Anatomy of a Trading Strategy

The Five Components of Every Trading Strategy

Every profitable algorithmic strategy, from trend-following to machine learning, shares five core components. Master these and you can build anything.

Component 1: Universe Definition

Which instruments will you trade? 100 large-cap stocks? All liquid options? Crypto? Your universe determines your opportunity set.

• Liquidity filter: Minimum daily volume (prevents getting stuck in trades)
• Sector/market classification: What types of assets
• Geographic restrictions: US equities, global, etc.
• Quality screens: Minimum price, dividends, etc.

Component 2: Signal Generation

Under what conditions do you enter and exit? This is where your edge lives.

• Entry signal: The trigger that says "BUY" or "SELL"
• Exit signal: When to close the position (profit target or stop loss)
• Filters: Additional conditions that must be met (trend, volatility, etc.)
• Timing: When during the day/week to check signals

Component 3: Position Sizing

How much capital to deploy per trade? This is where the math of profitability lives.

• Fixed fraction: Risk 1% per trade
• Volatility-adjusted: Risk more in calm markets, less in volatile
• Kelly Criterion: Mathematical optimal sizing (advanced)
• Equal weight: Same $ on every position

Component 4: Risk Management Rules

• Stop-loss level: Maximum acceptable loss per trade
• Portfolio max drawdown: Maximum portfolio loss before stopping
• Position limits: Maximum number of concurrent positions
• Sector/market exposure: Maximum allocation to any one correlation group

Component 5: Execution Logic

HOW and WHEN to actually place orders.

• Order type: Market (instant) vs. Limit (price-specific)
• Order timing: At open, at close, randomly throughout day
• Slippage assumption: How much worse price than signal
• Commissions: Account for trading costs

# Pseudo-code structure of a strategy
for each day in historical_data:
    # Component 1: Universe
    tradeable_symbols = filter_liquid_stocks(all_stocks)

    for symbol in tradeable_symbols:
        # Component 2: Signal
        if is_oversold(symbol) and time_of_day == "open":
            signal = BUY

        # Component 3: Sizing
        position_size = calculate_position_size(
            account,
            signal,
            entry_price,
            stop_loss_price
        )

        # Component 4 & 5: Execute with risk management
        if position_size > 0 and current_open_positions < max_positions:
            place_order(symbol, position_size, stop_loss_price)