 monthly observations were examined, and the best-fitting model was chosen using the Akaike Information Criterion (AICc). The Ljung-Box test verified that the ARIMA (0, 2, 1) model was the best model, as it had the lowest AICc and strong residual diagnostics (p > 0.05 for most lags). The forecasts showed anticipated pricing ranges with increasing uncertainty over time. Further statistical analysis was conducted to investigate the relationships between trading activity and stock price. While insights into the relationship between trade volume and price movement provide useful perspectives for market analysis, a strong forecasting model can help stakeholders make well-informed decisions. Overall, the integration of exploratory research with time series modeling provides a comprehensive framework for analyzing stock price behavior and forecasting future trends.

Investment Portfolios Optimization with Genetic Algorithm: An Approach Applied to the Spanish Market (IBEX 35)

2025-06-24

Sandra Millán-Palacios , Javier Sánchez-Soriano | Electronics

The results of this study validate the use of single-objective genetic algorithms as an effective tool for portfolio optimization in the Spanish market. Through an evolutionary approach with advanced objective … The results of this study validate the use of single-objective genetic algorithms as an effective tool for portfolio optimization in the Spanish market. Through an evolutionary approach with advanced objective functions and a phased structure (training, validation, and testing), the quality and stability of the generated portfolios were significantly improved. The single-objective genetic algorithms with a Complex Objective Function (SGA-COF-1) model delivered outstanding returns with high robustness and were adaptable to different risk profiles, including the ESG criteria. These contributions open multiple future research directions, such as the incorporation of predictive models, expansion to international markets, and the use of more sophisticated evolutionary algorithms. The proposed methodological framework (flexible and scalable) provides a solid foundation for the development of automated and sustainable quantitative investment solutions.

Stock Market Prediction Using Machine Learning and Deep Learning Techniques: A Review

2025-06-24

Mohammadreza Saberironaghi , Jing Ren , Alireza Saberironaghi | AppliedMath

The rapid advancement of machine learning and deep learning techniques has revolutionized stock market prediction, providing innovative methods to analyze financial trends and market behavior. This review paper presents a … The rapid advancement of machine learning and deep learning techniques has revolutionized stock market prediction, providing innovative methods to analyze financial trends and market behavior. This review paper presents a comprehensive analysis of various machine learning and deep learning approaches utilized in stock market prediction, focusing on their methodologies, evaluation metrics, and datasets. Popular models such as LSTM, CNN, and SVM are examined, highlighting their strengths and limitations in predicting stock prices, volatility, and trends. Additionally, we address persistent challenges, including data quality and model interpretability, and explore emerging research directions to overcome these obstacles. This study aims to summarize the current state of research, provide insights into the effectiveness of predictive models.

A Comparative Study of Deep Learning Approaches for Price Forecasting: A Case Study on Pomegranate

2025-06-23

S. Vishnu Shankar , V. Lavanya , P. Pandiyaraj +2 more | Deleted Journal

Public sentiment-based stock selection strategy using the BERT model and LightGBM algorithm

2025-06-23

Leqi Wang , Junwei Shen , Chenguang Shen +1 more | Journal of Computational Methods in Sciences and Engineering

This paper develops a novel approach to stock selection by integrating natural language processing techniques with machine learning algorithms to analyze public opinion data in financial markets. Specifically, we employ … This paper develops a novel approach to stock selection by integrating natural language processing techniques with machine learning algorithms to analyze public opinion data in financial markets. Specifically, we employ the Bidirectional Encoder Representations from Transformers (BERT) model to process and classify financial news and social media content, combined with the Light Gradient Boosting Machine (LightGBM) algorithm to select high-potential stocks within identified concept sectors. Using over 18 million Chinese financial text records from 2023 to 2024, we construct a comprehensive framework that captures both market sentiment and stock-specific characteristics. Our strategy consists of three core components: (1) a BERT-based sentiment analyzer that identifies promising concept sectors with strong momentum, (2) a LightGBM-powered stock selection mechanism utilizing a specially designed “Concept-Momentum-Combined” factor alongside conventional financial indicators, and (3) a risk management system combining sentiment anomaly detection with multi-stage Average True Range (ATR) trailing stop mechanisms. Empirical results demonstrate significant outperformance over CSI 800 across various timeframes, with annualized excess returns of 21.55% over a six-month period and maximum drawdowns of only 11.68%. Performance attribution analysis confirms that concept sector selection based on sentiment analysis is the primary driver of excess returns. Our results add to the expanding literature on the use of artificial intelligence in financial markets and provide actionable takeaways for investors who would like to incorporate public opinion data into their investment process.

