한국 알고리즘 트레이딩 규제의 기술혁신 및 안정성의 균형점 모색- Kingdon의 다중흐름모형을 중심으로 -

초록

This study analyzes how the competing objectives of promoting technological innovation and ensuring market stability can be reconciled in the regulation of algorithmic trading in Korea, using Kingdon’s Multiple Streams Framework as the analytical lens. Drawing on qualitative data from written expert group interviews with algorithmic trading practitioners, the study finds that, in the problem stream, repeated system failures during periods of extreme volatility, false positives in regulatory detection, and reduced predictability due to non-disclosure of thresholds are recognized as key issues, while in the politics stream, social demands for investor protection and market stability coexist with structural tensions created by existing legal and institutional frameworks such as personal data protection laws, and in the policy stream, a variety of policy alternatives, including enhanced ex ante controls, clearer regulatory criteria, risk-based differentiated regulation, technical standardization, and AI-based market surveillance, emerge and compete in parallel. AI-based surveillance appears as a promising policy option; however, a conditional acceptance logic is evident in that institutionalization is deemed possible only when explainability and clear allocation of responsibility are secured. Overall, the study suggests that regulatory change in algorithmic trading is more likely to occur not through a single regulatory measure but through a policy package that combines disclosure design, regime-specific calibration of thresholds, ex ante control systems, explainable AI surveillance, and risk-based internal governance.

키워드