Search Results (427)

Stock market price prediction remains a persistent challenge owing to the non-stationarity, high noise content, and intricate spatiotemporal dependencies that characterize financial time series. Existing approaches typically excel at either local pattern extraction or long-range dependency modeling, yet seldom reconcile both within a unified framework. This paper introduces a CNN–Transformer dual-channel architecture equipped with a dynamic attention fusion module for stock price forecasting. The convolutional channel applies hierarchical dilated convolutions to distill fine-grained local patterns from multi-indicator sequences while suppressing high-frequency noise. Simultaneously, the Transformer channel employs multi-head self-attention to capture long-distance temporal correlations and regime-shift dynamics. A learnable gating mechanism then fuses the two feature streams by adaptively weighting local detail against global trend information according to market conditions. Experiments conducted on four real-world stock datasets spanning the S&P 500, CSI 300, NASDAQ Composite, and Hang Seng Index show that the proposed model reduces mean absolute error by 9.7–15.3% and root mean square error by 9.5–13.8% relative to competitive baselines including LSTM, CNN–LSTM, Informer, and PatchTST. Ablation studies further indicate that both channels and the fusion module contribute to prediction accuracy, and the architecture remains effective across markets with differing volatility profiles. Full article

This study examines the asymmetric relationship between housing prices and stock market returns across China’s major economic regions. Specifically, it investigates whether positive and negative shocks exhibit different transmission dynamics and whether these dynamics vary across regions characterized by different levels of financial development and housing market maturity. Using monthly data from 2005 to 2024, the study employs region-specific asymmetric vector autoregression (VAR) models, asymmetric Granger causality tests, and generalized impulse response analysis based on asymmetric decompositions of housing prices and stock market returns. The results suggest that statistically significant housing-to-stock predictability is observed primarily following negative housing price shocks in selected regions, whereas positive shocks generally exhibit weaker or statistically insignificant predictive effects. Conversely, positive stock market shocks generally provide more consistent evidence of stock-to-housing predictability, particularly in the Eastern and Central regions, although the responses are more mixed in the Western region and vary in magnitude, statistical significance, and persistence across regional markets. Overall, the results provide evidence of heterogeneous dynamic transmission patterns across China’s major economic regions and suggest that housing-related downside risk may represent an important source of regional macro-financial vulnerability. Full article

Accurate next-day stock price forecasting remains challenging because daily price changes have a low signal-to-noise ratio and can be strongly affected by short-lived news shocks, order-flow imbalances, and abrupt changes in volatility or market sentiment. This study presents a controlled empirical comparison of deep learning architectures for next-day stock price forecasting using technical indicators. Using a decade-long daily dataset covering four large-cap NASDAQ equities (AAPL, META, SBUX, and TSLA), multivariate input sequences are constructed by combining historical prices with five widely used technical indicators: exponential moving average (EMA), relative strength index (RSI), moving average convergence divergence (MACD), on-balance volume (OBV), and average true range (ATR). Four deep sequence architectures—long short-term memory (LSTM), bidirectional LSTM (BiLSTM), gated recurrent unit (GRU), and convolutional LSTM (ConvLSTM)—are evaluated across multiple lookback windows (5, 15, and 30 trading days) and chronological train/validation/test splits (60–20–20, 70–15–15, and 80–10–10). Hyperparameters are optimized through random search, and forecasting performance is assessed on held-out test sets using normalized-scale root mean squared error (RMSE) and out-of-sample $R^2$. Within the examined fixed chronological partitions, ConvLSTM records the lowest observed RMSE for all four equities, attaining values between 0.0256 and 0.0394 and out-of-sample $R^2$ values above 0.90. Because the evaluation does not include walk-forward validation or formal statistical significance testing, these results should be interpreted as descriptive evidence within the present experimental setting rather than as proof of general architectural superiority. To assess practical utility, forecasts are translated into a transparent long-only trading rule that enters the market when the predicted next-day closing price exceeds the current closing price. Out-of-sample backtesting shows that the frictionless forecast-driven strategy achieves higher terminal cumulative returns than Buy-and-Hold for AAPL, SBUX, and TSLA, while Buy-and-Hold remains superior for META. Approximate five-day-frequency risk-adjusted estimates generally reinforce these relative patterns: the ConvLSTM strategy improves the Sharpe, Sortino, and Calmar ratios for AAPL, SBUX, and TSLA, although TSLA remains exposed to substantial drawdown risk. Transaction-cost sensitivity analysis further indicates that the terminal-return gains weaken under trading frictions and are particularly sensitive for AAPL. The findings demonstrate the value of evaluating forecasting architectures through both statistical and financial criteria, while emphasizing that lower point-forecast error does not necessarily translate into superior economic or risk-adjusted performance. Full article