A Full-Fledged Stock Market Prediction Framework using Adaptive TCN with a Bayesian Learning Network via Enhanced Good and Bad Groups-Based Optimizer

2025-06-23

Rakesh Roshan , Krishna Kumar N , Supraja Ballari +6 more | Computational Economics

Modeling and Pricing Exotic Options in Frontier Markets: A Computational Approach with Applications to Kenya’s Financial Sector

2025-06-23

Jackson Barngetuny | International Journal of Finance and Accounting

Purpose: This research examines how exotic options—such as Asian, lookback, and barrier options—are priced in Kenya’s emerging financial market. It looks into whether sophisticated computational models can work effectively in … Purpose: This research examines how exotic options—such as Asian, lookback, and barrier options—are priced in Kenya’s emerging financial market. It looks into whether sophisticated computational models can work effectively in a market challenged by limited data, low liquidity, and underdeveloped infrastructure. The study addresses a gap in existing literature, which mostly focuses on well-established markets, by assessing the feasibility of introducing complex financial instruments in frontier economies to encourage innovation, better risk management, and economic growth. Methodology: The study uses a quantitative approach, combining stochastic models with real market data from the Nairobi Securities Exchange (NSE) collected between 2019 and 2023. It also incorporates recent guidelines from the Capital Markets Authority (CMA). Pricing methods such as Monte Carlo simulations, Finite Difference Methods (FDM), and binomial/trinomial tree models are tailored to fit the local market context. To overcome the challenges of limited data, techniques like kernel smoothing for volatility estimation and bootstrapping to create synthetic data sets are applied. A mix of these methods helps improve pricing accuracy, especially under conditions of incomplete information and clustered volatility. Findings: The Monte Carlo method proves highly effective for pricing options that depend on the path of the underlying asset, while FDM, especially the Crank-Nicolson approach, handles early exercise options and price jumps well. Binomial and trinomial trees remain reliable in data-scarce environments. Despite infrastructural and regulatory hurdles, the study shows that calibrating pricing models is possible using resampling and non-parametric methods. The results highlight the potential benefits of exotic derivatives in managing risks within key sectors such as agriculture, energy, and trade. Unique Contribution to Theory, Practice and Theory: This work enhances the theoretical framework by adjusting traditional option pricing models to fit the challenges of frontier markets. It provides a practical toolkit for financial firms and offers regulatory recommendations to nurture a sustainable derivatives market. By aligning advanced modeling techniques with local market realities, the study paves the way for broader adoption of derivative products in underdeveloped financial systems.

Bino-Trinomial Tree Model for Determining Modified Barrier Option Prices: An Effort to Minimize Losses

2025-06-22

Rima Aulia Rahayu , Fitriani Agustina , Kuntjoro Kuntjoro | CAUCHY Jurnal Matematika Murni dan Aplikasi

A particular exotic option that is widely traded in the global financial market is the barrier option. Barrier options are attractive because they have a limit that must be reached … A particular exotic option that is widely traded in the global financial market is the barrier option. Barrier options are attractive because they have a limit that must be reached to activate the option. These limits may be utilized by investors as a point of reference to minimize potential losses. Accordingly, the researcher attempts to use the bino-trinomial tree model as a new approach to minimize losses. The purpose of this study is to analyze the bino-trinomial tree model to provide investors with more flexible hedging experience. The bino-trinomial tree model is obtained by combining the trinomial tree model at the first stage, then the binomial tree model at a further stage. This analysis was conducted by calculating the type of knock-out european call options. The results demonstrate that this model can effectively, accurately and flexibly manage the complex options required by modern investors, including multi-step single moving barrier options and single window barrier options.

Fractional Optimizers for LSTM Networks in Financial Time Series Forecasting

2025-06-22

Mustapha Ez-zaiym , Yassine Senhaji , Meriem Rachid +2 more | Mathematics

This study investigates the theoretical foundations and practical advantages of fractional-order optimization in computational machine learning, with a particular focus on stock price forecasting using long short-term memory (LSTM) networks. … This study investigates the theoretical foundations and practical advantages of fractional-order optimization in computational machine learning, with a particular focus on stock price forecasting using long short-term memory (LSTM) networks. We extend several widely used optimization algorithms—including Adam, RMSprop, SGD, Adadelta, FTRL, Adamax, and Adagrad—by incorporating fractional derivatives into their update rules. This novel approach leverages the memory-retentive properties of fractional calculus to improve convergence behavior and model efficiency. Our experimental analysis evaluates the performance of fractional-order optimizers on LSTM networks tasked with forecasting stock prices for major companies such as AAPL, MSFT, GOOGL, AMZN, META, NVDA, JPM, V, and UNH. Considering four metrics (Sharpe ratio, directional accuracy, cumulative return, and MSE), the results show that fractional orders can significantly enhance prediction accuracy for moderately volatile stocks, especially among lower-cap assets. However, for highly volatile stocks, performance tends to degrade with higher fractional orders, leading to erratic and inconsistent forecasts. In addition, fractional optimizers with short-memory truncation offer a favorable trade-off between computational efficiency and modeling accuracy in medium-frequency financial applications. Their enhanced capacity to capture long-range dependencies and robust performance in noisy environments further justify their adoption in such contexts. These results suggest that fractional-order optimization holds significant promise for improving financial forecasting models—provided that the fractional parameters are carefully tuned to balance memory effects with system stability.

