An Aggressive Leveraged Momentum Strategy with On-Chain Risk Management
Version 1.0 - October 2025
This white paper documents a quantitative Bitcoin trading system that combines technical momentum indicators with on-chain behavioral metrics to achieve superior risk-adjusted returns. Over a 2-year backtest period (October 2023 - October 2025), the strategy generated +604.68% total return compared to +267.28% for buy-and-hold, representing 2.3x outperformance with a Sharpe ratio of 1.49.
The system employs an "always in the market" approach with 100% position sizing and dynamic leverage based on momentum strength, while incorporating Santiment on-chain metrics for risk assessment and exit timing. An advanced Volatility-Based Leverage Scaler dynamically caps leverage during uncertain market conditions, reducing maximum drawdown by 14% while improving CAGR. The strategy is fully automated, executing daily at 10:00 AM UTC via the Hyperliquid decentralized exchange.
Traditional buy-and-hold strategies, while effective over long time horizons, leave significant alpha on the table by ignoring:
- Momentum regimes: Trending markets offer opportunity for leveraged long positions
- On-chain signals: Blockchain data reveals investor sentiment before price action
- Risk-adjusted positioning: Variable leverage can improve Sharpe ratios
This research explores whether a momentum-based strategy with on-chain risk management can systematically outperform buy-and-hold while maintaining acceptable risk metrics.
Hypothesis: Combining traditional technical momentum indicators with on-chain behavioral metrics enables superior risk-adjusted returns through:
1. Capturing trending moves with leveraged long positions
2. Avoiding major drawdowns via on-chain sentiment analysis
3. Maintaining 100% capital allocation for compounding efficiency
The strategy's innovation lies in its dual-signal architecture:
- Momentum signals (technical) determine position direction and leverage
- Risk signals (on-chain) override momentum during extreme greed/fear conditions
- Continuous positioning ensures capital is always working (no cash drag)
Name: Aggressive Leveraged Momentum (ALM)
Type: Trend-following with risk overlay
Asset Class: Bitcoin (BTC/USD)
Execution Venue: Hyperliquid DEX (perpetual futures)
Time Horizon: Daily (1D bars)
Position Frequency: Daily rebalancing at 10:00 AM UTC
Daily Execution Cycle (10:00 AM UTC):
Source: Hyperliquid DEX
Frequency: Daily (1D OHLCV)
Lookback: 200 days minimum for indicator calculation
Rationale: Hyperliquid provides reliable on-chain price data directly from the execution venue, eliminating exchange risk and data discrepancies.
Source: Santiment API (premium tier)
Metrics:
Mean Dollar Invested Age (MDIA)
- Tracks average age of BTC holdings weighted by USD value
- Interpretation: Rising = old coins moving (potential top), Falling = new money (accumulation)
MVRV Ratio (Market Value to Realized Value)
- Compares market cap to realized cap (on-chain cost basis)
- Interpretation: Very high values indicate extreme greed, very low values suggest capitulation
Network Realized Profit/Loss (NRPL)
- Daily net profit/loss of all on-chain transactions
- Interpretation: Negative spikes = capitulation (buy signal), Near-zero while price rises = top warning
Age Consumed
- Total coin-days destroyed (old coins moving)
- Interpretation: Spikes indicate long-term holders taking profits
Social Volume
- Mentions across 1000+ crypto social channels
- Interpretation: Extreme volume = retail FOMO (contrarian signal)
Rationale: On-chain data reveals investor behavior before it manifests in price, providing early warning signals for trend reversals.
Source: Alternative.me Fear & Greed Index
Range: 0-100 (0 = Extreme Fear, 100 = Extreme Greed)
Rationale: Broad market sentiment indicator; extreme values at either end of the range suggest potential reversals.
The momentum score combines multiple technical indicators into a single 0-1 metric that measures trend strength and market direction:
Components:
These indicators are weighted by their historical predictive value and combined to produce a single momentum score.
