Most people who ask us about algo trading have already read ten articles and come away more confused than when they started. This one skips the jargon and answers the questions in the order you will actually hit them.
What algo trading really is
Right now you sit in front of a screen at 9:15, watch a chart, and decide. Algo trading replaces the deciding part with a script that follows rules you wrote earlier.
You define the rule — say, buy Reliance if it drops below a level — and the script checks that condition continuously and places the order when it is met. No hesitation, no revenge trade after a loss, no missed entry because you were on a call.
That is the entire benefit, and it is a real one. The trade-off is that a bad rule now executes perfectly, at speed, every single time.
Is it legal in India?
Yes. SEBI's 2025 framework put formal rules around retail algo trading. As long as you are routing orders through a registered broker's official API — Zerodha, Angel One, Upstox and others all provide one — you are inside the regulated path.
What is not allowed is distributing a strategy to other people's accounts, or selling signals with performance claims, without the appropriate registration. Running your own algo on your own account is a different thing from operating an advisory.
What you need to learn
Less than you think, and it is worth being honest about the order.
- Python basics — variables, loops, functions, reading errors. A month of evenings is genuinely enough.
- Pandas — because everything in this field is a table of prices, and Pandas is how you slice one.
- Your broker's API — Kite Connect is the one most people start with.
- Strategy logic — the actual rules. This is the part people underestimate.
- Backtesting — testing those rules against history before they touch money.
Notice that only two of those five are programming. The coding is the easy half. Deciding what the rule should be, and proving it is not nonsense, is the hard half.
What it costs to start
Two separate numbers, and people usually only think about the first.
- Trading capital. Equity intraday is workable from around ₹25,000. Futures and options realistically need ₹1–2 lakh to be worth the effort, because margin requirements do not care how good your script is.
- Tooling. Kite Connect is ₹2,000 a month per app. Beyond that, a basic setup needs nothing you cannot get free.
A VPS becomes necessary once you go live — budget a few hundred rupees a month for that. Running a live algo on a laptop that sleeps, loses Wi-Fi, or restarts for an update is a mistake everyone makes exactly once.
A first script, end to end
This is deliberately the simplest thing that works — check a price, place an order if it is below a level.
from kiteconnect import KiteConnect
kite = KiteConnect(api_key="your_key")
kite.set_access_token("your_token")
ltp = kite.ltp("NSE:RELIANCE")["NSE:RELIANCE"]["last_price"]
if ltp < 2400:
kite.place_order(
tradingsymbol="RELIANCE",
exchange="NSE",
transaction_type="BUY",
quantity=10,
order_type="MARKET",
product="MIS",
variety="regular"
)
print("BUY order placed at", ltp)Read that and notice what is missing: there is no stop-loss, no check on whether you already hold a position, no handling for the API being down. A real bot is mostly those missing parts. The signal itself is usually ten lines.
The four mistakes that cost the most
- Going live first. Paper trade for a month. The gap between a backtest and reality shows up in that month, not in the spreadsheet.
- No stop-loss. An algo without one does not protect you from a bad day — it commits to it, automatically, faster than you could.
- Over-optimisation. If you tune the parameters until the backtest looks beautiful, you have fitted the strategy to the past, not found an edge. A curve that looks too good usually is.
- No plan for failure. Internet drops, tokens expire, the broker has an outage. Decide in advance what the script does when it cannot reach the API — because it will happen during a move, not during a quiet hour.
A sensible order of steps
- Open a Zerodha account if you do not have one.
- Create an app in the Kite developer console and get your API key.
- Install Python and run
pip install kiteconnect pandas. - Write one simple strategy — the example above is a fine starting point.
- Backtest it on two or three years of data.
- Paper trade it for a month, and actually look at the log every evening.
- Only then go live, with capital small enough that losing it teaches you something instead of hurting.
Where this usually goes
Algo trading is not difficult in the way people expect. The programming is ordinary. What separates people who stick with it from people who quit in six weeks is patience during the boring part — the backtesting, the paper trading, the evening spent reading logs.
If you would rather not build it yourself, that is what we do. Instacode.in builds custom trading systems around a strategy you already have, including the unglamorous parts — reconnection logic, order state handling and deployment.
One thing we say to everyone before starting: we build software, not returns. The system executes your rules. Whether those rules make money is your strategy and the market, and nobody honest will promise you otherwise.
FAQs
Do I need to know Python before starting?
You need the basics — variables, loops, functions and how to read an error message. A month of evenings with any decent beginner course is enough to write your first working script.
Does trading software guarantee performance in the market?
No. We are software developers, not registered investment advisers. An algo executes the rules you give it. Results depend entirely on your strategy and market conditions, and we do not guarantee any outcome.
Which broker is best for algo trading in India?
Zerodha's Kite Connect is the most widely used and best documented. Angel One's SmartAPI is a solid alternative, and Upstox and IIFL also offer working APIs. Pick the one where your account already is.
How much capital do I need to start?
Around ₹25,000 makes equity intraday practical. Futures and options realistically need ₹1–2 lakh because of margin requirements. Start smaller than feels exciting.
Need a custom solution?
Instacode builds production-grade software — algo trading, ecommerce, web apps. Let's talk.
Get in Touch
💬 Comments (0)
Be the first to comment 🚀