Enhanching Prophet Time Series Forecasting on Sparse Data via Hyperparameter Optimizattion: A Case Study in Retail

2025-06-22

Fadel Muhamad Hafid Atamimi , Wina Wintanti , Gunawan Abdillah | SinkrOn

In today’s competitive business landscape, accurate sales forecasting is crucial for retailers to optimize inventory, prevent overstock, and support strategic decision-making. However, many small to medium enterprises operate with sparse … In today’s competitive business landscape, accurate sales forecasting is crucial for retailers to optimize inventory, prevent overstock, and support strategic decision-making. However, many small to medium enterprises operate with sparse and irregular sales data, making conventional forecasting methods less effective. This study aims to evaluate the performance of the Prophet time series model in such non-ideal conditions and to investigate how hyperparameter tuning affects its forecasting accuracy. The research adopts the Prophet algorithm, an additive time series forecasting model developed by Facebook, which incorporates trend, seasonality, and holiday components. The model was implemented in two configurations: one using default parameters, and another with manually tuned hyperparameters, including changepoint prior scale (CP), seasonality prior scale (SP), and seasonality mode. A total of 32 experiments were conducted using historical transaction data from PT Eko Hejo. Results show that the default Prophet model achieved a MAPE of 9.50%, while the best-performing configuration (CP = 0.5, SP = 0.01, additive mode) reduced the MAPE to 6.80%. This indicates that hyperparameter tuning significantly improves forecast accuracy, even in sparse data environments. The study contributes both practically and scientifically by demonstrating that Prophet, when properly configured, is a robust and adaptable tool for business forecasting with limited data. It also highlights the value of manual tuning in enhancing model responsiveness and generalization, offering insights for further research in model comparison, automated optimization, and hybrid forecasting approaches.

Explainable reinforcement learning for trading decisions

2025-06-21

Narangarav Batbaatar | World Journal of Advanced Engineering Technology and Sciences

This article explores the application of Explainable Reinforcement Learning (XRL) in financial trading decisions, addressing the critical need for transparency and interpretability in AI-driven trading strategies. The study aims at … This article explores the application of Explainable Reinforcement Learning (XRL) in financial trading decisions, addressing the critical need for transparency and interpretability in AI-driven trading strategies. The study aims at understanding how to improve traditional reinforcement learning models which can be viewed as black-box systems such that they allow explainable insights without affecting performance. Through case studies, real-life applications, and comparative studies, the article investigates some of the XRL techniques, including the model-agnostic techniques and the hybrid techniques, to provide a better insight into the trading algorithms. The paper presents the following significant results, namely, explainable models are effective to enhance trust, mitigate risks, and allow human control over algorithmic trading. Moreover, the findings stress that explainable RL advances the transparency but creates complications concerning the model complexity and computational expenses. The article ends with the recommendations to continue investigating the hybrid XRL frameworks and outlines future research to make reinforcement learning models more ethical, accountable, and efficient in regards to the process of financial decision-making.

Multi-variate LSTM with attention mechanism for the Indian stock market

2025-06-21

Ashy Sebastian , Veerta Tantia | International Journal of Information Management Data Insights

Brazilian Stock Market Forecast with Heterogeneous Data Integration for a Set of Stocks

2025-06-20

Michele Jackeline Andressa Rosa , Marcos Roberto Souza , Camilo Machado +2 more | IEEE Latin America Transactions

Coefficient-Average Based Recursive Feature Elimination Modification to Optimize Feature Selection in Stock Price Prediction

2025-06-20

| International journal of intelligent engineering and systems

Decision Trees for Strategic Choice of Augmenting Management Intuition with Machine Learning

2025-06-20

Luo Guo-yu , Mohd Anuar Arshad , Gang Luo | Symmetry

Strategic financial decision-making is critical for organizational sustainability and competitive advantage. However, traditional approaches that rely solely on human expertise or isolated machine learning (ML) models often fall short in … Strategic financial decision-making is critical for organizational sustainability and competitive advantage. However, traditional approaches that rely solely on human expertise or isolated machine learning (ML) models often fall short in capturing the complex, multifaceted, and often asymmetrical nature of financial data, leading to suboptimal predictions and limited interpretability. This study addresses these challenges by developing an innovative, symmetry-aware integrated ML framework that synergizes decision trees, advanced ensemble techniques, and human expertise to enhance both predictive accuracy and model transparency. The proposed framework employs a symmetrical dual-feature selection process, combining automated methods based on decision trees with expert-guided selections, ensuring the inclusion of both statistically significant and domain-relevant features. Furthermore, the integration of human expertise facilitates rule-based adjustments and iterative feedback loops, refining model performance and aligning it with practical financial insights. Empirical evaluation shows a significant improvement in ROC-AUC by 2% and F1-score by 1.5% compared to baseline and advanced ML models alone. The inclusion of expert-driven rules, such as thresholds for debt-to-equity ratios and profitability margins, enables the model to account for real-world asymmetries that automated methods may overlook. Visualizations of the decision trees offer clear interpretability, providing decision-makers with symmetrical insight into how financial metrics influence bankruptcy predictions. This research demonstrates the effectiveness of combining machine learning with expert knowledge in bankruptcy prediction, offering a more robust, accurate, and interpretable decision-making tool. By incorporating both algorithmic precision and human reasoning, the study presents a balanced and symmetrical hybrid approach, bridging the gap between data-driven analytics and domain expertise. The findings underscore the potential of symmetry-driven integration of ML techniques and expert knowledge to enhance strategic financial decision-making.

