Our review
Detects and analyzes alpha decay in trading strategies using statistical indicators like Sharpe ratio decline, information coefficient drop, and hit rate degradation, providing severity warnings and recommendations.
Strengths
- Comprehensive decay indicators (Sharpe, IC, hit rate, capacity, regime).
- Customizable thresholds and analysis periods.
- Generates actionable recommendations such as reducing position sizing.
- Detailed mode for deeper analysis of time series and distribution shifts.
Limitations
- Relies on historical data and may not perfectly predict future decay.
- Requires integration with data sources for returns and strategy metadata.
- Interpretation of severity thresholds may vary by strategy context.
When monitoring active trading strategies for performance degradation and deciding whether to adjust or halt them.
When only qualitative assessment is needed or when historical data is insufficient or unreliable.
Security analysis
SafeThe skill performs analytical computations on strategy returns data within a Python environment. No destructive, exfiltrating, or system-modifying actions are described. No shell commands, network calls to external services without user intent, or obfuscated code are present.
No concerns found
Examples
/alpha-decay/alpha-decay --strategy momentum-001 --detailed/alpha-decay --threshold 0.5 --period 90name: alpha-decay description: Detect and analyze strategy alpha decay signals argument-hint: "[--strategy id|--all|--threshold pct|--period days]"
Alpha Decay Detection
Detect and analyze alpha decay in trading strategies using statistical methods.
Usage
/alpha-decay- Check all active strategies/alpha-decay --strategy momentum-001- Analyze specific strategy/alpha-decay --threshold 0.3- Custom decay threshold/alpha-decay --period 90- Analysis period in days/alpha-decay --detailed- Show detailed decay metrics
Decay Indicators
| Indicator | Description | Warning Level | |-----------|-------------|---------------| | Sharpe Decay | Rolling Sharpe ratio decline | > 30% decline | | IC Decay | Information coefficient drop | IC < 0.02 | | Hit Rate | Win rate degradation | < 45% | | Capacity | Returns vs AUM correlation | r < -0.3 | | Regime | Regime change detection | Confidence > 0.8 |
Related Files
scripts/risk_management/strategy_analytics.py- AlphaDecayDetector classscripts/risk_management/alpha_research.py- Signal evaluationservices/risk/risk_manager.py- Strategy monitoring
Instructions
When this skill is invoked:
-
Parse arguments:
- No args: Scan all active strategies
--strategy <id>: Single strategy analysis--threshold: Custom decay threshold (default 0.3)--period: Lookback period in days (default 60)
-
Load strategy data:
- Fetch returns from database/API
- Get strategy metadata and targets
- Load benchmark/factor returns
-
Run decay detection:
from risk_management.strategy_analytics import AlphaDecayDetector detector = AlphaDecayDetector( decay_threshold=0.3, confidence_level=0.95, lookback_window=60 ) signals = detector.detect_decay(returns) -
Display decay report:
Alpha Decay Analysis ═══════════════════════════════════════════════════════════ Strategy: momentum-001 Period: Last 60 days Status: ⚠️ WARNING - Decay signals detected DECAY SIGNALS ───────────────────────────────────────────────────────── Signal Severity Confidence Description ───────────────────────────────────────────────────────── Sharpe Decay 0.65 87% Sharpe dropped 42% IC Decay 0.45 72% IC now 0.015 (was 0.04) Hit Rate 0.30 65% Win rate 43% (target 52%) METRICS COMPARISON ───────────────────────────────────────────────────────── Metric Current Historical Change ───────────────────────────────────────────────────────── Sharpe Ratio 0.85 1.45 -41% IC Mean 0.015 0.042 -64% Hit Rate 43% 52% -17% Avg Return 0.02% 0.08% -75% RECOMMENDATIONS ───────────────────────────────────────────────────────── 1. Review regime indicators - potential regime change 2. Check for crowding in signal factors 3. Validate data inputs for drift 4. Consider reducing position sizing by 50% -
For
--detailed:- Rolling IC time series
- Distribution shift analysis
- Factor exposure changes
- Correlation regime changes
-
Severity thresholds:
- ACTIVE: No decay (severity < 0.3)
- WARNING: Moderate decay (0.3 <= severity < 0.6)
- CRITICAL: Severe decay (severity >= 0.6)
Prompt Engineering
Data & AI
Prompt engineering best practices and templates to maximize AI outputs.
Data Visualization
Data & AI
Generates data visualizations and charts tailored to your data.
RAG Architecture Setup
Data & AI
Setup guide for RAG (Retrieval-Augmented Generation) architectures.