Artificial Intelligence Applications in Stock Investment: The Impact of Machine Learning and Big Data on Investment Performance

The present study investigated the impact of Artificial Intelligence (AI) applications, in terms of the Machine Learning Integration (MLI) and the Big Data Analytics Utilization (BDAU), on stock investment performance in Small and Medium Enterprises (SMEs). The study adopts a broad perspective, focusing not only on algorithmic correctness but also on Dynamic Capability Theory (Teece, 2007) and the Technology-Organization-Environment (TOE) model (Tornatzky and Fleischer, 1990). In this view, the performance of investment is seen as a multidimensional process which is determined not only by technological tools, but also by the organizational capability and human preparedness. A purposive and stratified survey of 612 SME managers and investors in Vietnam, Malaysia, and Indonesia was undertaken to determine and collect data, and 385 valid responses were selected to proceed with the analysis. The regression outcomes revealed that Machine Learning Integration corresponded to the most powerful contributor of the enhanced investment performance (β = 0.736, p < 0.05), with Big Data Analytics Utilization also performing significantly (β = 0.714, p < 0.05). Moreover, Investor Technological Readiness (ITR) was identified to mediate the association between BDAU and investment performance (β = 0.520, p < 0.05), which highlights the significance of human capability and belief in transforming data-driven information to profitable investment policies. The results are theoretical as they reveal that digital technologies do not necessarily lead to success; their advantages are determined by the extent of their integration in the organizational processes and the investor’s willingness to handle the implementation. Pragmatically, the research proposes that SMEs ought to integrate the adoption of technology with training and capacity-building interventions, whereas policymakers ought to establish enabling ecosystems with low-cost digital infrastructure and human capital development. However, the study's weaknesses, including cross-sectional and self-reported data, as well as regional boundaries, limit the extrapolation of the findings. Future studies need to be longitudinal and cross-country to include the institutional and dynamic differences.