Stock Price Prediction Using Deep Neural Networks: A Comparative Study Between MLP and LSTM

2025-06-20

Yuchen Duan | Advances in Economics Management and Political Sciences

It is difficult to predict stock prices due to volatile financial markets and various economic factors. However, useful effective predictive techniques can offer great assistance to analysts and investors, and … It is difficult to predict stock prices due to volatile financial markets and various economic factors. However, useful effective predictive techniques can offer great assistance to analysts and investors, and therefore much research keeps being conducted in this area. This study analyzed the predictive capabilities of two popular deep learning methods: Long Short-Term Memory (LSTM) and Multi-Layer Perceptron (MLP), based on a historical stock price dataset. This study conducted experiments through the daily stock prices of Google and a larger dataset that covered multiple international companies. It also evaluated the rigorousness of LSTM and MLP under many conditions, such as different fluctuation mechanisms, high and low price levels, and datasets of varying scales. The study found that the prediction accuracy of both LSTM and MLP was satisfactory. However, during stable and low-fluctuation periods, LSTM achieved better performance than MLP, but on smaller datasets, MLP showed stronger generalization capabilities. Therefore, to improve predictive capabilities, which model to use should be based on market context and data scale.

Momentum Investing: A Comparison of Machine Learning and Markowitz Theory

2025-06-20

J. Yan | Advances in Economics Management and Political Sciences

This paper compares two momentum-based trading strategies: a traditional mean-variance approach and a machine learning method using XGBoost. The rule-based strategy selects stocks based on short-term returns and applies mean-variance … This paper compares two momentum-based trading strategies: a traditional mean-variance approach and a machine learning method using XGBoost. The rule-based strategy selects stocks based on short-term returns and applies mean-variance optimization. It achieved lower cumulative returns (17.55%) but showed more stable performance, with a maximum drawdown of -68.35% and a return-to-drawdown ratio of 0.26. However, its reliance on assumptions like normally distributed returns and static covariances limits its real-world applicability. In contrast, the XGBoost strategy forecasts returns using historical price and volume data. It produced higher returns (40.43%) and a stronger return-to-drawdown ratio (0.31), but with increased risk, reflected in a drawdown of -84.45%. While better at capturing nonlinear signals, the model is more fragile under volatility. The study highlights the trade-off between stability and return potential in traditional and machine learning based momentum strategies. Future work should investigate hybrid approaches that combine statistical rigor with machine learning adaptability to improve robustness in dynamic markets.

State Space Driven Financial Signal Decoupling: Empirical Evidence of the Structural Advantages of Mamba Model in Global- MultiAsset Forecasting

2025-06-20

Shaoyuan Shen | Advances in Economics Management and Political Sciences

The stock market is a critical component of the global economy, offering investors substantial economic opportunities through the accurate prediction of stock prices. While Transformer-based models have demonstrated considerable efficacy … The stock market is a critical component of the global economy, offering investors substantial economic opportunities through the accurate prediction of stock prices. While Transformer-based models have demonstrated considerable efficacy in forecasting financial time-series data, their application in real-time trading and the processing of extended sequences is constrained by significant computational demands and substantial memory requirements. This paper introduces the Mamba model, a novel financial signal separation framework that combines state space models with graph neural networks. The model utilizes bidirectional Mamba blocks to capture long-term dependencies in historical price data, alongside adaptive graph convolution to depict the interactions among daily stock attributes, thereby achieving near-linear computational efficiency. Focusing on the Shanghai Stock Exchange Composite Index and conducting five independent evaluations, the model demonstrates exceptional performance in evaluating the short-term profit potential of stocks, with an average Relative Importance Coefficient (RIC) of 0.497. Additionally, the average Mean Absolute Error (MAE) and Root Mean Square Error (RMSE) values are 0.0875 and 2.3518, respectively, highlighting the model's high accuracy in its forecasts. The unique advantages of the Mamba model in analyzing financial signal structures effectively address the limitations encountered by traditional time-series forecasting models in modeling long-cycle dependencies and the interrelationships among different assets, thereby providing a novel methodological framework for developing interpretable prediction systems across diverse asset classes.

Formulating resource and capability configurations in complex contexts: a hybrid AHP-ANN approach for state-owned banks

2025-06-20

Arwan Ahmad Khoiruddin , Manahan Siallagan , Utomo Sarjono Putro +1 more | International Journal of Applied Decision Sciences

Using ARIMA Model Predict Stock Market

2025-06-20

Rui Peng | Advances in Economics Management and Political Sciences

The dynamic nature of the stock market, shaped by a complex interplay of economic and non-economic variables, presents both opportunities and challenges for investors seeking to optimize their returns. Prudent … The dynamic nature of the stock market, shaped by a complex interplay of economic and non-economic variables, presents both opportunities and challenges for investors seeking to optimize their returns. Prudent participation necessitates careful consideration of risk mitigation strategies, often involving the construction of diversified investment portfolios. In the pursuit of enhanced predictive capabilities, investors frequently turn to mathematical models to discern underlying patterns and anticipate future stock movements. However, the inherent complexity of financial markets means that various analytical approaches can yield divergent and potentially biased forecasts. Among the array of available methodologies, the Autoregressive Integrated Moving Average (ARIMA) model stands out as a widely utilized statistical tool for time series analysis and forecasting. Its appeal lies in its relative simplicity, relying exclusively on the historical behavior of the target variable itself, thereby obviating the need for external explanatory factors. Furthermore, under appropriate conditions and parameter selection, the ARIMA model has demonstrated the potential for achieving a considerable degree of predictive accuracy. This research endeavors to develop and implement an ARIMA model specifically tailored to forecasting the future stock trends of JPMorgan Chase & Co., a prominent player in the financial sector.

