How much money do you need to start algo trading?

“How much do I need?” has two separate answers. The first is the cost of building and running the bot. The second is the trading capital the strategy needs so that each trade risks a sensible share of the account. Most guides give one number for both. The honest answer depends on your stop distance, your broker’s minimum lot and the risk you accept per trade.

All dollar figures in this guide are illustrative numbers. Replace them with your own broker’s and provider’s current prices.

What does it cost to build and run an EA or bot?

If you want the build done for you, our EA development service quotes a fixed price from your written rules.

Why do fixed costs matter more on a small account?

A VPS costs the same whether your account holds $500 or $50,000. Say it costs $20 a month (illustrative). That is $240 a year, which the strategy must earn before you make anything:

The same logic applies to a one-off development cost spread over the first year.

How does the minimum lot size set your minimum account?

The money at risk on a trade is: lots × stop distance × value per pip. Your broker sets the smallest trade it accepts, the minimum lot. On many forex accounts that is 0.01 lot, which is 1,000 units. On EURUSD with a USD account, 0.01 lot is worth about $0.10 per pip.

So with a 40-pip stop, the smallest possible risk is about $4, whatever your account size. Here is what a 1% risk target becomes after rounding to the broker’s lot step:

Wider stops and other markets make it worse. On gold, many brokers set 1 lot at 100 ounces, so 0.01 lot moves about $1 for every $1 change in price. A $15 stop at the minimum lot then risks $15: 3% of a $500 account. Contract sizes vary by broker, so check the symbol’s specification.

Here is the same calculation in Python, so you can run your own numbers. The position size calculator does it in your browser.

1import math2 3def lots_for_risk(balance, risk_pct, stop_pips, pip_value_per_lot,4                  min_lot=0.01, lot_step=0.01):5    """Lots for a % risk, rounded DOWN to the broker's step.6    Returns (lots, the real risk in % after rounding)."""7    risk_money = balance * risk_pct / 1008    lots = risk_money / (stop_pips * pip_value_per_lot)9    lots = math.floor(lots / lot_step + 1e-9) * lot_step10    if lots < min_lot:11        lots = min_lot  # the broker will not accept less12    real_risk = lots * stop_pips * pip_value_per_lot13    return round(lots, 2), round(real_risk / balance * 100, 2)14 15# EURUSD, USD account: about $10 per pip for 1.00 lot. 40-pip stop, 1% target.16print(lots_for_risk(300, 1, 40, 10))    # (0.01, 1.33)  forced above target17print(lots_for_risk(1000, 1, 40, 10))   # (0.02, 0.8)18print(lots_for_risk(5000, 1, 40, 10))   # (0.12, 0.96)

Why do tiny accounts force oversized risk?

When the minimum lot already risks more than your target, three things happen:

• Risk per trade is set by the broker, not by you. On $100 with a 40-pip stop, every trade risks 4%.
• Risk cannot shrink after losses. Percentage sizing normally reduces the lots as the account falls. At the minimum lot it cannot. Ten losses in a row at $4 each cost $40: 40% of a $100 account. On $2,000 at 1% per trade, the same streak costs under 10%.
• Several EAs multiply it. Two strategies at the minimum lot double the risk on the same small balance.

Small accounts also tempt traders to make the effort “worth it” with large lots or martingale. That is a fast way to lose a small account. See grid and martingale EA risks.

A worked example: one EA on EURUSD

Illustrative assumptions: one EA on EURUSD H1, stops between 30 and 50 pips, 1% target risk per trade, a VPS at $20 a month.

For this example, $500 is the technical minimum: the widest stop at the smallest lot still fits the 1% rule. But the size barely adjusts to the stop, and the fixed cost is a large hurdle. Around $2,000, the rounding and the costs stop dominating. Change the stop, the market or the costs and these numbers move, which is why you should run your own.

Is a prop-firm account a better route with little capital?

It can be. You pay an evaluation fee, trade the firm’s account under its rules, and if you pass, you share the profits. The fee is the most you can lose on that attempt, and the account is usually much larger than what you would fund yourself. Some firms refund the fee later; check the terms.

The costs are different, not absent:

• Each failed attempt means another fee.
• The daily loss and maximum drawdown rules are strict. Learn how your firm measures them in static vs trailing drawdown, and check your room with the prop-firm drawdown calculator.
• Not every firm allows every EA. Read can you use an EA on a prop-firm account? before you pay.

An EA for a funded account must enforce the firm’s limits in code. That is what our prop-firm EA development covers. If you are in India, check first which markets and brokers you may use; see is algo trading legal in India?

What should you budget before you start?

• Development: your time, or a fixed quote from a developer
• Demo testing for one to three months: free, apart from the VPS
• A VPS for demo and live trading
• Capital large enough that the minimum lot at your widest stop fits your risk per trade
• Only money you can lose without it affecting your life