Building Trading Strategies¶

What Makes a Trading Strategy¶

A trading strategy is a set of rules that determine when to buy, sell, or hold a security. These rules are based on market conditions, technical indicators, price patterns, or other factors.

A complete trading strategy specifies: 1. Entry conditions: When to open a position 2. Exit conditions: When to close a position 3. Position sizing: How much capital to allocate 4. Risk management: When to cut losses

In TzuTrader, strategies analyze market data and generate signals indicating whether to Buy, Sell, or Stay (do nothing). The backtesting engine then executes trades based on these signals.

Strategy Signals¶

A signal is a trading recommendation produced by a strategy. It contains:

Position: Buy, Sell, or Stay
Symbol: Which security to trade
Price: Current market price
Timestamp: When the signal was generated
Reason: Why the signal was generated (for analysis)

Example:

Signal(Buy AAPL @$150.25 at 2024-01-15 - RSI oversold: 28.5 < 30.0)

Using Pre-built Strategies¶

TzuTrader includes 16 pre-built strategies covering all major trading approaches:

Mean Reversion Strategies (6): - RSI Strategy - Classic overbought/oversold - Bollinger Bands Strategy - Volatility-based mean reversion - Stochastic Strategy - Momentum oscillator crossovers - MFI Strategy - Volume-weighted momentum - CCI Strategy - Statistical deviation from mean - Filtered Mean Reversion - RSI with trend filter

Trend Following Strategies (7): - Moving Average Crossover - Golden/death crosses - MACD Strategy - Momentum-based trend detection - KAMA Strategy - Adaptive moving average - Aroon Strategy - Time-based trend identification - Parabolic SAR - Trailing stop trend follower - Triple MA Strategy - Multi-timeframe confirmation - ADX Trend Strategy - Strength-filtered trends

Volatility Strategies (1): - Keltner Channel Strategy - ATR-based breakout/reversion

Hybrid Strategies (2): - Volume Breakout Strategy - Price + volume confirmation - Dual Momentum Strategy - ROC + trend filter

Let's explore each of these strategies in detail:

Original Four Strategies¶

RSI Strategy (Mean Reversion)¶

The RSI strategy trades based on overbought and oversold conditions:

import tzutrader

let strategy = newRSIStrategy(
  period = 14,        # Calculate RSI over 14 bars
  oversold = 30.0,    # Buy when RSI falls below 30
  overbought = 70.0   # Sell when RSI rises above 70
)

Trading logic: - Buy signal: RSI drops below the oversold threshold (default 30) - Sell signal: RSI rises above the overbought threshold (default 70) - Stay: RSI is between thresholds

When to use: - Range-bound markets (price oscillates without clear trend) - Mean reversion opportunities (price tends to return to average) - Shorter timeframes (minutes to days)

Characteristics: - Generates frequent signals in volatile markets - Can underperform in strong trends (price stays overbought/oversold) - Simple to understand and implement

Example:

let data = readCSV("data/AAPL.csv")
let strategy = newRSIStrategy(period = 14, oversold = 30, overbought = 70)
let report = quickBacktest("AAPL", strategy, data, 100000.0, 0.001)
echo report.summary()

Moving Average Crossover Strategy (Trend Following)¶

This strategy trades based on crossovers between fast and slow moving averages:

let strategy = newCrossoverStrategy(
  fastPeriod = 50,    # Fast moving average period
  slowPeriod = 200    # Slow moving average period
)

Trading logic: - Buy signal: Fast MA crosses above slow MA (golden cross) - Sell signal: Fast MA crosses below slow MA (death cross) - Stay: No crossover occurred

When to use: - Trending markets (price moves in one direction for extended periods) - Longer timeframes (days to weeks) - Following major trends

Characteristics: - Generates infrequent signals (only at crossovers) - Lags price movements (MAs use historical data) - Works well in strong trends, poorly in choppy markets

Common period combinations: - 50/200: Classic long-term crossover - 10/30: Medium-term trading - 5/20: Short-term trading