Product Price Optimization in Microeconomics: Exploring the Role of Artificial Intelligence Algorithms – A Literature Review

2025-06-20

Nur Fitri Hidayanti , Syaharuddin Syaharuddin , Zaenafi Ariani | Deleted Journal

In today’s data-driven economy, pricing strategies have become increasingly critical amid rapidly evolving market conditions. The integration of artificial intelligence (AI) offers new opportunities to optimize pricing decisions and strengthen … In today’s data-driven economy, pricing strategies have become increasingly critical amid rapidly evolving market conditions. The integration of artificial intelligence (AI) offers new opportunities to optimize pricing decisions and strengthen competitive advantage. This study investigates the use of AI algorithms in optimizing product pricing within microeconomic contexts. Using a qualitative method and systematic literature review, it draws on publications from the past decade indexed in Scopus, DOAJ, and Google Scholar. The findings highlight that AI-based price optimization is shaped by several key factors: data availability, algorithm complexity, and the alignment of AI systems with existing business models. However, major challenges such as data bias, limited computational resources, and insufficient organizational readiness often hinder successful implementation. Despite these barriers, AI shows great promise in enhancing pricing accuracy, efficiency, and adaptability to market fluctuations. This research offers a comprehensive overview of the limitations and potential of AI in price optimization, emphasizing the importance of addressing technical and organizational challenges. It contributes to a deeper understanding of how AI can transform traditional pricing strategies and encourages further empirical research to explore its real-world applications within dynamic microeconomic settings.Keywords - Artificial Intelligence, Microeconomics, Pricing Algorithms, Price Optimization.

The Role of Cloud-Based Vector Databases and Retrieval Augmented Generation (RAG) for Generative AI in Financial Markets Analysis

2025-06-19

S. Om Prakash | World Journal of Advanced Engineering Technology and Sciences

This scholarly article examines the transformative role of cloud-based vector databases and Retrieval Augmented Generation in enhancing generative artificial intelligence for financial markets evaluation. The convergence of these technologies creates … This scholarly article examines the transformative role of cloud-based vector databases and Retrieval Augmented Generation in enhancing generative artificial intelligence for financial markets evaluation. The convergence of these technologies creates powerful systems that overcome the constraints of standalone large language models by grounding outputs in specific, relevant financial information. Vector databases such as Pinecone, Weaviate, and Milvus enable efficient storage and retrieval of high-dimensional embeddings representing complex financial data, while RAG frameworks significantly improve accuracy, reduce hallucinations, and maintain temporal relevance in rapidly changing markets. Applications span semantic search of financial documents, enhanced sentiment assessment of market news, automated report generation, and more reliable financial forecasting. The advantages include improved accuracy and reliability, greater scalability and computational efficiency, enhanced explainability essential for regulatory compliance, and superior adaptability to changing market conditions. Despite significant benefits, implementation requires addressing challenges related to data security, regulatory compliance, technical integration, knowledge management, and organizational change. Financial institutions following best practices can leverage these technologies to gain deeper market insights and make more informed strategic decisions in increasingly complex global markets.

Fusion of Sentiment and Market Signals for Bitcoin Forecasting: A SentiStack Network Based on a Stacking LSTM Architecture

2025-06-19

Zhizhou Zhang , C. Jiang , Meiqi Lu | Big Data and Cognitive Computing

This paper proposes a comprehensive deep-learning framework, SentiStack, for Bitcoin price forecasting and trading strategy evaluation by integrating multimodal data sources, including market indicators, macroeconomic variables, and sentiment information extracted … This paper proposes a comprehensive deep-learning framework, SentiStack, for Bitcoin price forecasting and trading strategy evaluation by integrating multimodal data sources, including market indicators, macroeconomic variables, and sentiment information extracted from financial news and social media. The model architecture is based on a Stacking-LSTM ensemble, which captures complex temporal dependencies and non-linear patterns in high-dimensional financial time series. To enhance predictive power, sentiment embeddings derived from full-text analysis using the DeepSeek language model are fused with traditional numerical features through early and late data fusion techniques. Empirical results demonstrate that the proposed model significantly outperforms baseline strategies, including Buy & Hold and Random Trading, in cumulative return and risk-adjusted performances. Feature ablation experiments further reveal the critical role of sentiment and macroeconomic inputs in improving forecasting accuracy. The sentiment-enhanced model also exhibits strong performance in identifying high-return market movements, suggesting its practical value for data-driven investment decision-making. Overall, this study highlights the importance of incorporating soft information, such as investor sentiment, alongside traditional quantitative features in financial forecasting models.

