Lesson 13: Strategy Scoring System
⏱ Duration: 1.5 hours
🎯 Learning Objectives: Establish scientific strategy evaluation criteria
📚 Difficulty: ⭐⭐⭐ Strategy Optimization
📖 Course Overview
Facing numerous strategies, how can you objectively evaluate their strengths and weaknesses? This lesson will teach you to build a scientific strategy scoring system, comprehensively evaluate strategies through quantitative indicators, and help you select truly outstanding strategies.
13.1 Return Metrics (40% Weight)
Returns are the most intuitive performance of a strategy, but should not be the only criterion.
1. Total Return
Definition: Cumulative return over the entire backtesting period.
Scoring Criteria (30-day backtest):
Excellent: > 15% → 10 points
Good: 10-15% → 8 points
Pass: 5-10% → 6 points
Poor: 0-5% → 4 points
Loss: < 0% → 0 points
Weight: 15%
Calculation Example:
# Strategy A: Total return 18%
score = 10 × 0.15 = 1.5 points
# Strategy B: Total return 8%
score = 6 × 0.15 = 0.9 points
2. Annualized Return
Definition: Convert backtesting returns to annualized returns.
Calculation Formula:
Annualized Return = (1 + Total Return) ^ (365 / Backtesting Days) - 1
Example:
30-day backtest, 15% return:
Annualized Return = (1.15) ^ (365/30) - 1 = 7.43 = 743%
90-day backtest, 15% return:
Annualized Return = (1.15) ^ (365/90) - 1 = 0.68 = 68%
Scoring Criteria:
Excellent: > 100% → 10 points
Good: 50-100% → 8 points
Pass: 20-50% → 6 points
Poor: 0-20% → 4 points
Loss: < 0% → 0 points
Weight: 10%
3. Average Profit per Trade
Definition: Average profit per trade.
Scoring Criteria:
Excellent: > 1.5% → 10 points
Good: 1.0-1.5% → 8 points
Pass: 0.5-1.0% → 6 points
Poor: 0.2-0.5% → 4 points
Very Poor: < 0.2% → 2 points
Weight: 15%
Importance:
- Better reflects strategy quality than total return
- High average profit = efficient strategy
- Needs to be evaluated in combination with trade frequency
13.2 Risk Metrics (30% Weight)
Controlling risk is key to long-term profitability.
1. Maximum Drawdown
Definition: Maximum decline from fund's highest point to lowest point.
Scoring Criteria:
Excellent: < 5% → 10 points
Good: 5-10% → 8 points
Pass: 10-15% → 6 points
Poor: 15-20% → 4 points
Dangerous: > 20% → 0 points
Weight: 15%
Why is it important?
10% drawdown → Need 11.1% gain to break even
20% drawdown → Need 25% gain to break even
50% drawdown → Need 100% gain to break even
2. Sharpe Ratio
Definition: Risk-adjusted return.
Scoring Criteria:
Excellent: > 3.0 → 10 points
Good: 2.0-3.0 → 8 points
Pass: 1.0-2.0 → 6 points
Poor: 0.5-1.0 → 4 points
Very Poor: < 0.5 → 2 points
Negative: < 0 → 0 points
Weight: 10%
3. Calmar Ratio
Definition: Annualized Return / Maximum Drawdown.
Calculation Formula:
Calmar Ratio = Annualized Return / |Maximum Drawdown|
Example:
Strategy A: 60% annualized return, 10% drawdown
Calmar = 60 / 10 = 6.0 ✅ Excellent
Strategy B: 80% annualized return, 25% drawdown
Calmar = 80 / 25 = 3.2 ⚠️ Average
Scoring Criteria:
Excellent: > 5.0 → 10 points
Good: 3.0-5.0 → 8 points
Pass: 2.0-3.0 → 6 points
Poor: 1.0-2.0 → 4 points
Very Poor: < 1.0 → 2 points
Weight: 5%
13.3 Stability Metrics (20% Weight)
Stability determines whether a strategy can sustain long-term profitability.
1. Win Rate
Definition: Percentage of profitable trades to total trades.