Example:

let strategy = newCrossoverStrategy(fastPeriod = 50, slowPeriod = 200)
let report = quickBacktest("AAPL", strategy, data, 100000.0, 0.001)

MACD Strategy (Momentum Shifts)¶

The MACD strategy trades when the MACD line crosses the signal line:

let strategy = newMACDStrategy(
  fastPeriod = 12,     # Fast EMA for MACD
  slowPeriod = 26,     # Slow EMA for MACD
  signalPeriod = 9     # Signal line EMA
)

Trading logic: - Buy signal: MACD line crosses above signal line (bullish crossover) - Sell signal: MACD line crosses below signal line (bearish crossover) - Stay: No crossover occurred

When to use: - Detecting momentum shifts - Confirming trend changes - Medium timeframes (hours to days)

Characteristics: - More responsive than simple MA crossover - Combines trend and momentum information - Standard parameters (12/26/9) work across many markets

Example:

let strategy = newMACDStrategy(fastPeriod = 12, slowPeriod = 26, signalPeriod = 9)
let report = quickBacktest("AAPL", strategy, data, 100000.0, 0.001)

Bollinger Bands Strategy (Mean Reversion)¶

This strategy trades when price touches the upper or lower Bollinger Band:

let strategy = newBollingerStrategy(
  period = 20,      # Middle band (SMA) period
  stdDev = 2.0      # Number of standard deviations for bands
)

Trading logic: - Buy signal: Price touches or drops below lower band - Sell signal: Price touches or rises above upper band - Stay: Price is within bands

When to use: - Range-bound markets - Identifying extremes - Mean reversion opportunities

Characteristics: - Adapts to volatility automatically (bands widen in volatile markets) - Works best in ranging markets - Can generate false signals during breakouts

Example:

let strategy = newBollingerStrategy(period = 20, stdDev = 2.0)
let report = quickBacktest("AAPL", strategy, data, 100000.0, 0.001)

Additional Pre-Built Strategies¶

TzuTrader provides 12 additional strategies beyond the original four. Here's a quick overview:

Mean Reversion Strategies:

Stochastic Strategy - Uses %K/%D crossovers in overbought/oversold zones. More sensitive than RSI for short-term trading.
MFI Strategy - Volume-weighted RSI. Best for liquid markets where volume confirms price action.
CCI Strategy - Statistical deviation-based. Originally designed for commodities, works well for cyclical patterns.
Filtered Mean Reversion - RSI with trend filter. Only buys dips in uptrends, sells rallies in downtrends.

Trend Following Strategies:

Aroon Strategy - Time-based trend detection. Catches trend starts early by measuring time since highs/lows.
KAMA Strategy - Adaptive moving average. Automatically adjusts to market conditions - fast in trends, slow in ranges.
Parabolic SAR - Provides both signals and trailing stops. Always in the market, accelerates as trends mature.
Triple MA Strategy - Requires alignment of three MAs. Fewer but higher-quality signals for position trading.
ADX Trend Strategy - Filters trades by trend strength. Only trades when ADX shows strong trending conditions.

Hybrid/Volatility Strategies:

Keltner Channel - ATR-based bands with dual modes. Trade breakouts in low volatility or reversions in high volatility.
Volume Breakout - Requires price breakout AND volume confirmation. Reduces false breakout signals significantly.
Dual Momentum - Combines ROC momentum with MA trend filter. Only trades momentum moves aligned with the trend.

For complete documentation including parameters, examples, and detailed trading logic, see: - Mean Reversion Strategies Reference - Trend Following Strategies Reference - Hybrid Strategies Reference

Choosing the Right Strategy¶

Different market conditions favor different strategies:

Best: Moving Average Crossover, MACD, Triple MA, ADX Trend, Parabolic SAR, KAMA
Why: These strategies ride trends and filter out noise
Consider: ADX Trend filters weak trends; Triple MA requires strong alignment
Avoid: RSI, Bollinger Bands, CCI, Stochastic (give premature exit signals)

