SEBI Algo Trading Rules

• All algorithmic strategies must receive exchange approval before live deployment. 

• Algo Providers (especially those marketing Black Box algos) are strictly prohibited from promising guaranteed or assured returns. 

• Black box algorithms are available only on the SEBI Research Analyst registration. 

• Brokers are mandated to implement kill switches, order throttling, and maintain comprehensive audit trails for all algo trades. 

1. Low-Frequency Trading (LFT) 

• Strategies are on longer time frames with relatively smaller orders. 

• Common strategies include trend-following, arbitrage, and mean reversion models. 

• These strategies do not depend on speed as much as they do on price patterns, volume signals, or statistical relationships. 

• LFT is very popular amongst retail traders because of reduced infrastructure and latency requirements. 

2. High-Frequency Trading (HFT) 

• Strategies are used to execute a huge number of orders at very high speeds. 

• Includes market making, statistical arbitrage, and latency trading. 

• Requires advanced technology, collocation facilities, and important risk controls. 

• HFT activity is under close watch because of the potential effect on stability in the markets. 

1. Mandatory Exchange Approval for All Algorithmic Strategies  

• Brokers have to make sure that each of their algorithmic approaches is approved by the stock exchange before launching them. 

• No algorithm is allowed to run in live markets without the formal approval of the exchange. 

• This ensures that only strategies tested and certified are deployed, which has the effect of reducing errors and risks of manipulation. 

2. Unique Algo ID Tagging 

• Every order of algorithms will need to be associated with a unique Algo ID. 

• This can be used to allow the exchanges to account for algorithms, monitor behaviour, and identify issues in real-time. 

• Algo IDs are given after successful testing and approval of the strategy. 

3. Classification: White Box vs Black Box Algos 

• White Box Algos: Strategy logic is available and replicable by the users. 

• Black Box Algos: Strategy logic is not published and is subjected to more intensive investigation. Black box providers need to register as Research Analysts and keep detailed documentation. 

4. Registration of Algorithm Providers

• All algorithm providers will have to be empanelled with exchanges before brokers are able to onboard them. 

• This saves the market from unverified entities from offering algorithmic trading solutions to retail organisations. 

5. Deployment Only via Broker Infrastructure

• Unrestricted open API to controlled third-party access has been restricted. 

• Algorithms have to be hosted and delivered through broker-controlled infrastructure. 

• Broker systems need to support logging, pre-trade risk checks, and audit readiness. 

6. Broker Responsibility and Oversight 

• Brokers need to approve and register all client strategies. 

• Only exchange-approved algorithms are allowed to run. 

• Brokers need to monitor API usage, keep logs, and retain audit trails. 

• Algorithm-related complaints and issues have to be dealt with by brokers. 

7. Mandatory Risk Mitigation Controls brokers 

• Order throttling limit to control the excessive flow of orders. 

• Kill switch mechanisms to halt malfunctioning algorithms immediately 

• Enforcing mandatory 2FA and OAuth-based authentication for API strategies. 

• Static, whitelisted IP addresses to be used for controlled access to the API. 

8. Self-Developed Algorithms brokers 

• Retail traders may use self-developed algorithms for personal or immediate family accounts, and the NSE permits sharing a static IP among these family members with prior permission from the broker. 

• Systems that exceed a threshold of 10 orders per second (OPS) per market segment are classified as algorithmic trades and require mandatory registration.  

• Orders under 10 OPS still require a "Generic Algo ID" provided by the exchange. They are not "exempt" from being classified as algos. They are just exempt from the rigorous individual approval process that high-frequency strategies require. 

9. Regulations for Black Box Algorithms 

• Black box algorithm providers need to register themselves with SEBI as a Research Analyst. 

• Any material change in the logic of the algorithm requires fresh exchange approval. 

• This offers a way of preventing misuse through unauthorised or undocumented modifications from being carried out. 

10. Disclosures to Clients 

• Brokers must explicitly communicate details regarding APIs and third-party integration provided. 

• Clients have to be informed about risks, latency issues, and execution limitations. 

• All the expenses associated with algorithmic trading should be communicated transparently. 

• January 3, 2026: Deadline for brokers to participate in mandatory mock trading sessions to test their control systems. 

• January 5, 2026: Non-compliant brokers (those who failed mock testing or registration milestones) are barred from onboarding new retail clients for API-based algo trading. 

• April 1, 2026: The entire SEBI algorithmic trading framework becomes fully and legally binding for all stock brokers across India. 

1. Rise of Retail Automation: Explosive growth in API-based and third-party trading enhancements. 

2. Lack of Oversight: Lack of vigilance in monitoring the deployment of retail algorithms. 

3. Market Manipulation Risks: Faulty algorithms producing abnormal/overweight orders. 

4. Fairness and Level Playing Field: Institutional participants were already operating with strict norms. 

5. Investor Protection: Retail traders, for the most part, were not informed of algorithmic risk. 

1. For Retail Traders: 

• Increased safety and transparency in algorithmic trading 

• Enhanced investor awareness through disclosure and regulatory oversight 

2. For Brokers: 

• Increased responsibility for approvals, monitoring, and reporting. 

• Increased compliance, audit, and risk management requirements. 

3. For Algo Platform Providers: 

• Mandatory registration and exchange approvals. 

• Requirement to build compliant, broker-integrated solutions.

4. For the Market Ecosystem: 

• Increased transparency by means of tagging, monitoring, and audits. 

• Decreased system stress and abnormal trading behaviour. 

1.Algo Providers 

• Must register with recognised stock exchanges prior to offering algorithmic strategies. 

• Black box algorithms must register as Research Analysts and disclose on a periodic basis. 

• Strategy changes require new approval, which will increase the compliance and documentation requirements. 

2.Brokers 

•  Required to approve, monitor, and audit all algorithmic trading activity. 

• Responsible for the compliance of SEBI algo trading rules, risk controls, and client disclosures.