Scoring Criteria:
Excellent: > 80% → 10 points
Good: 70-80% → 8 points
Pass: 60-70% → 6 points
Poor: 50-60% → 4 points
Very Poor: < 50% → 2 points
Weight: 8%
Note:
- Win rate is not always better higher
- Needs to be evaluated in combination with risk-reward ratio
- 40% win rate + 3:1 risk-reward ratio > 80% win rate + 1:1 risk-reward ratio
2. Profit Factor
Definition: Total profit / Total loss.
Calculation Formula:
Profit Factor = Total Profit Amount / |Total Loss Amount|
Example:
Total profit: $1,200
Total loss: $400
Profit Factor = 1,200 / 400 = 3.0
Scoring Criteria:
Excellent: > 2.5 → 10 points
Good: 2.0-2.5 → 8 points
Pass: 1.5-2.0 → 6 points
Poor: 1.2-1.5 → 4 points
Very Poor: < 1.2 → 2 points
Loss: < 1.0 → 0 points
Weight: 7%
3. Return Standard Deviation (Consistency)
Definition: Degree of fluctuation in quarterly returns.
Calculation Method:
# Assume 4 quarters of returns
returns = [8%, 12%, 9%, 11%]
std_dev = np.std(returns) = 1.58%
# Smaller standard deviation = more stable
Scoring Criteria:
Excellent: < 3% → 10 points
Good: 3-5% → 8 points
Pass: 5-8% → 6 points
Poor: 8-12% → 4 points
Very Poor: > 12% → 2 points
Weight: 5%
13.4 Practicality Metrics (10% Weight)
Practicality determines whether a strategy is suitable for live trading.
1. Trade Frequency
Scoring Criteria (30-day backtest):
Ideal: 20-60 trades → 10 points
Acceptable: 10-20 trades → 8 points
Few: 5-10 trades → 6 points
Too few: < 5 trades → 4 points
Too many: > 100 trades → 4 points
Ultra-high: > 150 trades → 2 points
Weight: 4%
Reasons:
- Too few: Insufficient samples, unreliable
- Too many: Fees erode profits, overtrading
2. Average Holding Time
Scoring Criteria:
Ideal: 2-24 hours → 10 points
Acceptable: 1-2 hours → 8 points
Very short: < 1 hour → 6 points (high-frequency risk)
Very long: > 48 hours → 6 points (overnight risk)
Weight: 3%
3. Out-of-Sample Performance
Definition: Test period performance / Training period performance.
Scoring Criteria:
Excellent: > 90% → 10 points
Good: 75-90% → 8 points
Pass: 60-75% → 6 points
Poor: 40-60% → 4 points
Overfit: < 40% → 0 points
Weight: 3%
Calculation Example:
Training period return: 20%
Test period return: 17%
Out-of-sample score = 17 / 20 = 85% → 8 points
13.5 Comprehensive Scoring System
Scoring Formula
Total Score = (
# Return metrics (40%)
Total Return Score × 0.15 +
Annualized Return Score × 0.10 +
Average Profit Score × 0.15 +
# Risk metrics (30%)
Max Drawdown Score × 0.15 +
Sharpe Score × 0.10 +
Calmar Score × 0.05 +
# Stability metrics (20%)
Win Rate Score × 0.08 +
Profit Factor Score × 0.07 +
Return Std Dev Score × 0.05 +
# Practicality metrics (10%)
Trade Frequency Score × 0.04 +
Holding Time Score × 0.03 +
Out-of-Sample Score × 0.03
)
Strategy Scorecard Template
| Metric Category | Metric | Actual Value | Individual Score | Weight | Weighted Score |
|---|---|---|---|---|---|
| Return | Total Return % | 18% | 10 | 15% | 1.50 |
| Annualized Return % | 120% | 10 | 10% | 1.00 | |
| Average Profit % | 1.2% | 8 | 15% | 1.20 | |
| Risk | Max Drawdown % | 6% | 8 | 15% | 1.20 |
| Sharpe Ratio | 2.8 | 8 | 10% | 0.80 | |
| Calmar Ratio | 4.5 | 8 | 5% | 0.40 | |
