Backtest Command¶

The backtest command tests trading strategies against historical data to evaluate their performance and characteristics.

Synopsis¶

sqa-cli backtest [OPTIONS]

Description¶

Backtesting allows you to test how a trading strategy would have performed historically. This helps:

Evaluate strategy effectiveness
Compare different approaches
Understand risk/return characteristics
Identify strategy weaknesses
Optimize parameters

Options¶

Option	Type	Default	Description
`-t, --ticker SYMBOL`	String	AAPL	Stock ticker symbol
`-s, --strategy NAME`	String	RSI	Strategy to backtest
`-c, --capital AMOUNT`	Float	10000	Initial capital
`--commission AMOUNT`	Float	1.0	Commission per trade
`--compare`	Boolean	false	Compare all strategies
`-v, --verbose`	Boolean	false	Detailed output
`-h, --help`	Boolean	false	Show help

Available Strategies¶

Technical Indicator Strategies¶

RSI (Relative Strength Index) - Type: Momentum oscillator - Signals: Overbought (>70), Oversold (<30) - Best for: Mean reversion in ranging markets

SMA (Simple Moving Average) - Type: Trend following - Signals: Price crosses above/below moving average - Best for: Trending markets

EMA (Exponential Moving Average) - Type: Trend following (more responsive than SMA) - Signals: Similar to SMA but faster - Best for: Faster trending markets

MACD (Moving Average Convergence Divergence) - Type: Momentum and trend - Signals: MACD line crosses signal line - Best for: Identifying trend changes

Bollinger Bands - Type: Volatility-based - Signals: Price touches upper/lower bands - Best for: Identifying overbought/oversold in volatile markets

Stochastic - Type: Momentum oscillator - Signals: Similar to RSI - Best for: Identifying reversal points

Custom Strategies¶

Volume Breakout - Type: Volume-based - Signals: High volume with price movement - Best for: Breakout trading

KBS (Knowledge-Based Strategy) - Type: Rule-based inference - Signals: Multiple indicator combinations - Best for: Complex decision making

Consensus - Type: Ensemble method - Signals: Agreement of multiple strategies - Best for: Reducing false signals

Random - Type: Baseline/control - Signals: Random buy/sell - Best for: Performance comparison

Performance Metrics¶

Total Return¶

Percentage gain or loss over the backtest period.

Total Return: 15.3%

Annualized Return¶

Return normalized to annual basis for comparison.

Annualized Return: 12.8%

Sharpe Ratio¶

Risk-adjusted return (higher is better).

Sharpe Ratio: 1.45

Interpretation: - < 1.0: Poor risk-adjusted performance - 1.0-2.0: Good performance - > 2.0: Excellent performance

Maximum Drawdown¶

Largest peak-to-trough decline (lower is better).

Max Drawdown: -8.2%

Win Rate¶

Percentage of profitable trades.

Win Rate: 58.3%

Total Trades¶

Number of trades executed.

Total Trades: 24

Profit Factor¶

Ratio of gross profits to gross losses (>1 is profitable).

Profit Factor: 1.67

Examples¶

Basic Backtest¶

sqa-cli backtest --ticker AAPL --strategy RSI

Output:

======================================================================
Backtesting RSI on AAPL
======================================================================

Backtest Results:
  Total Return: 15.30%
  Annualized Return: 12.80%
  Sharpe Ratio: 1.45
  Max Drawdown: -8.20%
  Win Rate: 58.33%
  Total Trades: 24
  Profit Factor: 1.67

Different Strategy¶

sqa-cli backtest --ticker MSFT --strategy MACD

Custom Capital¶

sqa-cli backtest --ticker GOOGL --strategy SMA --capital 50000

Starts with $50,000 instead of default $10,000.

Custom Commission¶

sqa-cli backtest --ticker TSLA --strategy RSI --commission 5.0

Uses $5 commission per trade (default is $1).

