The answer people want is a single number. The answer that is actually useful is that your capital has to clear two separate hurdles: it must be large enough that fixed monthly costs are not eating the returns, and large enough that your risk rule produces a position size the market will let you trade.
Most people who ask this question have only thought about the second one. The first is what quietly kills small accounts, and it is simple arithmetic that nobody writes down.
Zero, for the first three months
Start here because it is true and it saves people money. Learning to build a system costs nothing in capital.
Historical data for backtesting is either free or comes with a subscription you can take for a single month. Paper trading against live prices costs nothing. The virtual trading approach — running your strategy on real market movement with simulated fills — is where the first several months should be spent regardless of how much money you have available.
Anybody who tells a beginner they need capital before they can start learning is selling something. The capital question only becomes real at the point where a tested strategy exists.
The fixed costs, and the arithmetic nobody spells out
This is the part that decides whether a small account can work at all.
Running an automated system in India has recurring costs that do not scale down with your account size. A broker API subscription is billed monthly. A small cloud server with a static IP costs a few hundred rupees a month. You may add a data or charting subscription.
Call it somewhere between 1,000 and 2,000 rupees a month all in, so 12,000 to 24,000 a year. Now look at what that means as a percentage of capital:
| Capital | Annual fixed cost at 18,000 | Return needed just to break even |
|---|---|---|
| 50,000 | 18,000 | 36% |
| 1,00,000 | 18,000 | 18% |
| 3,00,000 | 18,000 | 6% |
| 5,00,000 | 18,000 | 3.6% |
| 10,00,000 | 18,000 | 1.8% |
Read the first row again. On fifty thousand rupees, your strategy has to return 36% before you have made a single rupee — and that is before brokerage, STT and slippage on the trades themselves. There is no strategy worth running that reliably clears that bar.
This is the single most useful number in this article. Below roughly two to three lakh, the infrastructure is a tax you cannot outrun, and the sensible move is either to run the strategy manually without the server and API subscription, or to keep paper trading until the capital is there.
What each strategy type actually needs
The second hurdle is the market's own minimums. These vary by instrument and change with exchange rules and broker policy, so treat the shapes as guidance and check your broker's margin calculator for current figures.
Positional equity, delivery. The friendliest starting point. There is no minimum beyond the price of one share, and you control position size precisely because you can buy any quantity. A workable account starts around one to two lakh, mostly so the fixed costs make sense.
Intraday equity. Intraday margin lets you take a larger position than your capital, which cuts both ways. The real constraint is that your per-trade risk should be a small fraction of capital, and leverage makes it very easy to breach that without noticing.
Index option buying. You pay the premium, so the cost of one lot is the entry ticket. It is the cheapest way into derivatives and also the one with the highest rate of total loss on individual positions, because an option can expire worthless.
Index option selling. The expensive one. Selling requires span and exposure margin blocked per lot, which for index contracts runs to well over a lakh per lot, and more if you are not running a hedged structure. Hedged positions require materially less. This is where people discover their capital plan does not survive contact with the margin file.
Futures. Similar picture — margin is blocked per contract, and it is a large number relative to a retail account.
Why small accounts fail even with a good strategy
Suppose your risk rule says you never lose more than 1% of capital on a single position. That is a standard, sensible rule.
On ten lakh, 1% is ten thousand rupees of risk per position. You can build almost any strategy inside that budget.
On fifty thousand, 1% is five hundred rupees. Now try to find a position where your stop-loss distance costs you five hundred rupees or less. In equity you can, by buying a small quantity. In index options you cannot, because one lot moves by more than that on an ordinary day.
So the small account faces a choice, and both options are bad. Either break the risk rule and take a position sized far beyond 1%, or do not trade the instrument at all. Almost everybody chooses the first, which is why small derivative accounts do not fail slowly — they fail in a handful of positions.
The lot size is the hard edge here. Equity lets you scale down to one share; derivatives do not let you go below one lot, and that lot has a fixed dependence on the index level. Your risk rule has to fit around it rather than the other way round.
A sensible starting range
Bringing the two hurdles together, for someone with a tested strategy and no intention of losing money they need:
- Under 1 lakh. Keep paper trading, or trade equity manually without the API and server subscriptions. The fixed costs do not make sense yet.