Financial markets are increasingly shaped by heterogeneous information sources, including trading behaviors, order-book dynamics, textual events, investor sentiment, and macroeconomic conditions. These signals are often asynchronous, noisy, partially missing, and associated with different degrees of reliability, which makes trustworthy financial risk early warning challenging. To address these issues, this study proposes an uncertainty-aware multimodal financial sensing network, termed UAMF-Net. The model treats price series, trading volume, order books, news texts, investor sentiment, and macroeconomic variables as financial sensing signals and integrates them through three task-oriented modules. First, the asynchronous multimodal temporal alignment module uses absolute time encoding, relative interval modeling, event-lag representation, temporal gating, and target-time-guided cross-scale attention to align cross-frequency financial signals according to their relevance to the prediction time. Second, the risk-aware multimodal soft fusion module estimates modality-level risk contribution and signal reliability by combining fuzzy risk membership, modality confidence weights, and cross-modal consistency constraints. Third, the uncertainty-aware risk early warning module adopts evidential learning to generate nonnegative class evidence, derive risk-category probabilities from Dirichlet parameters, estimate predictive uncertainty from total evidence strength, and jointly predict continuous risk intensity. Experimental results show that UAMF-Net achieves the best overall performance, with Accuracy, Precision, Recall, F1-score, Macro-F1, ROC-AUC, and PR-AUC reaching $0.882$, $0.864$, $0.849$, $0.856$, $0.839$, $0.941$, and $0.824$, respectively, while ECE and Brier score are reduced to $0.037$ and $0.096$. Under severe temporal asynchrony, UAMF-Net maintains an Accuracy of $0.849$, a Macro-F1 of $0.797$, and a PR-AUC of $0.774$. Under the missing multiple modalities setting, it achieves an Accuracy of $0.842$ and a Macro-F1 of $0.788$. The uncertainty analysis further shows that Risk Precision@90% reaches $0.889$. Validation on FNSPID, Daily News, and StockEmotions also confirms its generalization ability across public financial benchmarks. These results indicate that UAMF-Net improves financial risk early warning by jointly modeling temporal asynchrony, modality reliability, and predictive uncertainty. Full article

This study provides new evidence on the predictability of confidence indices in China’s stock returns. We demonstrate that, during the sample period from January 2005 to December 2022, the business confidence index (BCI) positively and significantly predicted subsequent stock market returns, outperforming mainstream economic predictors and other confidence indices. Further, for the pricing effectiveness of the stock market, the BCI and investor sentiment provide complementary sources of information. The predictive power of confidence indices for stock market returns declined significantly during the COVID-19 pandemic. Meanwhile, confidence indices predicted better during bear market periods compared to bull market periods. Finally, in practical investment applications, the BCI and alternative confidence index produce appreciable economic gains for investors. These empirical results also pass the robustness test. Full article

Accurate stock price forecasting remains a challenging problem due to the noisy, nonlinear, and non-stationary characteristics of financial time series. Although recent advances in deep learning have improved predictive capabilities, most prior studies evaluate forecasting models primarily using statistical error metrics, with limited consideration of their practical value in trading and investment decision-making. This creates a gap between predictive performance and real-world economic utility. This study proposes a decision-oriented evaluation framework for multi-step stock price forecasting that jointly assesses predictive accuracy and trading profitability within a unified experimental setting. Using data from 91 ASX 100 stocks after data cleaning, with a testing period spanning 2019–2020, several deep learning architectures, including Multi-Layer Perceptron (MLP), Gated Recurrent Unit (GRU), Seq2Seq, and attention-based sequence models, are systematically compared under identical training and trading conditions. The results show that the Seq2Seq model achieved the best overall performance, obtaining the lowest average MAPE of 0.0293 and the highest ROI of 23.2%, while the attention-based model achieved a similar MAPE of 0.0294 but a lower ROI of 12.4%. Although differences in forecasting accuracy were relatively small, the Seq2Seq model achieved the highest observed trading profitability and generated a higher observed return than a passive market benchmark under the proposed evaluation framework. These findings suggest that evaluation based solely on prediction accuracy may not fully capture the practical value of forecasting models. Full article

