quant-trading

name	quant-trading
description	Quantitative trading strategy development, backtesting, and risk management
schema	1.0
version	1.1.0
triggers	{"keywords":{"primary":["quant","trading","quantitative","trade","backtest","backtesting","algo-trading","algorithmic trading"],"secondary":["strategy","strategy design","factor","factor model","arbitrage","arbitrage trading","hedge","hedging"]},"context_boost":["python","pandas","numpy","finance","investment","stock","futures"],"context_penalty":["design","marketing","frontend"],"priority":"high"}
keywords	["finance","trading","quantitative","investment"]
dependencies	{"software-skills":["python","database","data-analysis"]}
author	claude-domain-skills
metadata	{"mcpmarket-version":"1.0.0"}

Quant Trading

Systematic, data-driven trading strategy development

When to Use

• Developing trading strategies (trend following, mean reversion, arbitrage)
• Strategy backtesting and performance analysis
• Risk management and position control
• Factor research and alpha discovery

Strategy Development Process

Hypothesis → Data prep → Strategy coding → Backtest validation → Risk control → Live monitoring

Core Knowledge

Strategy Types

Type	Description	Risk
Trend following	Go with the trend, buy strength / sell weakness	Loses in choppy markets
Mean reversion	Revert after price deviates	Loses in trending markets
Statistical arbitrage	Pairs trading, spread convergence	Correlation breakdown
High-frequency trading	Capturing spreads at microsecond scale	High technical risk

Risk Metrics

Metric	Formula	Healthy Standard
Sharpe ratio	(Return - risk-free) / std dev	> 1.5
Maximum drawdown	Largest peak-to-trough decline	< 20%
Calmar ratio	Annualized return / max drawdown	> 1.0
Win rate	Winning trades / total trades	> 50%

Common Factors

Factor	Description	Rationale
Value	Low P/E, P/B	Cheap stocks outperform long term
Momentum	Past winners keep winning	Trend persistence
Quality	High ROE, low debt	Premium for good companies
Size	Small-cap premium	Liquidity compensation
Volatility	Low-volatility anomaly	Low risk, high return

Backtest Checklist

• Data quality: Is there survivorship bias?
• Look-ahead bias: Are you using future data?
• Overfitting: Are the parameters over-optimized?
• Trading costs: Are commissions and slippage included?
• Out-of-sample testing: Did you hold out a test set?

Best Practices

1. Simple before complex - Start with a simple strategy, add complexity gradually
2. Out-of-sample validation - Always hold out a slice of data for the final test
3. Account for trading costs - Add realistic commissions and slippage in backtests
4. Diversify risk - Don't bet all your capital on a single strategy
5. Monitor continuously - Track live performance and set stop-loss conditions

Common Mistakes

Mistake	Correct Approach
Going live because backtest returns look amazing	Check for overfitting
Ignoring trading costs	Add realistic commissions and slippage
Training on all the data	Split into train/validation/test sets
All-in on a single strategy	Diversify with a multi-strategy portfolio

Sharp Edges

SE-1: Backtest Overfitting

• Severity: critical
• Scenario: Strategy performs superbly in backtest (Sharpe > 3) but loses money live
• Cause: Parameters over-optimized to historical data, capturing noise instead of signal
• Symptoms: Backtest Sharpe > 3, live performance < 50%, parameter-sensitive, too few trades
• Detection: sharpe.*[3-9]\.|sharpe.*\d{2,}
• Fix: Use walk-forward validation, rolling tests across 5+ segments

SE-2: Look-ahead Bias

• Severity: critical
• Scenario: Inadvertently using future data in the backtest
• Cause: Time ordering not respected during data processing
• Symptoms: Perfect backtest returns, using same-day close as the same-day signal
• Detection: shift\(-|iloc\[-1\].*today
• Fix: Signals must be lagged by one period signal = prices.shift(1) > ma.shift(1)

SE-3: Survivorship Bias

• Severity: high
• Scenario: Backtesting using only currently-existing stocks
• Cause: Delisted/removed stocks are excluded, overstating strategy performance
• Symptoms: Backtest returns clearly higher than reality, small-cap strategies look especially good
• Fix: Use a complete database that includes delisted stocks; use point-in-time data

SE-4: Ignoring Trading Costs and Slippage

• Severity: high
• Scenario: High-turnover strategy looks very profitable
• Cause: Commissions and slippage not counted; real costs eat all the profit
• Symptoms: Annual turnover > 1000%, per-trade profit < 0.5%, backtest fills at the close
• Fix: Include commission (0.1%) + slippage (0.2%) = 0.3% per trade

SE-5: Overconfidence in In-Sample Performance

• Severity: medium
• Scenario: Deciding to go live based only on in-sample backtest results
• Cause: No independent test set was held out
• Symptoms: No out-of-sample test, validation set reused, test set too small
• Fix: Split data 60/20/20 (train/validation/test); use the test set only once

Risk Management Framework

Layer 1 – Trade level: 2% per-trade stop-loss, set take-profit based on volatility, trailing stop to protect profit Layer 2 – Strategy level: No single strategy over 20% of total capital, halve position when drawdown exceeds 15% Layer 3 – Portfolio level: Daily VaR no more than 2%, stress testing, keep 20% cash

Kelly Formula

f* = (p × b - q) / b
f* = optimal bet fraction, p = win rate, q = loss rate, b = odds
Example: win rate 55%, odds 1.5 → f* = 25% (in practice use Half-Kelly, 12.5%)

Execution Risk

Risk	Description	Countermeasure
Slippage	Fill price differs from expected	Limit orders, split execution
Liquidity	Can't fill enough size	Avoid small caps, set size caps
System failure	Program/network outage	Backup systems, manual monitoring
Black swan	Extreme events	Position caps, stop-loss mechanisms

Recommended Tools

• Backtrader - Python backtesting framework
• QuantConnect - Cloud quant platform
• TradingView - Chart analysis and strategy testing
• Zipline - Quantopian's open-source backtesting engine

Further Resources

• Code examples: extended/code-examples.md
• Templates and checklists: extended/templates.md

References

• Quantitative Trading - Ernest Chan
• Advances in Financial Machine Learning - Marcos Lopez de Prado

Disclaimer

Quantitative trading involves high risk; past performance does not indicate future results. This content is for educational reference only and does not constitute investment advice.