| Stability | Win Rate % | 78% | 8 | 8% | 0.64 |
| Profit Factor | 2.3 | 8 | 7% | 0.56 | |
| Return Std Dev | 4% | 8 | 5% | 0.40 | |
| Practicality | Trade Frequency | 42 trades | 10 | 4% | 0.40 |
| Holding Time | 8 hours | 10 | 3% | 0.30 | |
| Out-of-Sample % | 82% | 8 | 3% | 0.24 | |
| Total | 100% | 8.64 |
Rating Standards
⭐⭐⭐⭐⭐ S Grade (9.0-10.0): Top-tier strategy, highly recommended for live trading
⭐⭐⭐⭐ A Grade (8.0-9.0): Excellent strategy, recommended for live trading
⭐⭐⭐ B Grade (7.0-8.0): Good strategy, can be used for live trading
⭐⭐ C Grade (6.0-7.0): Passing strategy, use cautiously for live trading
⭐ D Grade (5.0-6.0): Poor strategy, needs optimization
❌ F Grade (< 5.0): Unqualified, not recommended for use
Actual Case Scoring
Case 1: MomentumTrendStrategy
| Metric | Value | Score | Weighted |
|---|---|---|---|
| Total Return | 23.5% | 10 | 1.50 |
| Annualized Return | 153% | 10 | 1.00 |
| Average Profit | 1.52% | 10 | 1.50 |
| Max Drawdown | 3.2% | 10 | 1.50 |
| Sharpe | 3.5 | 10 | 1.00 |
| Calmar | 8.2 | 10 | 0.50 |
| Win Rate | 89.7% | 10 | 0.80 |
| Profit Factor | 2.8 | 10 | 0.70 |
| Return Std Dev | 2.5% | 10 | 0.50 |
| Trade Frequency | 68 trades | 8 | 0.32 |
| Holding Time | 6 hours | 10 | 0.30 |
| Out-of-Sample | 88% | 8 | 0.24 |
| Total | 9.36 ⭐⭐⭐⭐⭐ |
Rating: S Grade - Top-tier Strategy
Comments:
- ✅ Excellent returns (23.5%)
- ✅ Extremely low risk (3.2% drawdown)
- ✅ Outstanding stability (90% win rate)
- ✅ Good out-of-sample validation
- 💡 Highly recommended for live trading
Case 2: Strategy001
| Metric | Value | Score | Weighted |
|---|---|---|---|
| Total Return | 15.3% | 10 | 1.50 |
| Annualized Return | 92% | 8 | 0.80 |
| Average Profit | 0.85% | 6 | 0.90 |
| Max Drawdown | 6.8% | 8 | 1.20 |
| Sharpe | 2.1 | 8 | 0.80 |
| Calmar | 3.5 | 8 | 0.40 |
| Win Rate | 72.4% | 8 | 0.64 |
| Profit Factor | 1.8 | 6 | 0.42 |
| Return Std Dev | 5.5% | 6 | 0.30 |
| Trade Frequency | 127 trades | 6 | 0.24 |
| Holding Time | 4 hours | 10 | 0.30 |
| Out-of-Sample | 72% | 6 | 0.18 |
| Total | 7.68 ⭐⭐⭐ |
Rating: B Grade - Good Strategy
Comments:
- ✅ Good returns (15.3%)
- ✅ Controllable risk (6.8% drawdown)
- ⚠️ Slightly high trading frequency (127 trades)
- ⚠️ Average out-of-sample performance (72%)
- 💡 Can be used for live trading but needs close monitoring
💡 Practical Tasks
Task 1: Create Strategy Scorecards
Select 3 strategies and create complete scorecards for each.
Steps:
- Backtest 3 strategies (30-day data)
- Record all metric data
- Use scoring formula to calculate total score
- Assign grade
Scorecard Template (Excel or Google Sheets):
Download Scorecard Template (needs to be created yourself)
Task 2: Compare Different Weighting Schemes
Try adjusting weights and observe impact on rankings:
Scheme A: Return Priority
- Return metrics: 50%
- Risk metrics: 20%
- Stability: 20%
- Practicality: 10%
Scheme B: Risk Priority
- Return metrics: 25%
- Risk metrics: 45%
- Stability: 20%
- Practicality: 10%
Scheme C: Balanced (Recommended)
- Return metrics: 40%
- Risk metrics: 30%
- Stability: 20%
- Practicality: 10%
Compare whether your strategy rankings change under the three schemes.