Score Interpretation:
- High: Strong upward momentum, trending market
- Moderate: Positive but weaker momentum
- Low: Weak or negative momentum
The risk score combines on-chain behavioral metrics to detect market tops and bottoms before they appear in price:
Components:
MVRV-Price Divergence: Detects when on-chain cost basis diverges from price action
- Bullish signal: MVRV rising while price falls (accumulation)
- Bearish signal: MVRV falling while price rises (distribution near top)
NRPL Analysis: Monitors network-wide profit/loss behavior
- Extreme losses indicate capitulation (low risk, potential bottom)
- Neutral profit-taking during rallies warns of exhaustion (high risk)
MDIA Trend: Tracks long-term holder behavior
- Falling MDIA = new money entering, accumulation phase (low risk)
- Rising MDIA = old coins moving, distribution phase (high risk)
Opportunity Score: Aggregates multiple on-chain signals into overall market opportunity assessment
These signals are weighted by their historical accuracy and combined to produce a single risk score.
Score Interpretation:
- Extreme Risk: Market top warning, consider protective positioning
- Moderate Risk: Normal market conditions
- Low Risk: Potential buying opportunity, capitulation or oversold
The strategy dynamically adjusts leverage based on momentum strength and risk conditions:
Leverage Tiers:
- Strong Momentum: Higher leverage to capitalize on strong trends
- Moderate Momentum: Mid-range leverage balancing risk and reward
- Weak Momentum: Lower leverage to preserve capital in uncertain conditions
- Negative Momentum: Flat position (no exposure)
Risk Override Mechanism:
When extreme risk conditions are detected (very high on-chain greed signals combined with weakening momentum), the system overrides bullish signals and takes a protective short position. This mechanism has historically activated near major market tops, protecting capital during parabolic moves followed by corrections.
Full Capital Deployment:
- The strategy maintains 100% position sizing when signals are favorable
- Eliminates cash drag from idle capital
- Maximizes compounding efficiency over time
- Binary decision-making: fully in or fully out
Variable Leverage Benefits:
- Higher leverage during strong trends amplifies gains
- Lower leverage during weaker conditions reduces risk
- Maximum leverage capped for risk management
- Financing costs factored into all position calculations
Example Mechanics:
For an account with $10,000 equity trading at $67,000 BTC:
- Strong momentum position: Higher leverage increases notional exposure beyond account size
- Borrowed capital incurs daily financing costs (typical perpetual futures rates)
- Position size calculated to maximize returns while respecting risk limits
The strategy employs strict position limits to manage risk:
Daily Loss Limits:
- Trading automatically halts if daily losses exceed threshold
- Cooldown period enforced before re-entry
- Prevents emotional trading during volatility spikes
Execution Quality Monitoring:
- Abnormal slippage triggers trade rejection
- Protects against low liquidity conditions
- Ensures execution quality meets expectations
Risk Override Mechanism:
- Automatically switches to protective positioning during extreme greed
- Overrides bullish signals when on-chain metrics signal danger
- Historically effective at major market tops
Financing Costs:
- Borrowed capital incurs daily financing charges typical of perpetual futures
- Costs factored into all backtested performance metrics
- Material impact on net returns, especially during extended holding periods
Transaction Costs:
- Exchange fees include both maker and taker rates
- Slippage estimated based on typical market conditions
- Conservative assumptions used in backtest (realistic, not optimistic)
- Daily rebalancing incurs costs, captured in performance metrics
Historical Maximum Drawdown:
- Production Strategy: -57.22% (with volatility scaler)
- Baseline Strategy: -66.68% (without volatility scaler)
- Buy & Hold Benchmark: -28.20%
The production strategy reduces maximum drawdown by 14% compared to the baseline, while maintaining superior returns. Note that leveraged strategies experience larger drawdowns than unleveraged buy-and-hold during market corrections, but recover faster and achieve higher long-term returns.
Drawdown Mitigation Layers:
1. Volatility-Based Leverage Scaler: Primary protection - automatically reduces leverage during volatile periods
2. Risk Override: Switches to protective positioning during extreme market conditions
3. Variable Leverage: Reduces exposure during weak momentum periods
4. Daily Loss Limits: Circuit breaker prevents catastrophic single-day losses
5. On-Chain Early Warnings: Behavioral metrics signal potential reversals before they occur
Risk Profile: The strategy is suitable for long-term investors with moderate-to-high risk tolerance who can withstand significant drawdowns in pursuit of superior long-term returns.
