Nithin Kamath Decodes Sebi’s New Algo Trading Rules For Retail Traders

Zerodha CEO Nithin Kamath took to X (formerly Twitter) to explain the Securities and Exchange Board of India’s (Sebi’s) latest algorithmic trading regulations, marking a significant shift in market access. Algo trading, which was previously only available to institutional participants, is now open to ordinary investors due to the new structure.

Kamath highlighted that this rule assures that regular traders can participate in algorithmic trading under specified conditions. He mentioned in his tweet, “If you’re a retail trader using broker APIs to automate trades, you can do so as long as your order frequency is below an exchange-prescribed threshold. This limit is yet to be decided.”

He also pointed out that traders who sell algorithms or trading strategies will now need to partner with brokers and register with the exchanges. To ensure more accountability and transparency, platforms that charge for algorithm sharing are required to get both an exchange registration and a research analyst (RA) license.

Kamath added, “Marketplaces where people share algos for a fee can’t publish strategies without exchange registration as well as an RA license.” Sebi is expected to release further operational guidelines by April 1, 2025, to support this transition.

Kamath also shared a link in the post’s comments for further insights into Sebi’s regulations.

The Sebi framework, which was issued earlier this week, will take into effect on August 1, 2025, with the goal of making algorithmic trading more accessible to regular investors while ensuring adequate risk management and market integrity. The new regulations require brokers to get exchange clearance for any algorithm they employ, and each automated order will require a tracking ID. Any changes to approved algorithms will require regulatory clearance. Also, retail traders who create their own algorithms must register them if their order frequency exceeds the specified limit.

In addition, Sebi has classified algorithms into two types: “white box” algos, which are transparent and replicable, and “black box” algos, which operate without transparency. Those offering black box strategies will be required to obtain an RA license and provide comprehensive research documentation.