Introduction
The transformation of financial markets in the last decade has been further marked by broad technology adoption. The most notable changes manifest in the explosion of algorithmic trading-computer programs and AI-driven strategies that rapidly make trades. Algorithmic trading and artificial intelligence (AI) in stock markets are increasingly onlookers, making stock market trading more efficient but posing many new challenges as well.
As the capital markets regulator of India, SEBI acts as a significant enabler for market stability and the protection of investors. However, the confluence of AI and algorithmic trading under the eyes of regulatory oversight undoubtedly contributes towards a very delicate balance between injecting technological innovation and protecting market integrity. This blog takes you through the regulatory framework on AI and algorithmic trading by SEBI, highlighting both the challenges and opportunities that such an evolving landscape brings about.
Understanding Algorithmic Trading and AI in Financial Markets
Algorithmic trading refers to the trading of securities using complicated mathematical models and computer systems that conduct trades with particular predefined criteria, such as price, timing, or volume. Thousands of orders are processed within seconds or even fractions of a second by analyzing vast amounts of data in real-time to identify profitable trading opportunities and execute them much faster than any human trader. Algorithms make up a large part of the total trade volume in India on both the National Stock Exchange (NSE) and Bombay Stock Exchange (BSE).
AI takes this even one step ahead by infusing machine learning and predictive analytics. Unlike most algorithms, AI can not only look at historical data to identify patterns but can also predict market movements before they happen. This makes trading strategies dynamic and adaptive. It has led to the rise of high-frequency trading. Traditionally, trades are placed within milliseconds of a prediction from an AI system.
While AI and algorithmic trading thus benefit from increased efficiency, liquidity, and lower transaction costs, they also introduce other potential risks of market manipulation, the spiking of volatility, and a breakdown in the gap between retail and institutional investors.
Do These Advances Need Regulation?
The adoption of AI and algorithmic trading brings a new set of risks that traditional regulation may be unable to cope with. Some concerns include:
- Algorithms can be created so that they create imbalances in demand and supply or manipulate stock prices, thereby turning the market prices distorted and bogus and inflicting severe loss upon the retail investors.
- Flash Crashes Algorithm-driven high-frequency trading causes sharp and extreme price moves that would immediately cause flash crashes. For example, the US Flash Crash in 2010 caused the Dow Jones Industrial Average to decline by nearly 1,000 points within minutes due to an algorithmic trading feedback loop.
- Private Advantage: Algorithmic and high-frequency traders have advanced tools and technologies that make them sustain higher speed than the general retail investor, which is raising questions about the fairness of the game.
- Lack of Transparency: Most algorithmic strategies, even most AI-based strategies, are opaque and consequently not clear enough to be understood by the regulators and sometimes not fully clear to their firms. This makes the regulation of the operations of such firms impossible.
Considering all the risks, SEBI has undertaken several steps to make the regulatory architecture of AI and algorithmic trading robust.
SEBI’s Regulatory Framework
The stand of SEBI on the regulation of algorithmic trading is developing in a manner wherein markets can innovate and, at the same time, restrain themselves from malpractices that might hurt market integrity. The three prime features of the SEBI’s approach are:
- Mandatory Registration of Algorithmic Traders
SEBI makes it mandatorily requisite that all algorithmic trade brokers should register with SEBI and seek prior approval for their trading systems. This limits only the regulated trade to high-frequency and algorithmic trading and prevents unregulated activities that can be seriously harmful.
Further, SEBI has specified the rules for deploying algos by the brokers. The same expects that the algorithms must have adequate risk management controls and must pass through certain checks in a controlled environment before being released into the market.
- Risk Management Systems
Algorithmic trading has been mandated with the implementation of strict risk management systems by SEBI to avoid any unwanted consequences. These systems include kill switches that automatically halt the trading if the algorithms begin to behave erratically, safeguards from algorithms touching pre-defined limits, and mechanisms to help minimize liquidity risks.
Therefore, risk management systems are critical to managing the volatility that can emerge from the large volume and speed of algorithmic trades. SEBI aims to prevent malfunctioning algorithms from causing possible market disruptions by making sure the brokers have robust controls for risks.
- Latency and Co-location Restrictions
As SEBI implemented these regulations, this kind of latency arbitrage, wherein the high-frequency traders can gain an unfair advantage over others just because they are closer to the servers of the stock exchanges, is worked on reducing. In other words, until 2018, SEBI did not allow co-location facilities unless there was fair access from the exchanges.
The main idea is to make the playing field level for all traders, with no particular group sporting an unfair technological advantage over the others. This also aligns well with the greater overall SEBI objective of promoting both market fairness and transparency.