An Evaluation of Machine Learning Models for Forecasting Short-Term U.S. Treasury Yields

2025-06-19

Y. Wang , Max Yue-Feng Wang , Liying Tu | Applied Sciences

This study explores the historical evolution and short-term predictive modeling of the U.S. 10-year Treasury bond yield, a critical indicator in global financial markets. Recognizing its sensitivity to macroeconomic conditions, … This study explores the historical evolution and short-term predictive modeling of the U.S. 10-year Treasury bond yield, a critical indicator in global financial markets. Recognizing its sensitivity to macroeconomic conditions, the research integrates economic variables, including the federal funds rate, core Consumer Price Index (CPI), real Gross Domestic Product (GDP) growth rate, and the U.S. federal debt growth rate, to assess their influence on yield movements. Four forecasting models are employed for comparative analysis: linear regression (LR), decision tree (DT), random forest (RF), and multilayer perceptron (MLP) neural networks. Using historical data from the Federal Reserve Economic Data (FRED), this study finds that the RF model offers the most accurate short-term predictions, achieving the lowest mean squared error (MSE) and mean absolute error (MAE), with an R2 value of 0.5760. The results highlight the superiority of ensemble-based nonlinear models in capturing complex interactions between economic indicators and yield dynamics. This research not only provides empirical support for using machine learning in economic forecasting but also offers practical implications for bond traders, system developers, and financial institutions aiming to enhance predictive accuracy and risk management.

Anomaly Detection in Financial Datasets Using Autoencoders

2025-06-19

Archit Yajnik , Vishal Sharma | International Journal of Science and Research (IJSR)

Enhancing BTST Trading with Sentiment Analysis and AI

2025-06-19

Rubik Ray , Subir Gupta , Vikas Bhandari +3 more | CRC Press eBooks

Stock investment selection via time series subpatterns and multi-relationship fusion

2025-06-18

Y. Shi , Jing Chi , Shu-Xu Yi +1 more | Pattern Recognition

Improved model of ELASTIC NET regularization for financial time series

2025-06-18

Roman Kvyetnyy , S.I. Borodkin | Optoelectronic Information-Power Technologies

This paper proposes a modification of Elastic Net regression for short-term forecasting of financial time series by introducing Gaussian weight decay. The new approach is designed to smooth the abrupt … This paper proposes a modification of Elastic Net regression for short-term forecasting of financial time series by introducing Gaussian weight decay. The new approach is designed to smooth the abrupt “jumps” between the last historical observation and the first forecast—an issue typical of standard regularization. To assess its effectiveness, we formally derive the Elastic Net model with four weighting schemes (no decay, linear, exponential, and Gaussian) and conduct empirical experiments on the S&P 500, Dow Jones Industrial Average, and Nasdaq Composite indices over the period 2020–2025. The results demonstrate that Gaussian decay minimizes the transition gap and achieves the lowest RMSE and Deviation for the S&P 500 and Nasdaq Composite, whereas exponential decay proves optimal for the Dow Jones Industrial Average.

The Application of Artificial Neural Networks in Financial Risk Assessment: A Methodical Approach

2025-06-18

Bounefla Alia , Bakhtiar Ali | Financial Engineering

This study explores the use of artificial neural networks (ANN) in finance, where they excel in risk assessment. An overall overview is presented where the basic principles of artificial neural … This study explores the use of artificial neural networks (ANN) in finance, where they excel in risk assessment. An overall overview is presented where the basic principles of artificial neural networks are defined. As well as the characteristics of these networks including how to solve complex problems that cannot be solved by classical mathematical techniques. This makes them well suited to their applications in the field of financial analysis and management. The study reveals the key elements for successful ANN adoption using a neural network model with five input factors to the analysis system (wide availability of top-notch data, strong and reliable technology setup, training qualified personnel, transparent moral principles and tactical execution strategy). The output of the system is the determination to what extent companies can apply this tool, helping to identify areas for improvement and facilitating the right choices. This methodical strategy offers a structured way to understand and apply ANN in today's business systems, ensuring the best use of this tool.

A contrastive deep learning framework for measuring central bank monetary policy scores

2025-06-18

Daqing Tian , Zhongjian Feng , Ran Jiang | International Journal of Financial Engineering

Machine learning and deep learning algorithms have recently been applied to analyze central bank communication texts, thereby providing valuable insights for financial market forecasting. However, as most deep learning methods … Machine learning and deep learning algorithms have recently been applied to analyze central bank communication texts, thereby providing valuable insights for financial market forecasting. However, as most deep learning methods require thousands or even more training examples, data scarcity often stands in the way when dealing with monetary policy report texts, especially for central banks in developing countries, which communicate their policies less frequently. To address this, we propose a contrastive deep learning framework designed to operate efficiently with small datasets. Despite being trained on fewer than 200 training samples, excellent performance was demonstrated by applying this modeling framework in two scenarios: Measuring China’s central bank monetary report’s hawkish-dovish score and predicting its next quarter’s tightening-easing moves.

Application of the Deep Learning Integrated Framework CEEMDAN-GRU-Informer in Financial Time Series Prediction

2025-06-18

Fang Liu , Chen Liang , Zhihan Jia | Computational Economics

Enhancing Equity Research with AI: A LangChain-Based News Analysis Framework

2025-06-18

K S Thejaswini | International Journal for Research in Applied Science and Engineering Technology

