Regulatory Effectiveness in Algorithmic and High-Frequency Trading: A Critical Assessment

Objective - Algorithmic and high-frequency trading have become a permanent fixture of modern financial markets over the past two decades — and regulators on both sides of the Atlantic have been trying to keep up, with mixed results. This paper asks how well current regulatory frameworks, specifically the EU's MiFID II and the SEC's supervisory approach, hold up against the challenges that autonomous trading algorithms create. These are not purely technical issues — market manipulation, flash crashes, and the technological gap between different market participants all have real consequences for how the financial system functions. The paper works through three questions: what empirical research tells us about HFT's effects on liquidity, price formation, and market stability; how MiFID II and the SEC each try to get these risks under control; and where exactly current regulation breaks down — whether in the rules themselves or in how they're enforced. Methodology - The paper draws on a structured literature review. This covered peer- reviewed empirical studies in market microstructure, reports from regulatory and international bodies (ESMA, SEC, BIS, IOSCO, FSB, FCA), and legal-policy analyses of the EU AI Act. Sources were pulled from five academic databases — IEEE Xplore, SpringerLink, ScienceDirect, SSRN, and Google Scholar — and each was assessed against four criteria: breadth of regulatory coverage, real-world enforceability, how well regulation adapts to technological change, and documented effects on market stability and integrity. Findings - The central finding is that regulatory effectiveness has less to do with whether formal rules exist and more to do with whether those rules can actually be applied in real time. The EU's preventive ex-ante model and the U.S. ex-post traceability approach both have blind spots — and technologically sophisticated trading firms know how to use them. The paper also argues that the ethical problems associated with algorithmic trading — manipulation, the information gap between institutional and retail participants, opaque AI models — are not really independent moral concerns. They are more like the visible surface of a deeper regulatory and technological failure. Part of the analysis maps how the EU AI Act interacts with MiFID II, tracing how requirements around high-risk system classification, transparency, human oversight, and post-deployment monitoring connect with existing RTS 6 obligations. From this, a four-pillar evaluation framework is proposed — fairness, accountability, transparency, and robustness — as a working tool for assessing algorithmic trading systems on ethical grounds. Practical relevance - The paper's conclusions have concrete implications for regulators, compliance teams, and trading firms alike. Supervisory authorities need to invest in real- time monitoring infrastructure, push for greater cross-border enforcement consistency, and bring AI Act conformity assessment into MiFID II compliance processes. Trading firms, meanwhile, should stop treating algorithmic impact assessments, independent model validation, and explainable AI as optional extras — these should be standard practice. The proposed framework is intended as a starting point for ethics audits in automated trading, and to help narrow the gap between what regulation promises on paper and what actually happens in the market.