This study builds a quality-minus-junk factor (QMJ) for Chinese A-shares stock via partial least square method. Of the pillars contained in this factor, only the profitability indicators deliver consistent portfolio return, while growth, safety and dividend add little. QMJ retains positive but modest alpha after standard factors, largely overlapping with RMW, and it does help elevate the performance measured with Sharpe ratio of various portfolio. Its distinct value emerges when paired with size. The return of QMJ portfolio shows a convex curvature relation with SMB, generating clear evidence of market timing. Since the existing studies emphasize size factor is the main driving force stirring market in China, investors should consider the combination of quality with size, not using it alone.
A fundamental challenge in quantitative finance and organizational sciences is to quantify the time-lagged relationship between a company's internal organizational dynamics and its tangible outputs. This study introduces an analytical framework that models this relationship using a quantum-inspired holographic concept. Applying a multi-stage statistical pipeline to 10 years of public data from six firms, we tested time-series models representing innovation and collaboration. The analysis revealed this new framework functions as an effective analytical filter for out-of-sample data. The results indicate that per-company calibrated models—which synchronize objective output data with calibrated, conceptually paired proxy metrics (like R&D spend for innovation)—achieved statistically significant predictive power (p = 0.021) against financial productivity data, whereas traditional, single-variable models did not. The findings demonstrate this Holographic Evaluation System for Topologic Interrelational Analysis (HESTIA) framework is a robust method for developing and validating firm-specific models for profits and valuation. The process is analogous to an imaging technique, transforming noisy proxy signals from a latent signal domain into an interpretable map of a firm's unique operational rhythm. The resulting models visualize a firm's value-creation cycle with a statistically high likelihood of predicting future “productivity,” which in this study is defined as profit and share price.
The volatility of gold prices significantly influences global financial stability, necessitating the development of reliable models capable of producing precise forecasts to minimize investment risks and maximize profitability. Recently, both machine learning and deep learning approaches have gained significant traction for time series forecasting across scientific and industrial domains. In this paper, we propose the ForCNN model, which utilizes grayscale image-based input rather than traditional numerical data. This algorithm integrates the advantages of visual image representation of a time series and deep 2D convolution neural network to analyze and extract important features and generate accurate forecasts. We carried out extensive experiments on two real-world gold closing price datasets and showed ForCNN outperformed most of the state-of-the-art deep learning techniques such as MLP, CNN1D, LSTM, CNN-LSTM, BiLSTM, CNN-BiLSTM in terms of accuracy measures. Furthermore, portfolio performance evaluation using Cumulative Return, Average Daily Return, and Sharpe Ratio indicates that ForCNN achieves superior profitability and stronger risk-adjusted performance, thereby underscoring its effectiveness in practical financial forecasting applications.
We establish that the actions of influential opinion leaders in the digital currency markets and potential investors following their lead drive abnormal cryptocurrency returns. We develop a psychological and behavioral factor, named the herd behavior index, that detects the herd instinct of the investors in cryptocurrency markets and captures anomalies in cryptocurrency returns. Our finding shows that the herd behavior index can explain the variation in cryptocurrency returns. Moreover, there exists a timeseries relationship between abnormal returns and the investors’ herd instinct, and the herd behavior index consistently forecasts future digital currency returns. Finally, we find notable gains even after considering transaction costs by implementing a long/short trading strategy based on the herd behavior index.
The present study seeks to propose an online portfolio selection algorithm based on a pattern-matching approach considering fuzzy returns. Online portfolio selection algorithms are concerned with capital allocation to several stocks to maximize the portfolio return over the long run by deciding the optimal portfolio in each period. Contrary to literature algorithms that calculate daily returns merely based on the closing price of the day, this study considers fuzzy returns to recognize similar patterns more efficiently. The Fuzzy C-Means (FCM) method has been used to cluster fuzzy returns in the sample selection step. In the second step, a log-optimal portfolio with transaction costs has been used to optimize the portfolio. The suggested models have been applied to five datasets with unique characteristics to evaluate their performance. Results indicated these algorithms outperformed the algorithms proposed previously in the literature on online portfolio selection in terms of total return and risk-adjusted return.