- 1 to 3 lakh. Workable for positional or intraday equity. Derivatives are difficult here because lot sizes do not fit a sane risk rule.
- 3 to 10 lakh. The range where an automated equity system genuinely makes sense, and where option buying strategies become manageable.
- Above 10 lakh. Option selling and multi-strategy approaches become feasible, with the margin requirements they carry.
None of this is a recommendation to deploy any particular amount. It is the arithmetic of what different amounts allow.
Money that must not go anywhere near this
Regardless of the numbers above:
- Borrowed money. A loan has a fixed repayment schedule and the market has no obligation to cooperate with it.
- Your emergency fund. The point of it is being available in a bad month, which is exactly when a drawdown is most likely.
- Business working capital. Once a supplier payment depends on a position closing well, the rules stop governing the position.
- Money with a date attached — school fees, a wedding, a down payment. A deadline turns an ordinary drawdown into a forced exit at the worst possible moment.
- Retirement capital, in a leveraged strategy. A drawdown a salaried person absorbs because income continues is not recoverable when it does not.
Scaling up by rule rather than by feeling
The most expensive mistake in the whole sequence is increasing size after a good month. It feels like confirmation and it is usually just variance.
Decide the rule for scaling before you go live, and write it down with the strategy: for example, increase size by a fixed step only after the system has completed a set number of trades and the live results are within a defined distance of what the backtest predicted. That is a condition your code can check. Confidence is not.
The same applies downwards. A rule that reduces size after a defined drawdown will do it without argument, at exactly the moment you would be least willing to.
Your account also needs a broker behind it, and that account is where both the historical data for testing and the margin requirements for live positions come from.
Margin numbers come from your broker
Requirements change with the index level and exchange rules. Account free to open
If you want the whole sequence taught in order
Position sizing, cost modelling and the deployment costs discussed above are all sections of our Algorithmic Trading with Python course at Rs 24,900 — 50+ modules from Python basics through the Kite Connect API, honest backtesting, a virtual trading system, and running the finished thing on a cloud server. One payment, lifetime access, taught in a Hindi-English mix, with a free demo on WhatsApp first.
If the demo suggests your capital is not there yet, you will be told to keep paper trading rather than sold something.
We do not sell tips or signal groups, we manage nobody's money, and we make no claim about returns. 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
How much money do I need to start algo trading in India?
Nothing to learn and to paper trade. To run a live automated system, the fixed costs of an API subscription and a cloud server make sense only from roughly two to three lakh upwards - below that, the infrastructure demands an unrealistic annual return just to break even.
Can I start algo trading with 50,000 rupees?
You can learn, backtest and paper trade with it. Running a live automated system on it is difficult: annual infrastructure costs of around 18,000 would require a 36% return simply to break even, and a 1% risk rule gives you only 500 rupees of risk per position, which no index option lot fits inside.
How much capital is needed to sell options?
Considerably more than to buy them. Span and exposure margin is blocked per lot and runs well over a lakh for index contracts, with hedged structures requiring materially less. Check your broker's margin calculator, since the figures move with the index level and exchange rules.
Why do small trading accounts fail so often?
Position sizing granularity. A 1% risk rule on a small account produces a risk budget smaller than one derivative lot can move in a day, so the trader either abandons the rule or abandons the instrument. Most abandon the rule, and a few positions then do the damage.
Should I increase my position size after a profitable month?
Not on the basis of the month. Decide a scaling rule before going live - a fixed step after a set number of trades, conditional on live results tracking the backtest - so that the decision is checked by code rather than by confidence.
What are the monthly costs of running a trading bot?
Typically a broker API subscription plus a small cloud server with a static IP, together somewhere around 1,000 to 2,000 rupees a month, before any optional data or charting subscriptions.
Related Reading
- How to start algo trading in India - the full sequence
- Backtesting a trading strategy in Python
- Zerodha API charges, broken down
- Straddle vs strangle - the margin difference
- Running a trading bot 24x7
- Algorithmic Trading with Python - full syllabus
- All course fees, stated plainly
Disclaimer: TheFinBaba provides educational content only - this is not investment advice. Trading involves risk of loss.
