Stock Market · Trading Systems

Building a rule-based system

A rule-based trading system is a set of WRITTEN, SPECIFIC, UNAMBIGUOUS instructions for entering, managing, and exiting trades. Anyone reading your rules should be able to execute them and produce the same trades you would.

The 5 components every system needs

1. UNIVERSE: what you trade (Nifty 50, Bank Nifty stocks, etc.)

2. SETUP: market conditions that signal a potential trade

3. ENTRY: exact trigger (price, indicator value, candle pattern)

4. RISK MANAGEMENT: position sizing, stop loss

5. EXIT: target, trailing stop, time-based exit

Example: simple moving average crossover

UNIVERSE: Nifty 50 stocks with daily volume > 1 crore shares.

SETUP: Stock above 200-day SMA (trending up).

ENTRY: Buy at close when 20-day SMA crosses above 50-day SMA.

STOP: 2× ATR below entry.

EXIT: Sell when 20-day SMA crosses below 50-day SMA OR stop hits.

POSITION SIZE: 2% of capital per trade.

> Anyone with these rules could trade this strategy. There's no judgment, no 'I think it'll go up'. Either the rules trigger or they don't.

Writing rules, the test

Bad rule: 'Buy when the stock looks strong.' (Subjective)

Good rule: 'Buy when RSI(14) crosses above 50 from below.' (Objective)

Bad rule: 'Exit when the trend ends.' (Subjective)

Good rule: 'Exit when price closes below the 20-day EMA.' (Objective)

Refinement happens after data

Don't try to design a perfect system upfront. Start simple, trade it for 50+ trades, then refine based on what the data shows. Tweaking before you have data is just guessing.

Takeaway. Rule-based systems have 5 components: universe, setup, entry, risk management, exit. Every rule must be objective. Testable, not subjective. Start simple, trade for 50+ trades, then refine based on data.