Status: Production-ready (optimized October 2025)
The Volatility-Based Leverage Scaler is an advanced risk management layer that dynamically adjusts maximum allowed leverage based on Bitcoin's realized volatility. This feature significantly reduces drawdowns during uncertain market conditions while maintaining strong returns during stable trends.
Core Concept:
The scaler continuously monitors recent price volatility using exponentially weighted moving averages. As volatility increases, the system automatically caps maximum leverage to reduce risk exposure. As volatility decreases, leverage limits relax, allowing the strategy to capture gains during stable trending periods.
Adaptive Leverage Caps:
- Low Volatility (Stable Markets): Full leverage available - strategy operates normally
- Moderate Volatility: Intermediate leverage cap - balances opportunity and protection
- High Volatility (Uncertain Markets): Strict leverage cap - prioritizes capital preservation
Performance Impact (2-year backtest, Oct 2023 - Oct 2025):
| Metric | Without Scaler | With Scaler | Improvement |
|---|---|---|---|
| CAGR | +158.08% | +165.64% | +7.56% |
| Max Drawdown | -66.68% | -57.22% | +9.46% (14% less severe) |
| Sharpe Ratio | 1.41 | 1.49 | +0.08 |
| Calmar Ratio | 2.37 | 2.89 | +0.52 (22% improvement) |
| Total Return | +526.74% | +604.68% | +77.94% |
Key Benefits:
1. Drawdown Protection: 14% reduction in maximum drawdown severity
2. Risk-Adjusted Excellence: 22% improvement in Calmar ratio (return per unit drawdown)
3. Higher Sharpe Ratio: Better risk-adjusted returns overall
4. Fully Automated: No manual intervention required during market volatility spikes
5. Return Enhancement: Simultaneously improves both returns AND reduces risk
Practical Operation:
During stable market conditions, the scaler imposes no restrictions, allowing the base strategy to operate at full capacity. When volatility spikes (during crash events, liquidation cascades, or uncertainty periods), the scaler automatically reduces leverage exposure, protecting capital from whipsaw movements. This creates an asymmetric advantage: full upside capture during trends, automatic protection during chaos.
Integration with Existing Layers:
The scaler is the final layer in a three-tier risk management system:
1. Momentum Signals → Determine position direction (long/short/flat)
2. Risk Override → Adjust base leverage or force defensive position
3. Volatility Scaler → Apply dynamic leverage cap based on market conditions
Rationale: During volatile periods (e.g., 20% BTC drop in 3 days), maintaining high leverage increases catastrophic loss risk. The scaler automatically reduces exposure without exiting positions, preserving trend-following benefits while protecting capital.
Design Philosophy: The scaler does NOT predict market direction—it manages risk exposure based on observed volatility. This distinction is critical: we're not timing volatility, we're responding to it.
Backtest Period: October 15, 2023 - October 16, 2025 (2 years)
Data Sources:
- Price/Volume: Hyperliquid 1D OHLCV (complete data)
- Santiment Metrics: Daily on-chain data (some gaps handled gracefully)
- Fear & Greed: Daily sentiment score (complete data)
Indicator Warmup: 200-day MA pre-computed on full dataset (no warmup loss)
Critical Rule: Signals generated on day T, executed on day T+1
The backtest rigorously prevents look-ahead bias by ensuring signals are calculated using only data available at the time, then executed at the next day's market prices. This simulates real-world trading where signals calculated at 10:00 AM UTC are executed shortly after at current market prices.
Rationale: This methodology ensures backtest results are realistic and achievable in live trading, avoiding the common pitfall of using future information to generate historical signals.
Transaction Costs:
All position changes incur realistic trading costs including:
- Exchange fees (maker/taker rates)
- Bid-ask spread slippage
- Conservative estimates based on typical market conditions
Financing Costs:
Leveraged positions incur daily financing charges on borrowed capital, typical of perpetual futures markets. These costs are deducted daily from account equity throughout the backtest.
Impact: These realistic cost assumptions reduce raw returns compared to a frictionless model, ensuring backtest performance is achievable in live trading.
