Intraday trading is the most popular thing in Indian retail markets and the one with the worst reported outcomes. Both facts have the same explanation, and it is arithmetic rather than psychology.
This is about why the approach is structurally harder than it looks, and what an intraday system genuinely has to handle. It is not a list of setups, because a setup without the arithmetic underneath it is the reason the outcomes look the way they do.
The arithmetic that decides it
Start here, because everything else follows.
Every round trip pays the full cost stack: brokerage, STT, exchange transaction charges, GST on those, SEBI turnover fees, stamp duty, and the spread you crossed to get filled. Call the total a few basis points per side on a liquid instrument — more on anything thin.
Now put that against an intraday move. If your average winning position captures half a percent, and the round trip costs a meaningful fraction of that, then a large share of what you produce goes to costs before anything reaches you. Take a position a day and you are paying that stack roughly two hundred and fifty times a year.
The consequence is precise: intraday requires a materially higher edge than a longer holding period does, simply to arrive at the same net result. Not more discipline, not more screen time. A better edge, because the friction is larger relative to the move.
This is the calculation almost nobody does before starting, and doing it explains the reported outcomes better than any discussion of emotion does.
Noise, and why short timeframes are harder to test
A daily candle summarises a session. A five-minute candle summarises five minutes, and a much larger share of what it contains is randomness rather than information.
Two things follow. Signals fire far more often on short timeframes, most of them meaning nothing, which is why intraday systems typically show a lower win rate on the same underlying logic. And backtests on short timeframes are easier to overfit, because there are so many more data points to accidentally fit to.
There is a practical data problem as well. Minute data is limited in how far back most brokers will give it, and it is heavier to store and slower to test across. A five-year daily backtest is quick; five years of minute data across a watchlist is a genuine engineering exercise before any strategy work begins.
What an intraday system must handle
Assuming the arithmetic above still works for your idea, the implementation is meaningfully harder than a swing system's.
Live data, properly. A WebSocket feed rather than polling, with strategy logic off the callback thread, reconnection with backoff, and a watchdog for the connection that stays open and stops delivering.
Speed that is adequate rather than fast. You are not competing on microseconds, but your loop must complete well inside your decision interval. Falling behind during a fast move means acting on stale prices at exactly the wrong time.
The auto square-off. Intraday product positions are closed by the broker near the end of the session, at whatever the market offers. Your system should close its own positions on its own schedule rather than being closed by somebody else's.
A daily loss limit that actually stops it. More important intraday than anywhere else, because the opportunity to keep trading after losses is continuous.
Restart safety. A crash at 11:40 with a position open is a real scenario. On restart the program must read its position from the broker rather than assume a clean slate.
Margin awareness. Intraday leverage makes it easy to take a position larger than your risk rule intended. Size from the stop distance, never from what the margin permits.
Testing an intraday idea without fooling yourself
The traps are the same as elsewhere but they bite harder, because the edge being measured is smaller relative to the errors.
Costs are not optional. Leaving them out of an intraday backtest does not make the result optimistic, it makes it meaningless. Subtract a pessimistic per-side figure before forming any opinion.
Assume you were filled badly. Not at the price on screen. Buying at the ask and selling at the bid, plus slippage during fast moves, because that is when your signals fire.
Check the trade count and then the distribution. An intraday system generates plenty of trades, which is good for significance and bad for costs. Look at whether the result comes from many small wins or a few large ones, because those two profiles behave very differently when conditions change.
Test the bad days separately. How did it perform on the ten worst days for the index? A system that is flat on ordinary days and loses heavily on volatile ones is a system that gives everything back periodically.
Who this actually suits
Worth being direct, because this is the approach most people start with and it suits the fewest of them.
It does not suit anybody with a job. Positions needing decisions during market hours cannot be managed from a meeting, and attention given in gaps produces the worst version of it.
It does not suit a small account. Fixed costs and the cost stack both weigh more heavily, and derivative lot sizes make a sane risk rule impossible to fit.
It does not suit somebody without a tested edge. The higher friction means an idea that would survive a monthly holding period may not survive a daily one.
It can suit an automated system with a tested edge, real cost modelling, a hard loss limit, and infrastructure that stays up — run by somebody who has already done this on a longer timeframe.
If you are drawn to intraday because it is where the activity is, that is worth examining. Activity and results are different things, and the costs make them actively opposed.
Any of it needs a broker account with the data and execution to support it.
Intraday needs live data, not snapshots
The WebSocket feed and the orders come from the same account. Free to open
Cost modelling, live data handling and the loss limits above are sections of our Algorithmic Trading with Python course at Rs 24,900 — one payment, permanent access, free demo on WhatsApp first.
We sell no tips and no signal group, we manage nobody's money, and we promise no returns. Nothing here recommends any position. Trading carries a real risk of loss.
Disclosure: the account-opening link on this page is under Atul Shrivastava's Zerodha Authorised Person registration (NSE AP Reg: AP2516003481; Zerodha Broking Ltd. SEBI Reg: INZ000031633) and earns a revenue share. TheFinBaba is not a SEBI-registered Investment Adviser — this content is educational, not investment advice.
Frequently Asked Questions
Why is intraday trading so hard in India?
Arithmetic. The full cost stack is paid on every round trip and is large relative to a typical intraday move, so the approach needs a materially higher edge than a longer holding period simply to reach the same net result.
Is intraday trading better than swing trading?
Not structurally. Intraday pays costs a fraction as often is false - it pays them far more often, faces more noise per signal, is easier to overfit, and requires presence during market hours. Swing is the better default for almost everybody who is not automating.
What does an intraday trading system need to handle?
A live WebSocket feed with strategy logic off the callback thread, reconnection and a stale-feed watchdog, closing its own positions before the broker's auto square-off, a hard daily loss limit, safe restarts that read positions from the broker, and sizing from the stop rather than from available margin.
How do I backtest an intraday strategy properly?
Subtract a pessimistic cost figure per side, assume fills at the wrong side of the spread with slippage during fast moves, judge on trade count and the shape of the distribution rather than the total, and check performance on the most volatile days separately.
Can a working professional do intraday trading?
Not well, unless it is fully automated. Positions needing decisions during market hours cannot be managed from a meeting, and intermittent attention produces the worst version of the approach.
Related Reading
- Swing trading in India
- How to structure a retail algo trading system
- Live tick data with the Kite Connect WebSocket
- Risk management in trading
- Common algo trading mistakes
- Algorithmic Trading with Python - full syllabus
Disclaimer: TheFinBaba provides educational content only - this is not investment advice. Trading involves risk of loss.
