[Rishabh Sharma is a IV Year student at NALSAR University of Law, Hyderabad]
The Algorithmic Trading Market: Global Industry Analysis, Trends, Market Size, and Forecasts up to 2025 Report forecasts the global algorithmic market in trading to expand at a compound annual growth rate of 11.8 percent between the period from 2019 to 2025. In the backdrop of such a rapidly booming global trading landscape, the use of Artificial Intelligence (AI) and Machine Learning (ML) has been soaring and is projected to rise further in the near future, and consequently the issues concerning their governance, application, regulation, and economic impact have become matters of central policy concern in many countries. The use of AI and ML as product offerings in the Indian financial market has also seen an exponential increase in the past few years. Growing access to machine learning tools combined with a greater computing competence has been an impetus for perpetuation of such a trend.
Presently, India has no policy framework in place for administration of AI and ML technologies in the financial market. The Securities & Exchange Board of India (SEBI), through its Circular on Reporting for Artificial Intelligence (AI) and Machine Learning (ML) applications and systems offered and used by Mutual Funds dated 9 May 2019, took first steps towards formulation of policies in future for governing their usage. While currently it can only be said that SEBI is ensuring preparedness for any future AI and ML policies that might arise in India, it can be reasonably surmised that policy framework for their use is in the pipeline, and not a distant dream. By way of the Circular, SEBI introduced mandatory reporting requirements, under the shroud of conducting a survey, for Market Intermediaries (MIs), Market Infrastructure Institutions (MIIs), and Mutual Funds (MFs) which use AI and ML based applications or systems, wherein they are required to make detailed quarterly submissions of such usage to Association of Mutual Funds in India (AMFI), beginning from quarter ending June 2019.
The definition of technologies that are comprehended by SEBI is fairly expansive, as all those applications or systems which are provided to the investors or employed internally for facilitation of investing and trading or any other purpose, dissemination of investment strategies and advice, or carrying out compliance/operations/activities, are included within the scope of the Circular. Even Fin-tech and Reg-tech initiatives by market participants which involve AI and ML have been covered by the Circular. Annexure B of the Circular also provides a comprehensive list of systems which will be categorized as AI and ML technologies.
All registered MIs, MIIs, and MFs which offer or use applications or systems listed in Annexure B are required to partake in the reporting process after completion of the AI or ML reporting form. Submissions are required to be made by these entities on a quarterly basis within 15 days from the expiry of the quarter to AMFI. The information reported by these entities on AI and ML applications and systems shall be consolidated by AMFI and submitted to SEBI within 30 days from the expiry of the concerned quarter. During the entire process, AMFI is required ensure that confidentiality is maintained regarding the information received from the entities.
While the Circular is a step in the right direction for formulation of a policy framework to govern the usage of AI and ML technology in the financial market, there are a few critical considerations that need to be evaluated before modeling any policy blueprint on the matter. It needs to be recognized that taking care of these considerations will go a long way in sparking a discourse among the stakeholders and catalyzing efforts to expand the horizons of AI and ML research in India. However, it also needs to be realized that showing any disregard to such considerations can lead to certain unintended consequences.
To begin with, attention needs to be drawn to SEBI’s implicit warning in the Circular to the intermediaries using AI and ML technology against any misrepresentation of financial benefits, primarily accentuating those arising out of black box systems, because quantification of their behaviour is a daunting task and an almost impossible one in some cases. It is an acknowledged fact that the inner functioning of these self-learning machines adds to the layers of complexity and opaqueness in terms of decoding the machine behavior. Once an ML algorithm is trained, understanding the system’s reasons for giving a particular response to a data set can be very difficult. In the author’s view, the Circular appears to be a prelude to framing of regulatory norms for ensuring greater transparency in the applications or systems using such complex AI and ML.
However, before ushering in such regulatory norms, SEBI needs to be cautious that coming up with any measures having an overarching policy objective of increasing transparency in use of AI and ML might prove to be a functional prohibition on the development of certain classes of technology which inherently lack transparency. More layers of dimensions and complexity are expected to add on to neural networks since the ML algorithms are becoming increasingly sophisticated with time and are developing improved capacity for balancing and optimizing tremendous amounts of data at once. To put it simply, there is not enough rationale supporting imposition of regulatory fiats on the financial institutions in order to achieve transparency in the usage of AI and ML technology. On the contrary, such fiats might pressure AI designers to employ shallower or less complex architecture for compliance of regulations, even if it results in poorer AI and ML performance. Such a ramification will be unwelcome and gravely detrimental to the future prosperity of AI and ML technology in the Indian financial market.
Further, SEBI also needs to keep in mind that in an emerging sector such as AI and ML technology in the financial market, the introduction of a regulatory framework aimed at promoting transparency is capable of bringing startups in the AI and ML technology to a halt. In the presence of unfavorable regulatory fiats, potential market entrants might lack the necessary catalyst to enter the Indian financial scenario. The requirement for complying with these regulations might entail bearing high costs for the potential entrants, in addition to constraining their freedom of developing new AI and ML designs. As on date, only a few large firms are in possession of the existing AI and ML talent in India. If a byzantine regulatory system is imposed on all the financial institutions, there are high prospects of these large firms continuing to maintain their monopoly over the Indian financial market in terms of using AI and ML technology, with potential market entrants being forced to remain on the sidelines as mere spectators, for lack of required means and motive to meet the compliance standards.
Accordingly, it can be said that SEBI’s future policy decisions pertaining to the ethical standards required to be met in using AI and ML technology will play an unfathomable role in dictating the fate of AI and ML in the Indian financial markets. In all likelihood, the framework of these future policies cannot be skewed to the distaste of future developers and market entrants of AI and ML applications or systems, lest India risks being deprived of prospering in an industry which subsumes boundless opportunities. To conclude, the author is of the opinion that SEBI’s present objective to “gain an in-depth understanding of the adoption of such technologies in the markets and to ensure preparedness for any AI/ML policies that may arise in the future” is prima facie a constructive one. However, it is crucial that any policy that is contemplated in the future based on such an understanding should not be moulded in a manner which, by any means, disincentivizes innovation and development in a fledgling field like AI and ML technology. Although regulatory mechanisms will surely come with their benefits, they will also entail their costs. Therefore, SEBI requires to be circumspect while introducing regulatory norms for financial institutions in a domain like AI and ML technology. None of this is to say that there is no need of regulating the use of AI and ML technology in the Indian financial market. But in shaping the policy framework, SEBI has the onus to formulate the most appropriate strategy for governance of AI and ML landscape in India, without impairing the incentives available for the potential future developers & market entrants of AI and ML technology.
– Rishabh Sharma
