← AITrading.in · Blog

What Is Algo Trading? A Plain Guide for India

Algo trading is trading in which orders are generated by automated rules instead of being typed in by hand. SEBI's own wording, in its circular of 30 March 2012, is short: "Any order that is generated using automated execution logic shall be known as algorithmic trading." In India, retail investors can use it only through a stock broker, within SEBI and stock exchange rules.

Risk note, up front: algo trading automates decisions. It does not make them better. A badly designed rule loses money faster when a computer runs it. This article is education, not investment advice, and it names no strategy, signal or expected return.

What does "algo" actually mean?

An algorithm is a fixed set of steps. In trading, those steps say when to place an order, for which instrument, at what price and in what quantity. Once the rules are written down, software can check them against market data and send the orders without a person clicking each time.

A few terms used throughout this guide:

How does algo trading work, step by step?

  1. Write the rules. Entry, exit, position size and the maximum loss you will accept, in plain words first.
  2. Turn them into code or a configured tool. This is where a developer, a coding course or a no-code product comes in.
  3. Test on past data. See how to backtest a trading algorithm without fooling yourself.
  4. Run it on paper, meaning live prices with no real money. Paper trading vs live trading explains what this stage can and cannot show.
  5. Connect to a broker's API. Your broker decides what access you get and under what conditions.
  6. Go live with small limits and monitoring. Cap order size and daily loss, log every order, and keep a way to stop the program quickly.

What are the main types of algo?

Type What it does Main thing to watch
Rule-based entry and exit Places orders when conditions you defined are met Rules that fit past data too closely
Order slicing and timing Splits or times a given order to place it in a planned way It carries out a decision; it does not make one
Black box The user cannot see the logic You cannot judge what you cannot see

SEBI's February 2025 circular uses a different split. It sorts algos into "Execution Algos or White box Algos" where the logic is "disclosed and replicable", and black box algos where "the logic is not known to the user and is not replicable". For black box algos, it says the algo provider must "register as a Research Analyst and maintain a detailed research report for each such algo."

Is algo trading legal in India for individuals?

Yes, through a broker. SEBI's circular of 4 February 2025, "Safer participation of retail investors in Algorithmic trading", says the aim is to let retail investors use algo facilities "with requisite safeguards". Key points from it:

The exact thresholds and dates have been set by the exchanges and have moved more than once. Our longer guide, is algo trading legal in India, walks through the NSE standards and the timeline, with the circular numbers.

What controls does the broker apply to algo orders?

SEBI's 2012 circular lists minimum risk checks that a broker must have before releasing algo orders. They include a price check against the exchange's price bands, a quantity check, an order value check, and an automated execution check. That last one says an algo "shall account for all executed, un-executed and unconfirmed orders" before releasing more, and the algo system should have pre-defined parameters for automatic stoppage if execution is "leading to a loop or a runaway situation".

Your own program sits behind these checks, so your code must handle a rejected order safely. For one broker's practical limits, see Kite Connect API for algo trading: what to check first.

What are the risks?

A short checklist before you try it

  1. Can you state your rules in two sentences, with the maximum loss you accept?
  2. Has your broker confirmed in writing that you can use its API, and under what conditions?
  3. Did you test on past data, then on paper, before using real money?
  4. Do you have an order size cap, a daily loss cap and a kill switch you can reach?
  5. If a vendor is involved, is it empanelled with the exchanges, and does it avoid promising returns?

Frequently asked questions

Is algo trading the same as high-frequency trading? No. Algo trading covers any order generated by automated logic, including a program that sends one order a day. High-frequency trading is a narrow, speed-focused subset. SEBI's definition does not mention speed.

Do I need to know coding? Not always. Some people write their own code, some use a broker or vendor tool, and some work with a developer. Whichever route you pick, you are still responsible for understanding the rules your program follows.

Can algo trading guarantee profits? No. Nothing in the SEBI circulars cited here suggests it can, and SEBI has acted against platforms advertising returns. Treat any guarantee as a warning.

Does a backtest tell me what will happen next? No. It shows how the rules behaved on past data. Markets change, and costs and fills in live trading differ from a simulation.

Where do I start? Start with the broker. Ask whether it offers API access to retail clients and what it requires. Then read the sources below.

Where a custom build fits

Some traders want software built to their own written rules, with backtesting and options analytics, connected to their broker's API. The home page describes the kinds of systems built, and the FAQ answers common questions. A developer can make your rules explicit and run them consistently. A developer cannot make a strategy profitable.

Limits of this article

It was written from the SEBI circulars listed below. It does not cover later circulars, exchange standards in detail, taxation, or your broker's terms. Read the current rules from SEBI, the exchanges and your broker before you trade.

Sources

This article is general information, not legal or financial advice. Trading involves risk of loss.