Compare All Strategies¶

sqa-cli backtest --ticker AAPL --compare

Output:

======================================================================
Comparing All Strategies
======================================================================

Strategy             Return%     Sharpe  Drawdown%  WinRate%    Trades
----------------------------------------------------------------------
MACD                   18.50       1.75      -7.20     61.50        26
RSI                    15.30       1.45      -8.20     58.33        24
BollingerBands         12.80       1.32      -9.50     55.00        20
SMA                    11.20       1.15     -10.30     52.50        18
EMA                    10.50       1.08     -11.00     51.20        19
Stochastic              8.30       0.95     -12.50     48.00        25
VolumeBreakout          6.50       0.82     -15.20     45.00        15
Consensus               5.20       0.75     -13.80     50.00        22
KBS                     4.80       0.68     -14.50     47.50        21
Random                 -2.30      -0.15     -20.00     48.00        30

Best Strategy: MACD (18.50% return)

Verbose Output¶

sqa-cli backtest --ticker AAPL --strategy RSI --verbose

Shows trade-by-trade details and intermediate calculations.

Interpreting Results¶

Good Strategy Characteristics¶

✅ Total Return: > 10% annually
✅ Sharpe Ratio: > 1.0
✅ Max Drawdown: < 20%
✅ Win Rate: > 50%
✅ Profit Factor: > 1.5

Red Flags¶

❌ Negative total return
❌ Sharpe Ratio < 0
❌ Max Drawdown > 30%
❌ Win Rate < 40%
❌ Profit Factor < 1.0
❌ Very few trades (< 10)

Common Workflows¶

Strategy Evaluation¶

# 1. Backtest single strategy
sqa-cli backtest --ticker AAPL --strategy RSI

# 2. If promising, compare with others
sqa-cli backtest --ticker AAPL --compare

# 3. If best performer, optimize parameters
sqa-cli genetic --ticker AAPL

Multi-Stock Analysis¶

# Test strategy across multiple stocks
for ticker in AAPL MSFT GOOGL TSLA; do
  echo "Testing $ticker..."
  sqa-cli backtest --ticker $ticker --strategy RSI
done

Parameter Sensitivity¶

# Test with different capital amounts
for capital in 10000 25000 50000 100000; do
  sqa-cli backtest --ticker AAPL --strategy RSI --capital $capital
done

Tips and Best Practices¶

Strategy Selection¶

Trending markets: Use SMA, EMA, MACD
Ranging markets: Use RSI, Stochastic, Bollinger Bands
Volatile markets: Use Bollinger Bands, Volume Breakout
Complex decisions: Use KBS, Consensus

Avoiding Overfitting¶

Use out-of-sample testing:
Train on first 70% of data
Test on last 30%
Keep strategies simple:
Fewer parameters = less overfitting
More rules = higher chance of curve fitting
Test across multiple stocks:
Strategy should work on different securities
Not just optimized for one ticker

Realistic Expectations¶

Commission matters: Small edges disappear with costs
Slippage: Real execution differs from backtest
Market impact: Large orders move prices
Past ≠ Future: Historical performance doesn't guarantee future results

Data Quality¶

Sufficient history: At least 1-2 years of data
Survivorship bias: Don't only test on successful stocks
Corporate actions: Account for splits, dividends
Data errors: Verify data integrity

Limitations¶

What Backtesting Can't Tell You¶

Future market conditions
Black swan events
Regime changes
Execution challenges
Psychological factors

Known Issues¶

Sample data is synthetic - use real data for actual trading
No slippage modeling in current version
No market impact modeling
Assumes instant execution at close price

Advanced Topics¶

Walk-Forward Analysis¶

Test strategy over rolling windows:

# Future enhancement
sqa-cli backtest --ticker AAPL --strategy RSI --walk-forward

Monte Carlo Simulation¶

Randomize trade order to test robustness:

# Future enhancement
sqa-cli backtest --ticker AAPL --strategy RSI --monte-carlo

analyze - Find entry/exit points to backtest
genetic - Optimize strategy parameters
pattern - Discover new trading patterns
optimize - Portfolio-level optimization