- Algo Flagging and Auditing
To enhance transparency, SEBI mandates every order arising out of an algorithm to be specifically so indicated as an algorithmic trade. This will enable regulators to track the level of algorithmic trade and identify any consistent pattern of manipulation or abuse of any kind.
Other than flagging, SEBI has also made mandatory audit trails of algorithmic trading systems. The brokerage firm must maintain log records of all algorithmic transactions involving parameters of the various algorithms used. In this regard, it is ensured that trading activities may be checked and audited as and when required.
- Limits on High-Frequency Trading
Such are the risks associated with the practice of HFT that SEBI has put several restrictions on certain HFT practices. In this, especially, one of the key measures is order-to-trade ratio regulation, which requires a broker to ensure that he limits the number of orders that he generates relative to the trades that he executes. That way, strategies that feed off orders for finding prices should not be able to overwhelm the market by the sheer volume of orders.
Secondly, SEBI has also introduced a minimum resting period for orders whereby, for example, an order stays in the open order book of an exchange for a certain minimum period before it can be canceled. This again discourages the rapid-fire entry and deletion of orders, so common in manipulative trading strategies.
- Regulatory Framework for AI in Trading
The regulation for AI-specific trading systems is still in its infancy. However, the regulator has started looking into how AI affects market integrity. SEBI has plans to help put guidelines around how an AI model should be developed, tested, and deployed with transparency and accountability.
And because AI systems are completely capable of autonomous learning and adaptation, SEBI thinks that special steps must be taken to prevent fraudulent market practices. This could be in the form of time-to-time monitoring of every trade executed by AI, audit of AI models from time to time, and more stringent penal provisions against AI systems engaged in manipulating the market.
- Sandbox Framework for Fintech Innovations
In 2020, SEBI began a Regulatory Sandbox where fintech companies were permitted to try out innovative products under controlled conditions, though still under relaxed oversight. The sandbox creates a safe space for firms to experiment while keeping in check the risks that could be posed to the wider market at large.
The sandbox attempts to get innovation working for both the technology frontier and market stability through SEBI.
Challenges Before SEBI
While the regulatory framework of SEBI sounds robust, it faces several challenges in the effective regulation of algorithmic trading and AI:
- Technological Complexity: Algorithms are getting too complex and sophisticated as driven by AI which becomes difficult for the regulators to understand what it does and how it may behave under different circumstances.
- International Coordination: Algorithmic and AI trading is increasingly borderless activity, even when the market participants and traders are situated in different jurisdictions, so coordination of such regulators is required. In the coordination process, the challenge is that it remains a huge problem to ensure the fact that Indian markets are not adversely affected by global strategies based on AI.
- Regulating Rapidly Evolving Technologies: Since the change is very fast in technology, constant updating of regulations becomes essential whenever one goes out of date. SEBI needs a balance of flexibility enough to fuel innovation but strong enough to safeguard the integrity of the market.
- Protection of the Retail Investors: Retail investors are always on the receiving end in a market and a time when algorithmic trading dominates. Protection of the retail investor’s interest remains a challenge before SEBI, but one that is continually faced since the market is increasingly shaped by algorithms and AI.
Global Regulatory Approaches and Role of SEBI
SEBI is under no illusion that, when it comes to AI and algorithmic trading regulation, it works in isolation. After all, on the world front, respective regulators such as the U.S. Securities and Exchange Commission (SEC) and the European Securities and Markets Authority (ESMA) – are fighting with regulators at respective levels. A comparative analysis of the framework adopted by SEBI with these global counterparts reveals India is not merely being proactive but forward-thinking with measures such as the Regulatory Sandbox.
However, it is an area of evolution as AI-driven trading is going to be one of the next phases, and hence SEBI will have no other alternative but to collaborate with international regulators more closely for rules harmonization and ensure that Indian markets are not exposed to undue risk from cross-border trading activities.
Way Forward:
This benchmark for AI and algorithmic trading from SEBI testifies that continuous innovation is still accompanied by the market integrity intended balance. Algorithmic and AI-driven trading have significant benefits but pose some risks that need strategic control. The primary function of SEBI is based on transparency, risk management, and fair market practices that ensure stable and secure Indian markets by constantly changing this financial landscape with advanced technology.
As AI and algorithmic trading continue to evolve, SEBI needs to be agile, update its regulatory framework to take in new challenges and build an environment where innovation can happen in such a way that it doesn’t compromise market fairness. In doing so, SEBI will ensure that the Indian financial markets remain at the apogee of technological advancement and regulatory prudence.
***
Author Srija Singh is a 4th-year student of Amity Law School, Noida
Author Bio