The advancement in chatbot technology and large-scale data processing has significantly transformed financial analysis and equity research and many other things. This research presents a LangChain-based framework to process financial … The advancement in chatbot technology and large-scale data processing has significantly transformed financial analysis and equity research and many other things. This research presents a LangChain-based framework to process financial documents, including company reports and market trends, to generate actionable investment insights. By integrating state-of-theart technologies such as Large Language Models (LLMs) and advanced Natural Language Processing (NLP) techniques, the framework will support efficient data extraction, semantic analysis, and response generation. The proposed system begins by extracting key information from PDF documents, such as company financials and stock performance reports, which are then processed and segmented into manageable text chunks. These chunks are embedded into high dimensional vectors using techniques like Word2Vec or Doc2Vec, allowing the system to capture semantic relationships and store them in a semantic knowledge base. The knowledge base is further enhanced with tools like FAISS for efficient similarity search and information retrieval. In the second phase, the system responds to user queries by analyzing the context and questions posed. It retrieves relevant information from the knowledge base and generates responses using a generative AI model like GPT, ensuring high relevance and accuracy. The research also compares the system’s efficiency in answering various types of investment-related questions, showcasing the chatbot's capability to assist users in making informed decisions. The framework’s versatility and scalability, supported by cutting edge AI models and semantic search, demonstrate its potential to revolutionize the way equity research is conducted, providing financial analysts and investors with a more efficient and accessible tool for market analysis in the digital age.

Stock Prices Forecasting by Using a Novel Hybrid Method Based on the MFO-Optimized GRU Network

2025-06-18

Xinjian Zhang , Guanlin Liu | Annals of Data Science

Decoding the Nifty50 Puzzle with ANN: FIIs, DIIs and Market Magic

2025-06-18

Neha Seth , Saif Siddiqui | Finance Theory and Practice

In the context of a dynamic and highly competitive financial market, understanding the behaviour of various categories of investors becomes a key factor in developing effective investment strategies and forecasting … In the context of a dynamic and highly competitive financial market, understanding the behaviour of various categories of investors becomes a key factor in developing effective investment strategies and forecasting market trends. This study examines the phenomenon of Nifty50 — a group of the 50 most liquid and significant stocks in the Indian stock market, which play an important role in forming the overall index. The purpose of the study is to determine the influence of foreign institutional investors (FIIs) and domestic institutional investors (DIIs) on the dynamics of Nifty50.The authors apply classical research methods: correlation analysis, a statistical model for analysing and forecasting the volatility of time series (GARCH), and artificial neural networks (ANN).The study is based on daily data on investments from the two specified groups of investors and the values of the Nifty50 index of the National Stock Exchange. The study period from 31.12.2019 to 30.11.2023 is divided into two sub-periods: before COVID‑19 and after. In periods of economic shocks, such as the COVID‑19 pandemic, the behaviour of these two types of investors becomes particularly contrasting. The results of the study showed that FIIs and DIIs are opposite to each other: when FIIs invest, DIIs are net sellers, and when FIIs sell, DIIs are net investors. In the context of the pandemic, FIIs often increased their investments in Indian assets, while DIIs, on the contrary, reduced their positions. However, in the post-pandemic period, the situation changed: DIIs began to play a more significant role in the dynamics of Nifty50, while the influence of FIIs decreased. Thus, the analysis of the interaction between FIIs and DIIs allows us to conclude the complex and multifaceted nature of the influence of institutional investors on the Indian stock market. Their strategies and behaviour have a significant impact on market indices and volatility, which requires careful monitoring and analysis for effective management of investment risks and making informed decisions in unstable conditions.

Forecasting Apple Inc. Stock prices: A comparative analysis of ARIMA, LSTM, and ARIMA-LSTM models

2025-06-18

Yiyang Zou | Deleted Journal

Stock price fluctuation and prediction is a problem that has attracted much attention. There exist many mathematical and statistical problems behind it. In essence, the key to solving this problem … Stock price fluctuation and prediction is a problem that has attracted much attention. There exist many mathematical and statistical problems behind it. In essence, the key to solving this problem lies in capturing the linear and nonlinear characteristics in the time series to predict future price movements. This study investigates the predictive capabilities of two distinct methodologiesLong Short-Term Memory (LSTM) networks and Autoregressive Integrated Moving Average (ARIMA) modelsusing Apple Inc. (AAPL) stock price data spanning 2016 to 2024. By synthesizing theoretical frameworks with empirical analysis, the research evaluates how each model captures linear trends and nonlinear fluctuations, ultimately proposing a hybrid ARIMA-LSTM architecture to enhance forecasting accuracy. Finally, according to the principal characteristics of the two models, the ARIMA-LSTM hybrid model is constructed. The results show that the hybrid model significantly outperforms single models in terms of RMSE and directional accuracy. Combined with error distribution visualization and volatility analysis, the hybrid model demonstrates efficient performance in achieving prediction optimization through the decomposition of linear and nonlinear components. It provides a new methodological perspective for financial time series modeling.

Enhancing stock timing predictions based on multimodal architecture: Leveraging large language models (LLMs) for text quality improvement

2025-06-18

Mingming Chen , Yifan Tang , Qi Qi +4 more | PLoS ONE

This study aims to enhance stock timing predictions by leveraging large language models (LLMs), specifically GPT-4, to filter and analyze online investor comment data. Recognizing challenges such as variable comment … This study aims to enhance stock timing predictions by leveraging large language models (LLMs), specifically GPT-4, to filter and analyze online investor comment data. Recognizing challenges such as variable comment quality, redundancy, and authenticity issues, we propose a multimodal architecture that integrates filtered comment data with stock price dynamics and technical indicators. Using data from nine Chinese banks, we compare four filtering models and demonstrate that employing GPT-4 significantly improves financial metrics like profit-loss ratio, win rate, and excess return rate. The multimodal architecture outperforms baseline models by effectively preprocessing comment data and combining it with quantitative financial data. While focused on Chinese banks, the approach can be adapted to broader markets by modifying the prompts of large language models. Our findings highlight the potential of LLMs in financial forecasting and provide more reliable decision support for investors.