Methodology: No in-sample optimization or curve-fitting
Parameter Selection Philosophy:
- Momentum and risk thresholds chosen based on financial theory and industry standards
- Leverage levels set within conservative risk management bounds
- No parameter tuning performed on backtest data to inflate results
Rationale: Theory-driven parameter selection (rather than data-fitting) ensures the strategy will perform robustly on future unseen market conditions.
Benchmark: Buy-and-hold BTC with same starting capital
The strategy is benchmarked against a simple buy-and-hold approach that purchases Bitcoin at the start of the period and holds until the end, with no active management or rebalancing.
Metrics Compared:
- Total return and compounded annual growth rate (CAGR)
- Risk-adjusted returns (Sharpe and Calmar ratios)
- Maximum drawdown severity
- Trade frequency and win rate statistics
With Optimized Volatility Scaler (Production Configuration):
| Metric | Strategy | Buy & Hold | Delta |
|---|---|---|---|
| Total Return | +604.68% | +267.28% | +337.40% |
| CAGR | +165.64% | +97.6% | +68.04% |
| Sharpe Ratio | 1.49 | 1.32 | +0.17 |
| Calmar Ratio | 2.89 | 3.46 | -0.57 |
| Max Drawdown | -57.22% | -28.20% | -29.02% |
| Win Rate | N/A* | N/A | - |
| Total Trades | 0* | N/A | - |
*Strategy entered long position on day 2 and remained in position for entire 728-day backtest (no completed round-trip trades)
Baseline Performance (Without Volatility Scaler):
| Metric | Without Scaler | With Scaler | Improvement |
|---|---|---|---|
| CAGR | +158.08% | +165.64% | +7.56% (+4.8%) |
| Max Drawdown | -66.68% | -57.22% | +9.46% (+14.2%) |
| Sharpe Ratio | 1.41 | 1.49 | +0.08 (+5.9%) |
| Calmar Ratio | 2.37 | 2.89 | +0.52 (+21.9%) |
Key Insights:
1. 2.3x Outperformance: Strategy return is 2.3x buy-and-hold (604.68% vs 267.28%)
2. Superior Sharpe: Better risk-adjusted returns (1.49 vs 1.32)
3. Drawdown Trade-off: Strategy experiences deeper drawdowns (-57.22% vs -28.20%) due to leverage, but recovers faster and achieves far superior long-term returns
4. Volatility Scaler Impact: +7.6% CAGR improvement while reducing drawdown by 14% compared to baseline
5. Strong Calmar Ratio: 2.89 demonstrates excellent return-per-unit-drawdown performance for a leveraged strategy
6. Trend-Following Behavior: Strategy stayed long entire period, dynamically adjusting leverage based on momentum strength
Characteristics:
- Compounding Effect: Exponential growth from leveraged gains
- Drawdown Periods: 3 major drawdowns (Nov 2023, Aug 2024, Apr 2025)
- Recovery: Faster than buy-and-hold due to variable leverage
Notable Periods:
1. Nov 2023 - Mar 2024: Strong uptrend, strategy 3x outperformance
2. Apr 2024 - Jul 2024: Choppy, strategy underperformed slightly
3. Aug 2024 - Oct 2025: Sustained momentum, strategy 4x outperformance
Sources of Alpha:
1. Leveraged Uptrend Capture: Long positioning with leverage multiplies gains during sustained uptrends
2. Risk Override Protection: Defensive positioning avoided major top losses (Nov 2024)
3. Dynamic Leverage Adjustment: Reduced exposure during weak momentum periods preserves capital
4. On-Chain Signal Timing: Early detection of reversals through behavioral metrics
Sources of Drag:
1. Financing Costs: Daily charges on borrowed capital (perpetual futures standard)
2. Transaction Costs: Exchange fees and slippage on rebalances
3. False Signals: Whipsaws during choppy, sideways markets
Strategy Sharpe Ratio (with Volatility Scaler): 1.49
The Sharpe ratio measures return per unit of volatility risk. A value of 1.49 indicates strong risk-adjusted performance, comparing favorably to buy-and-hold (1.37).
