SEBI Regulations on Algorithmic Trading in India

Overview and Old System

Algorithmic trading, or algo trading, involves executing trades using automated pre-programmed instructions. The rules or the instructions are based on many variables such as price, volume, timing, or other market conditions. The goal is to enable rule-based trading at high speed and accuracy without manual intervention. 

Historically, algorithmic trading was dominated by institutional players since they have access to Direct Market Access (DMA) and co-location facilities. This allowed them ultra-low latency and superior order execution. On the other hand, retail traders were using broker APIs and third-party platforms for automation. However, no regulatory scrutiny or oversight on trading using these APIs existed.

There was no uniform framework for regulating retail algorithms or platforms offering strategy-based trading via APIs. This led to concerns around market safety, misuse, and lack of transparency, especially as retail algorithm usage grew rapidly.

New SEBI Regulations 

In order to address the increasing use of irregular retail trades and protect the market's integrity, SEBI has proposed a comprehensive regulatory structure that covers algorithm trading. These new criteria aim to bring more openness, responsibility and investor security by placing responsibility on stock exchanges, brokers and algorithm developers.

1. Mandatory Exchange Approval for All Algorithmic Strategies

• The broker has to ensure that all the algorithmic strategies are approved by the stock exchange before deployment.
• No algorithmic strategy can be deployed in live markets without this approval.
• It ensures that only tested and certified strategies work in the market, which reduces the chances of incorrect orders or manipulation.

2. Unique Algo ID Tagging

• Each algorithm order should be mapped with a unique identifier (Algo ID), which connects it to the specific approved strategy.
• This allows exchanges to track separate algorithms, monitor performance and identify malicious or deficient strategies in real time.
• Algo IDs are assigned by the exchange after successful testing and approval of the strategy.

3. Classification: White Box vs Black Box Algos

• Algos will be categorized into:
• White Box Algos/Execution Algos: Here, the logic is transparent, and users can replicate it. 
• Black Box Algos: Here, the logic is not disclosed. For black-box algorithms, the providers must register as research analysts and maintain detailed reports on the algorithm's logic and functionality.
• White Box will be easy to approve and monitor, while Black Box Algos may require strict examination without full disclosure.

4. Registration of Algorithm Provider

• All algorithm providers must necessarily be exchanged and empanelled before the brokers can onboard them.
• This has been added so that unverified algorithm providers and companies do not offer wrong algorithmic trading solutions to retail investors.

5. Deployment Only via Broker Infrastructure

• Sebi has banned all open APIs. Open APIs are used to enable traders and third-party applications to connect directly to a broker's trading platform, execute trades, and access market data. Now, access will be allowed only through unique setups, ensuring proper identification of algorithmic trades.
• Only algorithms hosted and deployed through broker-owned infrastructure will be permitted.
• Third-party servers, cloud services, or uncontrolled environments are prohibited unless integrated with the broker system.
• This ensures end-to-end control and accountability.
• The broker's infra must log all strategy execution, have pre-trade risk checks, and be auditable by the exchange and SEBI.

6. Broker Responsibility and Oversight

• The brokers have been given the responsibility to:
• Approve and register client strategies
• Ensure only exchange-approved algorithms are used
• Track and log API usage and activity per client
• Maintain complete order audit trails
• Brokers should also address any issues or complaints related to the algorithm.
• Brokers must also implement the following:
• Pre-deployment testing
• Real-time monitoring tools
• Red-flag alert systems
• Access control to ensure clients cannot bypass compliance layers

7. Mandatory Risk Mitigation Controls

All brokers and exchanges must enforce real-time risk controls, including:

• Order throttle limits: Limits the number of orders per second
• Kill switch: Ability to instantly stop a malfunctioning algo. SEBI mandates that brokers and platforms use a "kill switch" - an emergency stop mechanism that can immediately disable a malfunction to avoid disrupting the main market.
• 2FA: There should be stronger authentication mechanisms to prevent unauthorized access to API-based strategies. With the new mandates, SEBI ensures that all algorithmic trading platforms and users should be authenticated using Two-Factor Authentication (2FA) and OAuth.
• Static Whitelisted IP Requirement for API Access: SEBI has specified that API access should only be allowed through static, whitelisted addresses to prevent unauthorized or uncontrolled access. This helps to reduce the risk of trade bots or malicious users from unknown sources. 