This study proposes a hybrid deep learning model named LSTM-CBAM-Time2Vec-KAN based on symmetry awareness and optimized multi-module collaborative features, aiming to improve the accuracy and stability of stock price prediction. To address common shortcomings in traditional forecasting models such as insufficient feature extraction, difficulties in parameter optimization, and inadequate utilization of temporal characteristics, the research innovatively exploits the symmetry inherent in financial time series, particularly their temporal periodicity and cross-dimensional feature consistency, to construct an intelligent prediction framework that integrates multiple modules. First, wavelet transform is applied to perform multi-scale decomposition and signal reconstruction on the raw stock price sequence, effectively extracting high signal-to-noise ratio features. Second, the Northern Goshawk Optimization (NGO) algorithm is employed to jointly optimize key hyperparameters of the model, including the LSTM hidden layer dimension and CBAM compression ratio, thereby resolving the challenge of parameter coupling across modules. Third, the CBAM attention mechanism enhances the importance of temporal features extracted by LSTM through a dual mechanism of channel and spatial attention, enabling the model to focus on critical price movement points. Meanwhile, Time2Vec encoding transforms temporal information into embedding representations with periodic properties, effectively capturing cyclical patterns at daily, weekly, and monthly trading intervals. Finally, the Kolmogorov–Arnold network (KAN) fuses multimodal features and produces precise predictive outputs. Experimental results show that the proposed model significantly outperforms all baseline models in four evaluation metrics, namely mean absolute error (MAE), root mean square error (RMSE), mean absolute percentage error (MAPE), and coefficient of determination (R²), which verifies its superior prediction accuracy and robustness. Furthermore, analyses of stock price forecasting under different time spans and simulated trading performance under various trading strategies further demonstrate that this study provides a feasible and effective technical solution for financial time-series forecasting, with important theoretical research value and practical application value. Full article

This study investigates whether volatility-related variables and traditional financial predictors can explain and forecast Taiwan stock returns. Using the Taiwan Stock Exchange Capitalization Weighted Stock Index (TAIEX) and eight major Taiwan industry indices, we examine the predictive ability of long-term adjusted volatility (LVadj), short-term volatility (SV), the volatility index (VIX), idiosyncratic volatility (IVOL), the earnings-to-price ratio (EP), the book-to-market ratio (BM), and the turnover ratio (TURN). The sample period spans from January 2007 to December 2025. Univariate and bivariate predictive regression models are estimated using ordinary least squares (OLS) and exponentially weighted least squares (EWLS). The empirical results show that SV exhibits the strongest in-sample explanatory power for TAIEX returns, whereas TURN plays a more important role in explaining industry-level returns. Out-of-sample forecasting performance varies considerably across industries. For TAIEX returns, the combination of LVadj and TURN provides the strongest forecasting performance, while models incorporating SV and VIX perform relatively well in several industry sectors. Overall, the results suggest that volatility-related variables provide useful predictive information under certain model specifications and industry sectors, with EWLS generally outperforming conventional OLS estimation. Full article

This study proposes a time series-aware Large Language Model (LLM)-driven feature engineering framework for tabular prediction tasks. Existing automated feature engineering methods, including LLM-based approaches such as CAAFE and OCT-Tree and established libraries such as tsfresh and Featuretools, can generate useful features for general tabular data but do not explicitly address temporal availability constraints in time series settings. This can lead to data leakage when variables that are only available after the prediction event are used directly during model training. To address this limitation, the proposed framework classifies variables into antecedent features, consequent features, and historical aggregated features. The key innovation is that consequent variables are not discarded to prevent leakage but are instead routed into a leakage-safe historical aggregation pipeline, recovering predictive signal from post-event variables through temporally valid past values. The framework guides an LLM to generate structured feature engineering configurations, applies temporally valid transformations, performs feature selection, and evaluates the selected features using predictive models. A formal leakage control mechanism ensures that all aggregations use strictly past observations, applied within entity groups and before the temporal train–validation–test split. The framework is evaluated on two time series tabular tasks: Tesla stock prediction and English Premier League match outcome prediction. The results show that the proposed approach improves predictive performance compared with raw-feature baselines and selected existing automated feature engineering methods. On the Tesla dataset, the framework reduced MAE compared with both the baseline and the reported OCT-Tree result. On the EPL dataset, it improved accuracy compared with the odds-only baseline and the reported Azure ML preprocessing result. These findings suggest that combining LLM reasoning with explicit temporal constraints is a practical direction for automated feature engineering in time series tabular machine learning. Full article

