Over time, SEBI has issued various circulars to establish a clear regulatory framework for algo trading. Below are the main components:
1. Approval of Algorithmic Strategies
Brokers and trading members must seek approval from exchanges before deploying any algorithm. Each strategy needs to be tested and documented, including its logic, risk parameters, and expected outcomes.
2. Exchange-Approved Infrastructure
Trading members must use infrastructure certified by stock exchanges, including servers, co-location facilities, and connectivity tools. Exchanges monitor and audit the systems regularly for compliance and performance.
3. Risk Controls and Checks
SEBI mandates a series of risk control mechanisms that must be built into algorithmic systems. These include:
Price and quantity limits
Order value checks
Fat-finger error prevention
Order throttling (order-to-trade ratio limits)
Circuit breakers and kill switches
Such controls are designed to minimise potential disruptions caused by faulty algorithms or operator errors.
4. Co-location Regulations
Co-location facilities — where traders place servers close to the exchange for faster execution — must be accessed fairly. SEBI ensures that all participants get equal access, and no unfair latency advantage is granted.
5. Unique Algo IDs and Audit Trails
Each algorithm must carry a unique identifier. Exchanges track these IDs for all orders placed, allowing authorities to audit activities in case of unusual trading patterns or complaints.
6. Order-to-Trade Ratio (OTR)
SEBI regulates the maximum permissible OTR to prevent traders from flooding the market with unexecuted orders. Excessive cancellation of orders is considered disruptive and may attract penalties.
7. Periodic Audits and Reporting
Brokers using algorithmic trading must undergo periodic audits by certified professionals. Detailed logs, strategy documentation, and system architecture must be maintained for review.