Forecasting commodity prices remains a challenging task due to market volatility, structural breaks, and changing economic conditions. This study evaluates the forecasting performance of classical econometric, deep learning, convolutional, and Transformer-based models for aluminum futures prices. Daily aluminum futures data are analysed using 11 forecasting approaches. Forecasting experiments are conducted for two distinct evaluation periods representing the years 2022 and 2025 to assess the robustness of model performance under different market environments. The empirical results reveal substantial differences in forecasting performance across model families. Recurrent neural network architectures, particularly GRU and RNN, consistently achieve the lowest forecasting errors across most horizons. The classical ARIMA model remains highly competitive despite its relative simplicity. In contrast, Transformer-based models generally fail to outperform simpler alternatives and frequently produce higher forecast errors. Statistical comparisons based on Diebold–Mariano tests and Model Confidence Set procedures indicate that performance differences among the best-performing models are often limited, suggesting that increased model complexity does not necessarily translate into superior predictive accuracy. Overall, the findings highlight the importance of empirical model evaluation and demonstrate that parsimonious forecasting approaches can remain effective competitors to substantially more complex machine learning architectures in aluminum futures price forecasting.
László Vancsura, Tibor Tatay, Tibor Bareith et al.· Decision Making Advances· 0 citations
Copper is a strategically important commodity whose price dynamics are increasingly affected by structural changes, geopolitical shocks, and the global energy transition. These conditions create substantial challenges for forecasting models and provide a useful setting for evaluating the practical value of machine learning predictions. This study compares statistical and artificial intelligence-based forecasting models for copper price prediction under different market regimes and structural break conditions. Model performance is assessed using a multi-dimensional evaluation framework that combines statistical accuracy (MAPE), dynamic pattern reproduction (Taylor diagrams and time-lagged cross-correlation analysis), and the economic performance of forecast-driven trading strategies. The results reveal a consistent error–profit paradox: models with the highest statistical forecasting accuracy do not necessarily generate the best trading outcomes. In several cases, models with larger prediction errors achieve superior economic performance because they capture directional market dynamics more effectively. The analyses further show that structural breaks substantially alter model rankings and predictive usefulness, highlighting the importance of regime-aware evaluation. These findings suggest that forecast accuracy alone provides an incomplete assessment of model quality in financial and commodity forecasting applications. The study contributes to machine learning evaluation research by proposing an integrated framework that jointly considers predictive accuracy, temporal dynamics, model robustness, and economic utility, thereby offering a more comprehensive approach to assessing forecasting systems in real-world decision-making environments.
László Vancsura, Tibor Tatay, Tibor Bareith· Machine Learning and Knowled...· 0 citations
The value-creation capacity of active asset management remains one of the most debated issues within modern portfolio theory. While the relationship between costs and performance is typically negative in developed markets, in smaller and less liquid markets—where information asymmetry is more pronounced—higher fees may also be interpreted as signals of managerial ability. This study investigates this apparent contradiction in the context of the Hungarian equity mutual fund market, using a panel dataset covering 79 funds over the period 2017–2024, with a particular focus on identifying non-linear effects among performance determinants. The methodological framework combines fixed-effects panel regression with Driscoll-Kraay robust standard errors, complemented by quantile regression estimates to examine different segments of the return and alpha distributions. The results indicate that growth dynamics (NAV_change) and cumulative historical performance (Yield_from_start) consistently enhance fund performance, while the negative effect of past returns suggests the dominance of mean reversion. The impact of the total expense ratio (TER) proves to be non-linear and specification-dependent—a finding con-firmed by an extensive battery of robustness checks—thereby rejecting the cost-signalling hypothesis with respect to risk-adjusted excess returns (Jensen’s alpha). Quantile estimates further reveal that the effects of economies of scale and cost structure differ significantly between underperforming and top-performing funds, confirming that analyses based on average effects obscure the heterogeneity of market dynamics. By jointly modelling returns and risk-adjusted performance across the full conditional distribution, the study contributes a distribution-sensitive theoretical account of active management’s limitations in a small, less liquid market, showing that these limitations are conditional on fund size and cost structure rather than uniform across the fund population.
László Vancsura, Tibor Tatay, Tivadar Zakár et al.· Economies· 0 citations