This study examines the relationship between valuation multiples and investment performance in the U.S. stock market. Specifically, it tests whether portfolios constructed with high-multiple stocks consistently outperform portfolios with low-multiple stocks. The analysis spans the Technology, Consumer Staples, and Healthcare sectors from 2018 to 2022. A sector-based portfolio construction framework was employed using quarterly portfolio-return data. Quantitative financial modelling, including regression analysis and descriptive statistics, was applied to assess the correlation between portfolio returns and valuation multiples (P/E and EV/EBITDA), while interpreting results within the broader context of market volatility and the COVID-19 period. The results show no statistically significant relationship between valuation multiples and portfolio performance. Low-multiple portfolios demonstrated marginally higher average returns over the period, offering weak support for value-based investment strategies. Results further suggest limited standalone predictive power in high-multiple valuations. Drawing on the Efficient Market Hypothesis, Value Investing, Growth Investing, and the Fama-French Three-Factor Model, this paper empirically tests the impact of valuation multiples within a sector-based portfolio framework. Accordingly, the study adds to the asset pricing literature by offering a structured null-result framework, demonstrating that valuation multiples, when applied in isolation, may not provide sufficiently reliable standalone signals for portfolio performance. The COVID-19 period is interpreted as an economically meaningful contextual regime characterized by elevated volatility, liquidity intervention, and sectoral divergence, rather than as a formally estimated event-study framework. Full article

Artificial intelligence (AI) and machine learning (ML) are increasingly important in financial technology (FinTech) applications involving large datasets, uncertainty, and complex decision-making. First, this paper presents a review of AI- and ML-based approaches in FinTech from 2010 to 2025, with particular emphasis on the relationship between predictive analytics and optimization-based decision-making. The review identifies two major research streams: (i) predictive AI/ML models for financial forecasting, stock price prediction, risk management, and fraud detection and (ii) optimization approaches for constrained financial decision problems, including portfolio optimization, asset–liability management, and risk-based decision-making. These two streams have largely evolved independently, which creates challenges in real financial environments, where uncertainty in predictions directly affects decision quality. Secondly, the paper also provides a decision-oriented perspective on how AI/ML-based predictions can support optimization under uncertainty and practical financial constraints. It highlights the role of uncertainty-aware optimization, simulation-based methods, and hybrid approaches such as simheuristics in improving the robustness of financial decision-making. Finally, the paper identifies open research directions toward integrated financial decision-support frameworks that combine predictive analytics, optimization, and simulation to address dynamic and uncertain FinTech environments. Full article

Stock index price series are composed of superimposed multi-frequency components, including long-term trends, cyclical fluctuations, and stochastic noise. Effectively decoupling these heterogeneous components and modeling them separately is key to improving forecasting accuracy. Existing methods under the “decomposition–prediction” paradigm mostly employ fixed-scale decomposition, and the forecasting models are not specifically adapted to the non-stationary and high-noise characteristics of financial data, resulting in limitations in adaptivity and local dynamic capture. This paper proposes a frequency-aware adaptive multi-scale decomposition Transformer hybrid model (FAMS-Transformer). At the decomposition level, the fast Fourier transform is used to dynamically identify dominant cycles, thereby adaptively decoupling trends and fluctuations, overcoming the limitations of fixed-scale decomposition. At the forecasting level, a lightweight depthwise separable convolution is embedded between the self-attention and feedforward network of the Transformer encoder, enhancing the model’s ability to capture local temporal dynamics and achieving collaborative modeling of global dependencies and local information. Comparative experiments with 15 baseline models including LSTM, Transformer, TimesNet, and FreTS on three representative Chinese market indices—Shanghai Composite Index, Shenzhen Component Index, and Small and Medium Enterprises 100 Index—across four prediction horizons from one step to 15 steps demonstrate that FAMS-Transformer achieves the best forecasting accuracy in all scenarios. The coefficient of determination for 15-step prediction remains stably between 0.730 and 0.928. Moreover, the model still performs well on the S & P 500 dataset. Ablation studies and significance tests further validate the effectiveness of each core module and the statistical significance of the performance improvements. Full article