Range-Bound Markets¶

Best: RSI, Bollinger Bands, Stochastic, MFI, CCI, Keltner (reversion mode)
Why: These strategies profit from price oscillations
Consider: MFI adds volume confirmation; Stochastic catches short-term swings
Avoid: Moving Average Crossover, MACD, Aroon (generate false signals)

High Volatility¶

Best: Bollinger Bands, Keltner Channel, Parabolic SAR
Why: These adapt to volatility automatically
Consider: Wider bands/stops in volatile conditions
Considerations: Wider stops, smaller positions

Low Volatility¶

Best: Volume Breakout, Keltner (breakout mode), Aroon
Why: Low volatility often precedes significant moves
Consider: Volume Breakout waits for volume confirmation
Considerations: May need patience for signals

Mixed/Uncertain Conditions¶

Best: KAMA, Filtered Mean Reversion, Dual Momentum, ADX Trend
Why: These adapt to conditions or use multiple filters
Consider: KAMA adjusts automatically; ADX only trades strong trends
Avoid: Single-indicator strategies without filters

By Trading Style¶

Day Trading (Short Timeframes): - Stochastic, MFI, CCI (fast oscillators) - Volume Breakout (intraday momentum) - Keltner Channel breakout mode

Swing Trading (Medium Timeframes): - RSI, MACD, Bollinger Bands - Dual Momentum, Filtered Mean Reversion - Aroon, KAMA

Position Trading (Long Timeframes): - Moving Average Crossover (50/200) - Triple MA Strategy - ADX Trend, Parabolic SAR

No strategy works in all conditions. The market environment, timeframe, and your trading style determine which approach is likely to succeed.

Creating Custom Strategies¶

To create a custom strategy, inherit from the base Strategy class and implement the onBar() method:

import tzutrader
import std/strformat

type
  MyCustomStrategy* = ref object of Strategy
    # Add your strategy's state here
    threshold*: float64
    ma*: MA

proc newMyCustomStrategy*(threshold: float64, maPeriod: int = 20): MyCustomStrategy =
  result = MyCustomStrategy(
    name: "My Custom Strategy",
    threshold: threshold,
    ma: newMA(maPeriod)
  )

method onBar*(s: MyCustomStrategy, bar: OHLCV): Signal =
  ## Process one bar and generate signal
  # Update indicators
  let maVal = s.ma.update(bar.close)

  var position = Position.Stay
  var reason = ""

  # Your trading logic here
  if not maVal.isNaN:
    if bar.close > s.threshold and bar.close > maVal:
      position = Position.Buy
      reason = &"Price ${bar.close} above threshold ${s.threshold} and MA ${maVal}"
    elif bar.close < s.threshold or bar.close < maVal:
      position = Position.Sell
      reason = &"Price ${bar.close} below threshold ${s.threshold} or MA ${maVal}"

  result = Signal(
    position: position,
    symbol: s.symbol,
    timestamp: bar.timestamp,
    price: bar.close,
    reason: reason
  )

Key points: - Store strategy parameters and indicators in the type definition - Implement onBar() to process each bar - Update indicators inside onBar() as data arrives - Always return a Signal with position, price, and reason - Use streaming indicators (newMA, newRSI, etc.)

Example: Dual RSI Strategy¶

A more practical custom strategy that uses two RSI periods:

import tzutrader
import std/strformat

type
  DualRSIStrategy* = ref object of Strategy
    shortPeriod*: int
    longPeriod*: int
    shortRSI*: RSI
    longRSI*: RSI

proc newDualRSIStrategy*(shortPeriod: int = 7, longPeriod: int = 21): DualRSIStrategy =
  result = DualRSIStrategy(
    name: "Dual RSI Strategy",
    shortPeriod: shortPeriod,
    longPeriod: longPeriod,
    shortRSI: newRSI(shortPeriod),
    longRSI: newRSI(longPeriod)
  )

method onBar*(s: DualRSIStrategy, bar: OHLCV): Signal =
  let shortVal = s.shortRSI.update(bar.close)
  let longVal = s.longRSI.update(bar.close)

  var position = Position.Stay
  var reason = ""

