고빈도 알고리즘에 의한 시세조종성 행위 규제의 현행법상 한계와 개선방안*- 고도의 인공지능(AI)을 이용한 시세조종에의 함의를 포함하여 -

As seen in the recent case of sanctions against Citadel Securities, this article starts from the need to prepare appropriate legislation for the act of damaging the market price determination structure of the capital market with high-frequency algorithmic trading. This is because the current Korean Capital Markets Act is silent on algorithmic trading regulation, and although the Korea Exchange regulates algorithmic trading with its own regulations, there are significant restrictions. Therefore, this article reviews recommended revisions to the current Korean Capital Markets Act in order to effectively regulate and sanction algorithmic trading, especially high-frequency algorithmic trading, which can have a great impact on the market order, mainly by referring to EU laws. In addition, the standards for calculating criminal fines and/or administrative penalties for violations of Articles 176, 178, and 178-2 of the Capital Markets Act, which prohibit market price manipulation, etc., need to be greatly revised. Since these provisions prohibiting market manipulation are intended to ensure fairness in stock trading and smooth distribution, the amount of profits obtained by violators should not be a key factor in determining the statutory penalties. In addition, as algorithmic trading by autonomous AI using machine learning becomes common, proper regulation will be difficult with conventional market price manipulation regulations that depend on intention, purpose, and causation. However, it is not clear whether the bills pending in the Korean National Assembly to regulate AI will be helpful in protecting users in terms of regulating market manipulation by AI involved algorithmic trading. In this case, further discussions on liability attribution and sharing of damages should be conducted and appropriately reflected in the AI regulatory legislation.