Position Sizing for Algo Trading: Fixed Lots vs Risk Percent

Sizing is the strategy

Take any profitable signal stream and you can still destroy an account with it - just size badly. Sizing determines drawdown depth, recovery time, and whether the strategy survives its inevitable losing streak. In automation this matters double: the machine will take every signal at whatever size you configured, including the ninth loser in a row at 3am.

Three models cover almost every automated setup: fixed lots, percent of balance, and risk-based sizing. They differ in one thing - what stays constant as your balance and your stop distances change.

The three models

Risk-based sizing is what professionals default to, and in signal terms it is one field: size_type set to risk with a risk percentage - the EA does the arithmetic per trade using your stop.

• •Fixed lots: every trade is the same size (0.10 lots). Constant: the lot size. Simple, predictable, but risk per trade swings wildly with stop distance, and it never compounds - your size at 50k equals your size at 5k
• •Percent of balance: size scales with account value. Constant: exposure relative to equity. Compounds up and de-risks down automatically, but still ignores stop distance
• •Risk-based: you define the loss (say 1% of balance) and the size is computed from the stop distance - a 20-pip stop gets a bigger position than an 80-pip stop, both losing exactly 1% if hit. Constant: the thing that actually matters

Choosing the risk number

The famous 1-2% rule exists because of losing-streak math. At 1% risk, ten consecutive losses - which every real strategy eventually produces - costs under 10% of the account, recoverable with an 11% gain. At 5% risk the same streak costs 40%, needing a 67% gain to recover. The asymmetry of drawdowns is the entire argument: losses hurt more than equivalent gains help, so the risk number should be set for the losing streak, not the average day.

For strategy-mode automation where TradingView controls sizing via placeholders, do the same math inside your Pine strategy - and keep a terminal-level daily loss limit as the backstop either way.