Against the backdrop of intensifying global climate change, climate policy uncertainty (CPU) and investor sentiment have become critical factors influencing the stability of financial markets. In this study, a quantitative index of investor sentiment is constructed using stock trading volume, turnover rate, price-to-earnings ratio, circulating market value, and the consumer confidence index. The QVAR-DY model is employed to analyze the risk contagion mechanisms among CPU, investor sentiment, and China’s financial sub-markets across different quantiles. Furthermore, five machine learning models—LSTM, BiLSTM, CNN, XGBoost, and LightGBM—are used to forecast risk spillover indices, and their performance is compared with three benchmark models (ARIMA, Persistence, and HistMean) to systematically evaluate the advantages of machine learning models in capturing tail risk spillover effects. The findings reveal significant cross-market risk contagion in financial markets, characterized by asymmetry. The level of risk spillover under extreme conditions is substantially higher than under normal conditions, indicating high sensitivity to extreme events and major policies. CPU exhibits the most pronounced spillover effect on the money market, while investor sentiment has the greatest impact on the stock market. The stock, real estate, and commodity markets act simultaneously as sources of risk and receivers of shocks. In terms of forecasting performance, LightGBM performs best under normal conditions, whereas LSTM achieves the highest prediction accuracy under extreme conditions. Full article

This study proposes a hybrid forecasting framework that integrates Kalman Filtering (KF), Markov Switching (MS), and nonlinear recurrent learning for stock-index prediction. The KF component smooths short-term price noise, the MS model identifies latent return–volatility regimes, and the LSTM/GRU components learn nonlinear temporal patterns from regime-conditioned information. The framework is evaluated using the CSI 300, S&P 500, and Nikkei 225 indices through forecasting-accuracy measures, Bootstrap Diebold–Mariano tests with Modified Bayes Factor evidence, out-of-sample trading simulations, and robustness checks. The empirical results show that regime conditioning is the primary source of forecasting and economic improvement. KF–MS–LSTM performs best for the CSI 300 and Standard MS performs strongest for the S&P 500, while KF–MS–LSTM and KF–MS–GRU are more competitive for the Nikkei 225. In contrast, models without regime information, including pure LSTM/GRU and the standalone Transformer, generally exhibit weaker forecasting and trading performance. The findings suggest that latent market-state information is more important than neural-network complexity alone for robust financial forecasting, while the incremental value of Kalman filtering and recurrent learning remains market dependent. Overall, the results support regime-aware forecasting as an interpretable and economically meaningful approach for stock-index prediction under heterogeneous market environments. Full article

Volatility is a fundamental indicator for assessing the risk of financial assets. By integrating unstructured data, such as earnings call transcripts, the limitations of traditional time series data can be transcended, enabling collaborative forecasting from multiple data sources, enhancing the robustness of volatility prediction, and improving the efficiency of risk management. Although current research has effectively utilized earnings call data to predict asset volatility, price trends, and stock correlations, it often overlooks the inherent challenges of integrating textual and time series data, as well as the self-exciting and clustering characteristics of financial events. While conventional Long Short-Term Memory (LSTM) networks excel in processing fused data, they lack the structural capacity to explicitly model event-driven temporal decay, often failing to differentiate the varying influence of historical shocks over time. To surmount this limitation, we have significantly enhanced the predictive model by focusing on extracting salient information and integrating temporal dependency modeling with dynamic state adjustment mechanisms. The core innovation is introducing the Hawkes process to explicitly capture the self-exciting effect of financial events, which is the key to modeling volatility clustering around earnings releases. The proposed Hawkes LSTM model introduces a decay gating module and a textual information knowledge enhancement module. The decay gating module is specifically designed to more effectively capture the temporal dependencies between events within an event sequence. This allows the model to focus more on recent significant events, with the influence of an event on subsequent events typically diminishing as the temporal interval between them increases. By integrating temporal dependency modeling, the model is enabled to utilize historical data in a more flexible manner. The dynamic state adjustment mechanism further enhances its capacity to capture dynamically changing characteristics. Together, these features provide a more robust and precise solution for volatility prediction. Experimental results on two real-world earnings call datasets show that this approach significantly outperforms existing benchmark models on most prediction horizons, achieving competitive and superior performance and verifying its effectiveness and robustness. Full article