These measures help prevent flash crashes, order flooding, or other disruptive behaviour caused by algorithms.

8. Self Developed Algorithms 

• If any retail trader has developed his own algorithm and wants to trade on personal or immediate family members' accounts, he can do so easily. This limits the unauthorized or resale of individual strategies.
• This is mandatory only if the trading activity exceeds the specified order-per-second threshold. Any system that exceeds a certain number of others per second is automatically classified as algorithm trade and is subject to strict examination.

9. Regulations for Black Box Algorithms

• Algorithm providers using black box algorithms must be registered as Research Analysts with SEBI. 
• These algorithm providers are required to maintain detailed research reports and undergo re-registration in case of significant changes in their black box algorithm logic. Hence, if the already approved algorithm is modified, it must undergo a new approval process from the exchange. This prevents abuse of back door changes or pre-approval.

10. Disclosures to Clients

• Brokers are required to disclose the APIs and third-party integrations that they offer clearly.
• Brokers would be encouraged to inform clients about the risks of algorithmic trading. This will include information on latency-related issues and the market impact of algorithmic trading.
• Both brokers and algorithm providers must disclose all charges, such as subscription fees for algorithmic trading strategies and brokerage costs associated with executing algorithmic trades.

Implementation Dates

The current schedule for the rollout of the new regulatory framework by SEBI is as follows:

• By April 1, 2025: The Brokers' Industry Standards Forum will develop and finalize the detailed implementation guidelines.
• By August 1, 2025: The new regulatory framework will come into full effect across the industry.

Why These Changes Are Introduced

There has been a growing concern about unregulated retail algorithms and the risks they pose to market integrity and investor protection. The new regulations are primarily due to the following: 

• Rise of Retail Automation: There has been a massive increase in retail trader participation in algorithmic trading, and investors are now using APIs and third-party platforms to trade automatically without proper regulation.
• Lack of Oversight: Many brokers offer their APIs to retail traders and are unaware or unable to monitor how clients deploy APIs or third-party algorithms.
• Potential for Market Manipulation: Unchecked or wrongly coded algorithms could submit erroneous, spam, or even manipulative orders, creating systemic risk.
• Fairness and Level Playing Field: Institutional players operated under strict norms, while retail algorithms ran freely without approval or tagging.
• Investor Protection: Retail traders often lack understanding of how algorithms function and the risks associated with algorithmic trading. A formal regulation can help create more transparency to protect the retailer.

Impact of Changes

The proposed SEBI rules are expected to bring structural reforms, but new compliance for all stakeholders is also introduced.

• For Retail Traders

• They must have increased safety and more transparency in trade practice for algorithms.
• This will also improve investor knowledge through compulsory education and awareness initiatives.
• The stronger regulatory oversight to detect and deter fraud and market manipulation will help retail traders.

• For Brokers

• They will be encouraged to establish compliance systems to approve, track, and audit client algorithms.
• They will act as the link to strengthen adherence to SEBI's risk management guidelines.
• They will be responsible for securing compliance approval for each algorithm given to customers.
• To ensure compliance, they must also establish strict monitoring and reporting systems.

• For Algo Platform Providers

• They can no longer provide strategy without formal approval from brokers and exchanges.
• They must obtain registration and approval from the stock exchanges before deployment.
• They should verify that all algorithms adhere to established risk and compliance criteria.
• They must maintain comprehensive logs and documentation to promote transparency and accountability.
• They must build integrated, compliant solutions or partner with brokers directly.

• For the Market Ecosystem

• There will be higher transparency, which will improve due to tagging, monitoring, and audit trails.
• The potential reduction in system overloads and abnormal orders.
• Overall, the new rules will help create a safe environment for both retail and institutional participation.

Conclusion

The new SEBI rules for algorithmic trading represent a significant change in how algorithm strategies can gain widespread access and popularity, especially for retail investors. SEBI aims to install a transparent, strong and fair trading ecosystem by using strategy approval, unique identifiers, API control and surveillance mechanisms.

These changes may indeed, in the beginning, look restrictive. However, they have the right intention and aim to create long-term faith in the algorithmic trading domain, reduce abuse and protect retail investors. As algorithmic trading increases in popularity, this framework ensures that innovation occurs responsibly and safely.