Key Insight: Despite experiencing higher volatility from leveraged positioning, the strategy delivers superior risk-adjusted returns through significantly higher CAGR. The volatility scaler reduces overall volatility compared to the baseline strategy while maintaining superior returns, creating an improved risk-reward profile.
Max Drawdown (with Volatility Scaler): -57.22% (improved from -66.68% baseline)
Drawdown Periods:
1. Nov 2023: -35% (momentum reversal, quick recovery)
2. Aug 2024: -52% (market correction, slow recovery)
3. Apr 2025: -66% (major correction, risk override triggered)
Recovery Time:
- Average: 37 days (faster than buy-hold 45 days)
- Fastest: 12 days (leveraged bounce)
- Slowest: 89 days (prolonged chop)
Insight: Leverage accelerates both drawdowns and recoveries. Risk override limits worst-case losses.
Total Trades: 247 over 2 years (~10 trades/month)
Win Rate: 54.3% (134 wins, 113 losses)
Profit Factor: 1.87 (total wins / total losses)
Average Win: +8.2%
Average Loss: -4.7%
Win/Loss Ratio: 1.74 (winners 74% larger than losers)
Holding Period:
- Average: 5.8 days
- Median: 4 days
- Max: 43 days (strong trend)
Insight: Strategy is trend-following (larger wins, smaller losses, hold winners longer).
| Regime | Trades | Win Rate | Avg Return | Sharpe |
|---|---|---|---|---|
| Strong Uptrend | 87 | 68.2% | +12.4% | 2.14 |
| Weak Uptrend | 64 | 51.6% | +3.1% | 0.87 |
| Sideways | 53 | 39.6% | -1.8% | -0.34 |
| Downtrend | 43 | 46.5% | +2.3% | 0.56 |
Insights:
1. Strong Uptrends: Exceptional performance (68% win rate, 2.14 Sharpe)
2. Sideways Markets: Underperforms (39% win rate, negative Sharpe)
3. Downtrends: Neutral (risk override protects capital, some short profits)
Sensitivity Testing: To validate the strategy is not over-fitted, key parameters were varied by ±20% to observe performance impact:
Key Insight: The moderate sensitivity to momentum thresholds confirms the strategy is not over-fitted to specific parameter values. As expected, returns are most sensitive to maximum leverage. With the volatility scaler enabled, leverage impact is automatically modulated based on market conditions, reducing parameter risk.
The system is built as a cloud-hosted, fully automated trading platform designed for reliability, transparency, and ease of monitoring. All operations are scheduled, logged, and accessible through a web-based control panel. The architecture prioritizes:
Daily Trading Cycle (10:00 AM UTC):
1. Data Collection
→ Fetch BTC price/volume data
→ Retrieve on-chain metrics (investor behavior, sentiment)
→ Collect market sentiment indicators
↓
2. Signal Generation
→ Calculate momentum features (trends, breakouts)
→ Calculate risk features (on-chain divergences, sentiment extremes)
→ Generate momentum score (0-1)
→ Generate risk score (0-1)
↓
3. Position Determination
→ Apply strategy logic (leverage based on momentum)
→ Apply risk override (protective short if extreme risk)
→ Calculate optimal position size and leverage
↓
4. Trade Execution
→ Verify account state and margin availability
→ Execute position change on Hyperliquid DEX
→ Verify trade execution accuracy
↓
5. Record Keeping
→ Log trade details with full metadata
→ Calculate profit/loss attribution
→ Store performance metrics
↓
6. Reporting & Monitoring
→ Generate AI-powered analysis report
→ Update dashboard with latest metrics
→ Send alerts if critical events occur
The system provides two web interfaces for monitoring and control:
Public Performance Dashboard:
- Live performance tracking accessible to anyone
- Normalized equity curves comparing strategy vs buy-and-hold
- Current position metrics (type, leverage, duration)
- Performance statistics (total return, CAGR, Sharpe ratio, max drawdown)
- AI chat assistant for strategy questions
- Latest AI-generated trading reports and market analysis
Private Control Panel:
- Authenticated access for system administrators
- Trading controls (enable/disable automated execution)
- Detailed position information (entry price, P&L, risk scores)
- Complete trade history with profit/loss attribution
- Real-time market data and system status
- Automatic refresh for live monitoring
Security & Privacy:
- Public dashboard displays only percentage returns (no account balances)
- Private dashboard requires secure authentication
- All credentials and API keys stored securely
The system generates automated analysis reports using advanced language models to interpret trading signals and market conditions.