  if not shortVal.isNaN and not longVal.isNaN:
    if shortVal < 30.0 and longVal < 40.0:
      position = Position.Buy
      reason = &"Both RSIs oversold: short={shortVal:.1f}, long={longVal:.1f}"
    elif shortVal > 70.0 and longVal > 60.0:
      position = Position.Sell
      reason = &"Both RSIs overbought: short={shortVal:.1f}, long={longVal:.1f}"

  result = Signal(
    position: position,
    symbol: s.symbol,
    timestamp: bar.timestamp,
    price: bar.close,
    reason: reason
  )

method reset*(s: DualRSIStrategy) =
  s.shortRSI = newRSI(s.shortPeriod)
  s.longRSI = newRSI(s.longPeriod)
  s.history = @[]

This strategy only generates signals when both short-term and long-term RSI agree, potentially reducing false signals.

Strategy Development Tips¶

Start Simple¶

Begin with straightforward logic. Add complexity only if it clearly improves results. Simple strategies are: - Easier to understand - Less likely to overfit - More robust across different markets

Use Meaningful Reasons¶

Always provide descriptive reasons in your signals:

# Good
reason = &"Golden cross: fast MA ({fastMA:.2f}) > slow MA ({slowMA:.2f})"

# Bad
reason = "Buy signal"

Good reasons help you understand why the strategy made decisions and identify issues during analysis.

Test Incrementally¶

Don't write a complete complex strategy and then test it. Instead: 1. Start with basic logic 2. Backtest and verify it works 3. Add one feature 4. Backtest again 5. Repeat

This makes it easier to identify what helps and what doesn't.

Consider Edge Cases¶

Handle situations where indicators aren't ready:

if not rsiValue.isNaN:
  # Use RSI value
else:
  # Not enough data yet
  position = Stay

Avoid Look-Ahead Bias¶

Only use information available at the time of the decision. For example, don't use today's close price to make decisions earlier in the day.

Next Steps¶

Now that you understand how to build strategies, the next chapter covers portfolio management - how strategies interact with capital allocation, position sizing, and performance tracking.

Key Takeaways¶

A trading strategy defines rules for when to buy, sell, or hold
TzuTrader includes 16 pre-built strategies across 4 categories:
6 Mean Reversion strategies: RSI, Bollinger, Stochastic, MFI, CCI, Filtered Mean Reversion
7 Trend Following strategies: MA Crossover, MACD, KAMA, Aroon, Parabolic SAR, Triple MA, ADX Trend
1 Volatility strategy: Keltner Channel (dual mode)
2 Hybrid strategies: Volume Breakout, Dual Momentum
Choose strategies that match market conditions and your trading style
Trending markets favor trend-following; ranging markets favor mean reversion
Strategies with filters (ADX Trend, Dual Momentum, Filtered Mean Reversion) reduce false signals
TzuTrader uses streaming architecture - same code for backtesting and live trading
Create custom strategies by inheriting from the Strategy base class and implementing onBar()
Start simple and add complexity only when justified
Test incrementally and handle edge cases properly
Avoid look-ahead bias in your trading logic

Building Trading Strategies¶

What Makes a Trading Strategy¶

Strategy Signals¶

Using Pre-built Strategies¶

Original Four Strategies¶

RSI Strategy (Mean Reversion)¶

Moving Average Crossover Strategy (Trend Following)¶

MACD Strategy (Momentum Shifts)¶

Bollinger Bands Strategy (Mean Reversion)¶

Additional Pre-Built Strategies¶

Choosing the Right Strategy¶

Trending Markets¶

Range-Bound Markets¶

High Volatility¶

Low Volatility¶

Mixed/Uncertain Conditions¶

By Trading Style¶

Creating Custom Strategies¶

Example: Dual RSI Strategy¶

Strategy Development Tips¶

Start Simple¶

Use Meaningful Reasons¶

Test Incrementally¶

Consider Edge Cases¶

Avoid Look-Ahead Bias¶

Next Steps¶

Key Takeaways¶