Dynamic-aware hypergraph learning for stock recommendation

2025-06-18

Shaoqi Ma , Jiacheng Han , Chao Luo | Engineering Applications of Artificial Intelligence

TD-HCN: A trend-driven hypergraph convolutional network for stock return prediction

2025-06-18

Lexin Fang , Tianlong Zhao , JaeYong Yu +3 more | Neural Networks

The impact of social media and the pandemic on stock price prediction

2025-06-18

Hui Shi , Dazhi Chong , Weijun Zheng | Journal of Management Analytics

Integrating Financial Knowledge for Explainable Stock Market Sentiment Analysis via Query-Guided Attention

2025-06-18

Chuanyang Hong , Qingyuan He | Applied Sciences

Sentiment analysis is widely applied in the financial domain. However, financial documents, particularly those concerning the stock market, often contain complex and often ambiguous information, and their conclusions frequently deviate … Sentiment analysis is widely applied in the financial domain. However, financial documents, particularly those concerning the stock market, often contain complex and often ambiguous information, and their conclusions frequently deviate from actual market fluctuations. Thus, in comparison to sentiment polarity, financial analysts are primarily concerned with understanding the underlying rationale behind an article’s judgment. Therefore, providing an explainable foundation in a document classification model has become a critical focus in the financial sentiment analysis field. In this study, we propose a novel approach integrating financial domain knowledge within a hierarchical BERT-GRU model via a Query-Guided Dual Attention (QGDA) mechanism. Driven by domain-specific queries derived from securities knowledge, QGDA directs attention to text segments relevant to financial concepts, offering interpretable concept-level explanations for sentiment predictions and revealing the ’why’ behind a judgment. Crucially, this explainability is validated by designing diverse query categories. Utilizing attention weights to identify dominant query categories for each document, a case study demonstrates that predictions guided by these dominant categories exhibit statistically significant higher consistency with actual stock market fluctuations (p-value = 0.0368). This approach not only confirms the utility of the provided explanations but also identifies which conceptual drivers are more indicative of market movements. While prioritizing interpretability, the proposed model also achieves a 2.3% F1 score improvement over baselines, uniquely offering both competitive performance and structured, domain-specific explainability. This provides a valuable tool for analysts seeking deeper and more transparent insights into market-related texts.

The Calendar Effect of Online Shopping Festivals on China’s Financial Market: Evidence from China’s Singles’ Day Shopping Festival

2025-06-18

Tianyu Fu | Scientific journal of economics and management research.

Online Shopping Festival is a promotional activity created by those e-commerce platforms to set up a blockbuster which helps increase the general market value. In financial market, Calendar effect is … Online Shopping Festival is a promotional activity created by those e-commerce platforms to set up a blockbuster which helps increase the general market value. In financial market, Calendar effect is one of the most common effect of behavioral finance. In China, anomaly in November financial market is that the stock price will go up without a reason corresponding to the traditional Efficient Market Hypothesis. In this passage, Online Shopping Festival, particularly Single’s Day Shopping Festival, is treated as a traditional event of calendar effect. By using the Event Study Method framework and 83636 pieces of statistics from 489 companies, it is verified that Single’s Day Shopping Festival has a positive calendar effect on China’s financial market, causing abnormal returns during the period of festival. The psychological and behavioral cause of this anomaly is also discussed.

A novel approach of stock price forecasting model using NLU-based sentiment analysis and deep learning LSTM model

2025-06-17

Ujjwal Mishra , Anupa Sekhar Dash | Neural Computing and Applications

Investigating the dynamics and uncertainties in portfolio optimization using the Fourier-Millen transform

2025-06-17

Muhammad Hilal Alkhudaydi , Aiedh Mrisi Alharthi | PLoS ONE

Many investors and financial managers view portfolio optimisation as a critical step in the management and selection processes. This is due to the fact that a portfolio fundamentally comprises a … Many investors and financial managers view portfolio optimisation as a critical step in the management and selection processes. This is due to the fact that a portfolio fundamentally comprises a collection of uncertain securities, such as equities. For this reason, having a solid understanding of the elements responsible for these uncertainties is absolutely necessary. Investors will always look for a portfolio that can handle the required amount of risk while still producing the desired level of expected returns. This article uses feature-based models to investigate the primary elements that contribute to the optimal composition of a specific portfolio. These models make use of physical analyses, such as the Fourier transform, wavelet transforms and the Fourier–Mellin transform. Motivated by their use in medical analysis and detection, the purpose of this research was to analyse the efficacy of these methods in establishing the primary factors that go into optimising a particular portfolio. These geometric features are input into artificial neural networks, including convolutional and recurrent networks. These are then compared with other algorithms, such as vector autoregression, in portfolio optimisation tests. By testing these models on real-world data obtained from the US stock market, we were able to obtain preliminary findings on their utility.