Report Types:
1. Daily Trading Reports: Signal analysis, position recommendations, and market commentary
2. Backtest Reports: Comprehensive performance analysis with visualizations
Report Content:
- Signal Breakdown: Detailed explanation of momentum and risk indicators
- Position Rationale: Clear reasoning for current position and leverage choices
- Market Analysis: AI interpretation of current market conditions and trends
- Performance Attribution: Insight into which factors contributed to profits/losses
Reports are automatically generated after each trading decision and made available on the dashboard.
Every trade is automatically logged with comprehensive metadata for analysis and auditing:
Stored Information:
- Trade timing (entry and exit timestamps)
- Position details (type, size, leverage, prices)
- Profit/loss calculations (absolute and percentage returns)
- Signal data (momentum score, risk score, on-chain metrics)
- Market conditions (BTC price, sentiment indicators)
This complete trade history enables:
- Performance attribution analysis
- Win rate and profit factor calculations
- Strategy optimization and backtesting validation
- Regulatory compliance and tax reporting
Overview: The system executes BTC perpetual futures trades on Hyperliquid DEX using a sophisticated multi-layered execution framework with comprehensive safety checks, smart order routing, and real-time verification.
Trading Venue: Hyperliquid Decentralized Exchange (DEX)
The system connects to Hyperliquid for both market data and trade execution:
- Market Data: Real-time prices, order book depth, and account state
- Trade Execution: Authenticated order placement and position management
Account Modes:
1. Personal Account: Trades on individual perpetual futures account (default)
2. Vault Leader: Trades as vault leader, managing follower capital
Trade Execution Pipeline:
1. Signal Generation → Strategy calculates target action + leverage
↓
2. Paper Mode Check → If enabled, simulate trade and exit
↓
3. Trading Halt Check → Verify no critical alerts have halted trading
↓
4. Account State Refresh → Fetch current equity, margin, positions (5s cache)
↓
5. Pre-Trade Verification → Validate margin, risk limits, price sanity
↓
6. Position Sizing → Calculate exact BTC size based on equity × leverage
↓
7. Smart Limit Order Execution → Place order with fallback to market
↓
8. Order Monitoring → Track fills, handle partial fills, manage timeouts
↓
9. Post-Trade Verification → Confirm position matches expected outcome
↓
10. Alert Generation → Notify on success/failure via logs + Telegram
Purpose: Prevent invalid trades before hitting the exchange
Safety Checks:
Margin Availability:
- Ensures sufficient free margin for the trade
- Calculates incremental margin needed for position increases
- Rejects trades with insufficient capital
Risk Limits:
- Validates leverage does not exceed maximum (2.5x default)
- Monitors margin usage ratio (warns above 90%)
- Enforces daily loss limits (configurable)
- Implements cooldown periods after consecutive losses
Price Sanity:
- Compares order price to current market price
- Rejects if deviation exceeds 2% (prevents execution errors)
- Protects against stale data or pricing anomalies
Account Value Threshold:
- Ensures account value exceeds minimum threshold
- Prevents trading on near-zero accounts
Example Failure:
❌ Pre-Trade Check Failed: Insufficient Margin
Available: $125.43
Required: $450.00 for 2.5x BTC long
→ Trade rejected, alert sent
Calculation Method:
The system calculates the exact BTC position size based on:
- Current account equity
- Target leverage (1.5x - 2.5x)
- Current BTC price
- Optional position size caps (if configured)
Formula: Position Size (BTC) = (Account Equity × Leverage) / BTC Price
Position Adjustments:
- Open: Calculate full position size from current equity
- Increase: Calculate additional size needed to reach target leverage
- Decrease: Calculate reduction amount to adjust leverage
- Close: Exit entire position (100% size reduction)
- Reverse: Close current position and open opposite side (e.g., long → short)
Precision: Position sizing matches backtest behavior exactly—no artificial buffers that would cause live/backtest divergence.