Task 3: Create Strategy Recommendation List
Based on scoring results, create recommendation list:
# My Strategy Recommendation List
## 🥇 S Grade Strategies (9.0-10.0 points)
1. **MomentumTrendStrategy** - 9.36 points
- Advantages: High returns, low risk, good stability
- Disadvantages: Slightly high trading frequency
- Recommended position: 30%
## 🥈 A Grade Strategies (8.0-9.0 points)
(Fill in your strategies)
## 🥉 B Grade Strategies (7.0-8.0 points)
(Fill in your strategies)
## ⚠️ Strategies Needing Optimization
(Fill in strategies needing improvement)
📚 Python Automated Scoring Script
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Strategy Automated Scoring Tool
"""
import numpy as np
class StrategyScorer:
def __init__(self):
self.weights = {
'total_return': 0.15,
'annual_return': 0.10,
'avg_profit': 0.15,
'max_drawdown': 0.15,
'sharpe': 0.10,
'calmar': 0.05,
'win_rate': 0.08,
'profit_factor': 0.07,
'consistency': 0.05,
'trade_freq': 0.04,
'holding_time': 0.03,
'out_of_sample': 0.03
}
def score_total_return(self, value):
"""Total return scoring"""
if value > 15: return 10
if value > 10: return 8
if value > 5: return 6
if value > 0: return 4
return 0
def score_max_drawdown(self, value):
"""Maximum drawdown scoring (negative value)"""
value = abs(value)
if value < 5: return 10
if value < 10: return 8
if value < 15: return 6
if value < 20: return 4
return 0
def score_sharpe(self, value):
"""Sharpe Ratio scoring"""
if value > 3.0: return 10
if value > 2.0: return 8
if value > 1.0: return 6
if value > 0.5: return 4
if value > 0: return 2
return 0
def score_win_rate(self, value):
"""Win rate scoring"""
if value > 80: return 10
if value > 70: return 8
if value > 60: return 6
if value > 50: return 4
return 2
def calculate_total_score(self, metrics):
"""Calculate total score"""
scores = {
'total_return': self.score_total_return(metrics['total_return']),
'max_drawdown': self.score_max_drawdown(metrics['max_drawdown']),
'sharpe': self.score_sharpe(metrics['sharpe']),
'win_rate': self.score_win_rate(metrics['win_rate']),
# Add scoring for other metrics...
}
total = 0
for key, score in scores.items():
if key in self.weights:
total += score * self.weights[key]
return total
def get_grade(self, score):
"""Get grade"""
if score >= 9.0: return 'S'
if score >= 8.0: return 'A'
if score >= 7.0: return 'B'
if score >= 6.0: return 'C'
if score >= 5.0: return 'D'
return 'F'
# Usage example
scorer = StrategyScorer()
strategy_metrics = {
'total_return': 18.5,
'max_drawdown': -6.2,
'sharpe': 2.8,
'win_rate': 75.5,
# ... other metrics
}
score = scorer.calculate_total_score(strategy_metrics)
grade = scorer.get_grade(score)
print(f"Strategy Score: {score:.2f}")
print(f"Strategy Grade: {grade} Grade")
📌 Key Points Summary
- Multi-dimensional evaluation: Don't just look at returns, risk and stability are equally important
- Weight allocation: Returns 40%, Risk 30%, Stability 20%, Practicality 10%
- Out-of-sample validation: Must validate in test period, avoid overfitting
- Comprehensive rating: Only scores above 9 are top-tier strategies
- Dynamic adjustment: Adjust weights according to personal risk preference
- Continuous monitoring: Scoring is not one-time, needs regular re-evaluation
➡️ Next Lesson Preview
Lesson 14: Risk Management and Capital Management
In the next lesson, we will learn:
- Position management strategies
- Stop loss and take profit settings
- Capital allocation methods
- Risk control mechanisms
Preparation:
- ✅ Complete scoring for at least 3 strategies
- ✅ Select 1-2 high-scoring strategies
- ✅ Prepare demo trading account
🎯 Learning Assessment Criteria:
- ✅ Understand various indicators in strategy scoring system
- ✅ Can independently score and rate strategies
- ✅ Can select appropriate strategies based on scoring
- ✅ Understand impact of different weighting schemes
After completing these tasks, you have acquired the ability to scientifically evaluate strategies! 🎯
Top comments (0)