Financial Dynamics and Time Series: A Bibliometric and Systematic Study of Academic Production

This article presents a systematic bibliometric study on the application of time series models in the analysis of corporate finance over the period 2010–2025. Through the examination of 84 scientific articles indexed in academic databases, five main thematic axes are identified and synthesized: theoretical evolution of the models, methodological approaches, application areas, emerging trends, and the integration of artificial intelligence. The results demonstrate a shift from classical univariate models, such as ARIMA and SARIMA, to more robust and complex hybrid architectures, including LSTM-ARFIMA, which combines deep learning with traditional econometric techniques. These methodologies are applied in diverse contexts, including bankruptcy prediction, financial risk management, public policy design, and corporate performance evaluation. Furthermore, a growing relevance of artificial intelligence and machine learning is observed as complementary tools to address financial problems with high levels of volatility and nonlinearity. This methodological and thematic overview identifies research gaps and potential synergies between related disciplines. The study concludes by highlighting the need to develop more integrated and adaptive models that can improve the accuracy of financial projections and strengthen strategic decision-making in dynamic and uncertain environments. The theoretical and practical implications of the analysis are relevant to both academics and financial professionals interested in advanced methodologies for understanding current corporate dynamics.