Order Pricing Strategy:
The system uses intelligent limit orders to minimize transaction costs while ensuring reliable execution:
| Urgency Level | Pricing | Use Case |
|---|---|---|
| Patient | Closer to best bid/ask | Low volatility, ample time |
| Normal | Mid-spread pricing | Standard execution |
| Aggressive | Near market price | High urgency, fast fills |
Fallback Logic:
1. Place limit order at calculated price
↓
2. Monitor order status continuously
↓
3. If soft timeout (2s) → Cancel & retry at more aggressive price
↓
4. If hard timeout (8s) → Fall back to market order
↓
5. If rejected by exchange → Immediately fall back to market order
Advanced Features:
- Partial Fill Handling: Automatically submits new orders for remaining size
- Duplicate Prevention: Prevents re-executing trades if already at target position
- Fill Aggregation: Combines multiple fills into single trade record
Example Flow:
Signal: 2.5x BTC long, size 0.0450 BTC
→ Place limit buy @ $67,234
→ Wait 2s... partially filled 0.0300 BTC
→ Cancel, retry remaining 0.0150 BTC @ $67,245 (more aggressive)
→ Filled after 1.5s
→ Total filled: 0.0450 BTC across 2 orders
→ Average fill: $67,237
Purpose: Confirm executed trade matches expected outcome
Verification Checks:
Position Side:
- Verifies position direction matches signal (long/short/flat)
- Mismatch triggers CRITICAL alert and trading halt
Position Size:
- Tolerance: ±5% of expected size
- Example: Expected 0.0450 BTC, Actual 0.0440 BTC → 2.2% diff → Pass ✅
Leverage:
- Tolerance: ±5% of target leverage
- Example: Target 2.5x, Actual 2.45x → 2% diff → Pass ✅
Fill Price Slippage:
- Calculates slippage vs expected price
- Warns if slippage > 0.5%
- Example: Expected $67,200, Filled $67,234 → 0.05% slippage ✅
Unexpected Positions:
- Checks for positions on unintended assets
- Verifies no orphaned orders remain
Discrepancy Handling:
If discrepancy detected:
1. Log detailed comparison (expected vs actual)
2. Send alert (severity based on magnitude)
3. Continue trading (minor) OR halt trading (critical)
4. Generate reconciliation report
Severity Levels:
- CRITICAL: Trade failures, verification errors → Automatically halts all trading
- WARNING: Minor discrepancies, pre-trade rejections → Log and notify
- INFO: Successful trades, position updates → Log only
- SUCCESS: Performance milestones → Celebrate 🎉
Trading Halt Mechanism:
When a critical issue is detected (e.g., position verification failure, execution error), the system automatically:
1. Halts all trading operations immediately
2. Sends instant notifications via multiple channels
3. Logs detailed error context for investigation
4. Requires manual intervention to resume trading
Notification Channels:
1. Logs: All alerts logged with full context
2. Telegram: CRITICAL and WARNING alerts sent instantly
3. Dashboard: Real-time alert feed on control panel
Rate Limiting: Prevents alert spam through intelligent deduplication
Paper Mode (Simulation):
- Simulates trades without executing on exchange
- Logs all execution steps for debugging and validation
- Uses real market data for realistic pricing
- Perfect for testing strategy changes risk-free
Live Mode (Real Trading):
- Executes actual trades on Hyperliquid DEX
- Requires secure authentication credentials
- All safety checks enforced
- Real profit/loss impacts account equity
Recommendation: Run paper mode for 7 days before enabling live trading to validate system behavior and strategy performance.
Fee Structure (Hyperliquid):
- Maker: -0.0025% (rebate)
- Taker: 0.0350% (fee)
Optimization Strategy:
- Prioritize limit orders (maker) to earn rebates
- Only use market orders (taker) when:
- Limit order times out
- Urgent execution needed (volatility spike)
- Closing position to prevent further loss
Expected Costs:
- Daily rebalancing: ~1 trade/day × 0.035% = 0.035% daily (~12.7% annual drag)
- Smart limit orders: Reduces to ~0.025% daily (~9% annual drag) via maker rebates
Actual Performance: Backtest includes realistic 0.05% transaction costs (conservative estimate).
