Between 2024 and 2026, something important changed in India’s financial markets. What was once a niche activity reserved for tech-savvy enthusiasts and institutions has become a regulated and widely adopted retail practice.
Today, thousands of individual traders in India are exploring algorithmic trading, attracted by the promise of automation and efficiency. But the transition from manual trading to systematic execution is not just about writing code. It involves navigating regulations, understanding exchange mechanics, accounting for real-world costs, and building reliable infrastructure. These realities often surprise beginners.
In India, algorithmic trading simply means placing trades automatically via broker APIs rather than manually clicking Buy and Sell. Whether it is a basic rule-based strategy or a machine learning model, once an order is triggered without manual intervention, it falls under the umbrella of automated trading. And succeeding in this space requires much more than technical curiosity; it requires structural awareness.
The Regulatory Landscape: How SEBI Shapes Retail Automation
Automated trading in India is legal for retail investors but it operates under strict exchange guardrails. While institutions must get their specific ‘algorithms’ approved by the exchange, retail investors leverage broker-provided APIs to automate their strategies, ensuring all trades still pass through the broker’s real-time risk checks.
SEBI formalised retail API-based trading to ensure transparency and accountability. Traders cannot connect directly to exchanges like NSE or BSE. All orders must pass through a registered broker’s infrastructure. That broker becomes the gatekeeper, monitoring order frequency, enforcing risk checks, and ensuring compliance.
Each order generates unique identifiers at the broker and exchange levels, creating an audit trail that enables traceability and regulatory oversight. This creates a clear audit trail, ensuring traceability of strategy origin and execution responsibility. In short, the system is designed to protect market integrity, which means traders must operate within constraints.
Market Microstructure: The Reality of Competing in an Automated Arena
Understanding how orders are matched inside the exchange is non-negotiable.
According to the National Stock Exchange of India (NSE), algorithmic participation in 2025 reached approximately:
-
73% in Stock Futures
-
67% in Equity Derivatives
-
54% even in the Cash Market
This means that when a retail strategy fires a signal, it enters a highly competitive environment dominated by fast, optimised institutional systems.
The execution path looks simple:
Strategy → Broker API → Broker System → Exchange → Matching Engine → Execution → Feedback
But inside those milliseconds, speed and efficiency matter. A slight delay or poorly structured order can lead to slippage or adverse fills. Directional accuracy alone is not enough. Execution quality determines survival.
A Real Transition: From Software to Systematic Trading
Kranti Potluri’s journey illustrates how structured learning can bridge technology and finance. An IT graduate with 1.5 years of experience as a web developer, he used the COVID-19 work-from-home period to rethink his career direction. With a strong interest in automation, machine learning, and programming, he recognised that auto trading sat at the intersection of his skills and his fascination with markets.
Rather than casually experimenting, he committed to fully completing EPAT and transitioning into professional algorithmic trading. His story reflects a broader pattern: structured, disciplined education often makes the difference between curiosity and career transformation.
The Real Cost of Trading: Where Many Strategies Break
One of the biggest misconceptions in algorithmic trading in India is underestimating trading costs. In India, the Securities Transaction Tax (STT) is often the difference between a winning and losing strategy. Unlike US markets, where costs are negligible, Indian intraday strategies must clear a high ‘breakeven hurdle’ of taxes before booking a single rupee of profit.
Beyond basic brokerage, your backtest must explicitly deduct:
-
Securities Transaction Tax (STT): The largest component for equity delivery and F&O.
-
Exchange Transaction Charges: Fees levied by NSE/BSE.
-
GST (18%): Applied on brokerage and transaction charges.
-
SEBI Turnover Fees: A smaller but mandatory cost.
If your edge does not survive these deductions, it was never an edge to begin with.
Then comes slippage, the difference between the expected and actual execution price. In volatile or illiquid conditions, this can quietly erode returns.
There is also market impact cost. Larger or aggressively executed orders can move the market against the trader, particularly in instruments with limited order book depth. Institutions use order-slicing algorithms to minimise this footprint. Retail traders must factor these realities into their backtests. If your edge does not survive transaction costs, it was never an edge to begin with.
Operational Constraints: Engineering Discipline Is Essential
India’s regulatory and broker ecosystem introduces several technical guardrails.
Brokers enforce strict rate limits, often capping retail APIs at 1-2 orders per second to prevent flooding. Strategies relying on ‘machine-gun’ execution will hit these throttles instantly, leading to rejected orders and unhedged risk.
Other practical constraints include:
-
Static IP whitelisting for API access
-
Broker-imposed API rate limits
-
Risk of downtime, internet failures, or faulty code
Technology is a single point of failure. A coding mistake can create runaway order loops. Professional traders implement monitoring systems and emergency kill switches to protect capital.
This is where automation shifts from hobby to engineering discipline.
Strategy Decay and Overfitting: The Quiet Killers
There is a popular myth that algorithms generate effortless income. In reality, strategies decay. As more capital chases the same inefficiency, returns compress.
Overfitting is another major risk. When a model is excessively tuned to historical data, it captures noise instead of signal. Backtests look impressive, but live performance deteriorates quickly.
Robust strategies require:
-
Out-of-sample validation
-
Regime testing
-
Simplicity in design