SHAP analysis reveals that trading activity, lagged liquidity, and market uncertainty are the main determinants of liquidity forecasts, and the findings highlight the complementary role of explainable machine learning in empirical finance.
Abstract
This paper examines the forecasting of liquidity dynamics in European stock markets by means of traditional econometric models and machine learning techniques. It uses daily data for the DAX, CAC 40, FTSE 100, FTSE MIB, and IBEX 35 over 2010–2026, liquidity being measured by the logarithmic Amihud illiquidity indicator. The empirical framework compares ARIMA models and a dynamic panel specification with Random Forest, Extreme Gradient Boosting (XGBoost), and Support Vector Regression (SVR) within a common rolling one-step-ahead forecasting framework. The results show that liquidity is highly persistent and that the dynamic panel model achieves the lowest forecast errors, although Diebold–Mariano tests indicate no significant predictive advantage over the leading machine learning models. SHAP analysis reveals that trading activity, lagged liquidity, and market uncertainty are the main determinants of liquidity forecasts. The findings highlight the complementary role of explainable machine learning in empirical finance.
This paper utilises high‐frequency data from the Chinese stock market and a panel of individual stocks to compare the forecasting performance of machine learning with popular econometric models across different periods. For short‐term volatility forecasting, most machine learning models outperform econometric models with limited explanatory variables, though they do not exhibit a significant advantage over the econometric model incorporating all features. For medium‐ and long‐term volatility forecasting, Light gradient boosting machine (LGBM) in machine learning substantially dominates econometric models. We also explore a simple‐to‐implement forecast combination method that leverages the best machine learning model and the best econometric model to explore if model averaging leads to any improvement. Our findings indicate that over a longer forecasting horizon, this method achieves the best performance among all forecast combinations and dominates the best econometric model.
Predicting stock prices is a complex challenge, par-ticularly in volatile and low-liquidity markets like the Bangladesh Stock Market, where economic, political, and market-specific factors significantly influence outcomes. This study compares state-of-the-art machine learning models for stock price predic-tion, utilizing historical market data to forecast short-term and long-term price movements. Algorithms including Long Short-Term Memory (LSTM) networks, Support Vector Regression (SVR), Prophet, and Autoregressive Integrated Moving Average (ARIMA) were employed. The dataset comprised historical price and trading volume data to enhance prediction accuracy. Model performance was evaluated using prediction accuracy, Root Mean Square Error (RMSE), Mean Absolute Percentage Error (MAPE), and Mean Absolute Error (MAE). Results indicate that the ARIMA model outperformed the other models considered. This research contributes to the field of stock market forecasting and provides a framework for improving investment strategies through advanced AI techniques
Hasan Mahmud· Science Set Journal of Econo...· 0 citations
This study examines the daily closing level of the Borsa Istanbul 100 (BIST100) Index using a unified framework that combines machine-learning (ML) and deep-learning (DL) methods with macro-financial, behavioral and market-based indicators.
The empirical model analyzes a dataset consisting of 3,222 trading days spanning January 2010 to October 2022. The explanatory variable set includes foreign ownership, the policy rates of the Central Bank of the Republic of Türkiye and the US Federal Reserve, the consumer price index, the USD/TRY exchange rate, the Amihud illiquidity measure, and an investor-sentiment index derived from securities investment-trust discounts. Forecasting performance is evaluated within a common model-comparison framework.
The results show that both ML and DL models provide strong forecasting performance for the BIST100 Index. Across specifications, macro-financial variables contain the strongest predictive information, while liquidity and sentiment contribute complementary explanatory content. Overall, the findings suggest that BIST100 dynamics are shaped by the joint influence of economic fundamentals, market sentiment and liquidity conditions. The study contributes to the BIST100 and emerging-market forecasting literature by evaluating these determinants within an integrated data-driven framework.
This study pioneers a unified forecasting framework that simultaneously evaluates macroeconomic fundamentals, market illiquidity and behavioral sentiment, moving beyond the isolated approaches common in the literature. Furthermore, it bridges the gap between high predictive accuracy and economic interpretability, employing rigorous lagged-return validations to capture genuine predictive alpha without look-ahead bias.
Alptekin Erdağ, M. F. Uçar, İ. Tarhan· Review of Behavioral Finance· 0 citations
Volatility forecasts play an important role in financial markets. They are used for setting position limits, determining margin requirements, and guiding short-term trading decisions. Financial volatility is known to exhibit persistence and regime dependence, which makes short-horizon forecasting challenging. In addition to historical volatility measures, implied volatility indicators such as the Volatility Index (VIX) are often used as forward-looking proxies for market risk expectations. However, it is not clear whether the predictive usefulness of the VIX is consistent across different neural network architectures. This study investigates the role of the VIX in one-day-ahead realized volatility forecasting and examines whether its contribution varies across alternative neural network models. The empirical analysis focuses on major global equity markets, including the S&P 500, DAX, FTSE 100, Nikkei 225, and Hang Seng Index. Using market data from these indices, the study evaluates the forecasting performance of artificial neural networks (ANN), long short-term memory networks (LSTM), and gated recurrent unit (GRU) models. All models are estimated within a rolling out-of-sample framework and are tested with and without the VIX variable. Forecast accuracy is evaluated using Mean Squared Error (MSE), Root Mean Squared Error (RMSE), Quasi-Likelihood (QLIKE) loss functions, and Diebold–Mariano tests. The results show that memory-based models outperform feedforward ANN models, while the contribution of the VIX varies across models and markets.
Orhan Özaydın, A. Kamil· Journal of Intelligent Decis...· 0 citations
This paper compares direct-yield and factor-based approaches to U.S. Treasury yield curve forecasting using a common high-dimensional macroeconomic information set. Forecasts are evaluated on monthly zero-coupon yields over the 2015-2025 out-of-sample period. Gains over the random walk are concentrated at short maturities and in slope forecasts, and decline with the forecast horizon. Direct-yield models perform best for slope forecasts and are relatively stronger at short horizons, while factor-based models become more competitive at longer horizons. Macroeconomic predictors provide clear incremental predictive power, strongest for slope-related movements. A trading simulation reinforces that macro-augmented models perform best in slope trades. The simulation also highlights a gap between statistical and economic performance, as the random walk is a strong benchmark under statistical loss but performs poorly as a trading signal.
Forecasting stock market returns remains a central challenge in finance, and much of the existing literature has relied on linear econometric models, which often struggle to capture nonlinear and time-varying patterns in equity markets. Prior studies have shown mixed evidence regarding the predictive value of technical indicators, leaving uncertainty about whether they contain meaningful information for short-term forecasting. This study addresses that gap by examining the predictability of weekly S&P 500 returns using traditional market-based indicators combined with modern machine learning methods. Weekly price and volume data from January 2000 to October 2025 are used to evaluate whether momentum, moving averages, volatility, and trading volume provide forecasting power for aggregate returns. Two ensemble algorithms, Random Forest and Histogram Based Gradient Boosting, are applied within an expanding window validation design. Results indicate that Gradient Boosting explains approximately 70% of the variation in weekly returns, demonstrating that adaptive learning approaches can reveal meaningful short-term market dynamics.
George Chang, Esther Ochieng· International Journal of Eco...· 0 citations