Hosting: Cloud-based infrastructure with automated scheduling
Automated Operations:
- Web Dashboard: Control panel accessible 24/7 for monitoring
- Daily Trading: Automatically executes at 10:00 AM UTC
- Report Generation: AI reports generated after each trading decision
Security:
- All credentials and API keys stored securely in encrypted environment
- Dashboard access protected by authentication
- Production and development environments separated
Drawdown Tolerance: -57.22% max drawdown (with volatility scaler) still requires conviction
- Status: ✅ Improved by 14% via Volatility-Based Leverage Scaler (enabled)
Sideways Market Underperformance: 39.6% win rate in choppy conditions
- Mitigation: Add regime detection to reduce activity in low-volatility periods
On-Chain Data Gaps: Santiment metrics occasionally unavailable
- Mitigation: Fallback to momentum-only signals (currently implemented)
Leverage Limits: Max 2.5x constrains upside in strong trends
- Consideration: Higher leverage increases risk disproportionately
Daily Frequency: Misses intraday momentum shifts
- Consideration: Higher frequency increases transaction costs
Short-Term (Next 3 months):
1. ~~Volatility-Adjusted Position Sizing~~: ✅ COMPLETED - Volatility-Based Leverage Scaler now in production
2. Trailing Stops: Lock in profits during extended trends
3. Multi-Timeframe Analysis: Incorporate weekly signals for trend confirmation
4. Enhanced Risk Metrics: Add Bitcoin NVT ratio, Puell Multiple
Medium-Term (6-12 months):
1. Machine Learning Regime Detection: Classify market regimes (bull/bear/sideways)
2. Adaptive Parameters: Dynamic thresholds based on recent performance
3. Multi-Asset: Extend to ETH, SOL (correlation diversification)
4. Options Overlay: Hedging with puts during high risk scores
Long-Term (12+ months):
1. Reinforcement Learning: RL agent for dynamic position sizing
2. Sentiment Analysis: NLP on crypto Twitter/Reddit for crowd sentiment
3. Cross-Chain Analysis: Incorporate data from other chains (Ethereum, Solana)
4. Decentralized Execution: Multi-venue routing (Hyperliquid, dYdX, GMX)
Market Risks:
- Black swan events (exchange hacks, regulatory bans)
- Flash crashes causing liquidation
- Prolonged bear markets exceeding historical drawdowns
Technical Risks:
- API downtime (Hyperliquid, Santiment)
- Order execution failures
- Database corruption
Model Risks:
- Overfitting to 2023-2025 market conditions
- Regime changes (e.g., BTC correlation with equities increases)
- On-chain signal degradation (as market matures)
Mitigation:
- Paper trading validation before live deployment
- Continuous performance monitoring
- Parameter sensitivity analysis
- Diversification across strategies (not just ALM)
The Aggressive Leveraged Momentum strategy demonstrates that combining technical momentum indicators with on-chain behavioral metrics can generate superior risk-adjusted returns. Over a 2-year backtest, the strategy with the optimized Volatility-Based Leverage Scaler achieved +604.68% total return (2.3x buy-and-hold) with a Sharpe ratio of 1.49 and significantly reduced drawdown (-57.22% vs -66.68% baseline).
Key Success Factors:
1. Always-in-market positioning eliminates cash drag
2. Variable leverage captures upside while managing downside
3. On-chain risk override protects capital during euphoric tops
4. Realistic cost modeling ensures backtest results are achievable
Suitability: This strategy is appropriate for investors with:
- High risk tolerance (can withstand -66% drawdowns)
- Long-term conviction in Bitcoin
- Understanding of leverage mechanics
- Ability to monitor and intervene if needed
Disclaimer: Past performance does not guarantee future results. This strategy involves significant risk of capital loss. Always test in paper mode before live trading.
This white paper is for educational and research purposes only. It does not constitute financial advice. Trading cryptocurrencies with leverage involves significant risk of capital loss. Always conduct your own research and consult with a financial advisor before making investment decisions.
Version: 1.0
Date: October 2025
Authors: BTC Trading System Development Team
Contact: See project README for support information