EXPERIMENTAL_STRATEGY - Strategy Analysis
I. Strategy Overview
EXPERIMENTAL_STRATEGY is a typical multi-indicator combination strategy within the Freqtrade quantitative trading framework. It belongs to the hybrid category of trend-following and momentum reversal strategies. By comprehensively employing various classical technical analysis indicators, this strategy captures trading opportunities across different market environments, achieving dual coverage of trending markets and oversold rebound opportunities.
1.1 Strategy Positioning
This strategy is labeled "experimental," meaning it is not a fully validated production-grade strategy but serves as a technical demonstration and reference for strategy development. However, through in-depth analysis of the strategy code, its logic design demonstrates considerable professionalism and completeness, incorporating considerations across multiple dimensions: trend identification, momentum confirmation, and overbought/oversold recognition.
1.2 Core Design Philosophy
The core design philosophy can be summarized as "trend is king, momentum is secondary." In markets with a clear trend, the strategy tends to trade in the direction of the trend; when the market reaches extreme oversold or overbought conditions, the strategy seeks reversal opportunities. This dual-track design enables the strategy to adapt to different market cycles.
1.3 Applicable Scenarios
This strategy is primarily suited for the following trading scenarios:
- Cryptocurrency markets with noticeable medium-to-short-term trends
- Trading instruments with moderate but not excessive volatility
- 5-minute-level short-term trading timeframe
- Exchange environments that support limit order trading
II. Strategy Configuration Analysis
2.1 Timeframe Settings
The strategy adopts 5-minute candlesticks as the base timeframe. This setting positions the strategy for short-term trading, suitable for capturing intraday volatility opportunities while avoiding overly frequent trading signals.
The choice of 5-minute period is reasonably justified in the cryptocurrency market:
- Sufficient to capture short-term price fluctuations
- Relatively low signal noise
- Balance point between execution efficiency and trading costs
2.2 Take-Profit Settings (minimal_roi)
The strategy employs a phased take-profit mechanism with the following configuration:
| Holding Time | Take-Profit Target |
|---|---|
| 0 minutes | 4% |
| 20 minutes | 2% |
| 30 minutes | 1% |
| 40 minutes | 0% |
This decremental take-profit design reflects the strategy's time preference: pursuing quick profits while avoiding prolonged holding. Immediately after opening a position, a 4% return is demanded; as time passes, expectations gradually lower, ultimately closing unconditionally after 40 minutes. The advantages of this design include:
- Lock in quick gains: If the market moves favorably and quickly, timely profit-taking is achieved
- Control holding time: Avoid prolonged risk exposure in uncertain markets
- Progressive exit: Provide sufficient profit margin while setting a time floor
2.3 Stop-Loss Settings
The strategy sets a fixed -10% stop-loss, which is a relatively wide stop-loss level for a 5-minute trading framework. Under the 5-minute timeframe, a 10% stop-loss provides ample room. This setting's characteristics:
- Advantage: Allows sufficient price fluctuation space, avoiding being shaken out by normal volatility
- Risk: Individual trade losses may be significant
- Recommendation: Consider combining with position management to reduce per-trade risk in practical use
2.4 Order Type Configuration
order_types = {
'buy': 'limit',
'sell': 'limit',
'stoploss': 'limit',
'stoploss_on_exchange': False
}
The strategy exclusively uses limit orders for trading, reflecting a cautious execution approach:
- Buy limit orders: Ensure execution at preset or better prices
- Sell limit orders: Avoid slippage losses
- Stop-loss limit orders: Executed locally, not uploaded to the exchange
III. Entry Conditions Details
3.1 Overview of Technical Indicator System
3.1.1 Trend Strength Indicator: ADX (Average Directional Index)
ADX is the most core trend-judging indicator in the strategy. It does not indicate trend direction but measures the strength of the trend.
Indicator Logic:
- ADX value > 25: Indicates a clear trend exists
- ADX value > 50: Indicates a strong trend
- ADX value > 70: Indicates an extremely strong trend
Application in Strategy:
- Entry signals require ADX > 30, ensuring entry in a trending environment
- Another entry signal requires ADX > 65, capturing extremely strong market conditions
- Exit signals require ADX > 10, ensuring minimal trend confirmation
- Another exit signal requires ADX > 70, identifying extreme trend reversals
3.1.2 Direction Indicators: +DI/-DI (Positive/Negative Directional Indicators)
Direction indicators work in conjunction with ADX to determine the specific direction of a trend.
Indicator Logic:
- +DI > -DI: Bullish trend
- -DI > +DI: Bearish trend
- Larger DI values indicate more definitive directional clarity
Application in Strategy:
- Entry requires plus_di > 0.5, confirming bullish direction
- Exit requires minus_di > 0 or minus_di > 0.5, identifying bearish pressure
3.1.3 Momentum Indicator: RSI (Relative Strength Index)
RSI is a classic overbought/oversold indicator and plays an important role in the strategy.
Indicator Logic:
- RSI > 70: Overbought zone
- RSI < 30: Oversold zone
- RSI near 50: Balanced state
Application in Strategy:
- Entry conditions require RSI < 35, capturing oversold rebound opportunities
- Exit conditions detect RSI crossing above 70, identifying overbought reversals
3.1.4 Stochastic Indicator: Stochastic Fast
The strategy uses the fastd and fastk values of Stochastic Fast.
Indicator Logic:
- Value > 80: Overbought
- Value < 20: Oversold
- Crossovers of fastd and fastk can serve as signals
Application in Strategy:
- Entry requires fastd < 35, confirming oversold condition
- Exit detects fastd crossing above 70, confirming overbought reversal
3.1.5 Trend Confirmation Indicator: MACD
MACD is an important trend confirmation indicator in the strategy, comprising three lines:
- MACD Line: Fast line (12-period EMA - 26-period EMA)
- Signal Line: Slow line (9-period EMA of MACD)
- Histogram: Difference between MACD and signal line
Application in Strategy: Although the strategy calculates the complete MACD indicator, it is not directly used for signal generation in the current version, possibly for future optimization or as auxiliary reference.
3.1.6 Volatility Indicator: Bollinger Bands
The strategy uses standard Bollinger Bands settings (20-period, 2x standard deviation).
Indicator Logic:
- Upper, middle, and lower bands form a price channel
- Price touching upper/lower bands may signal overbought/oversold conditions
- Bandwidth narrowing foreshadows potential breakouts
Application in Strategy: Similarly, Bollinger Bands are calculated but not used for signal generation in the current version, possibly reserved for future optimization.
3.1.7 Moving Averages: EMA and SMA
The strategy calculates two types of moving averages:
- EMA(10): Fast exponential moving average
- SMA(40): Slow simple moving average
Potential Uses:
- Golden cross/death cross signals
- Trend direction judgment
- Dynamic support/resistance levels
3.2 Entry Signal 1: Oversold Rebound Mode
(
(dataframe['rsi'] < 35) &
(dataframe['fastd'] < 35) &
(dataframe['adx'] > 30) &
(dataframe['plus_di'] > 0.5)
)
Condition Breakdown:
| Condition | Meaning | Design Intent |
|---|---|---|
| RSI < 35 | Relative strength in oversold zone | Price may be excessively sold |
| fastd < 35 | Stochastic confirms oversold | Multiple oversold confirmations |
| ADX > 30 | Sufficient trend strength | Avoid catching a falling knife in a trendless market |
| plus_di > 0.5 | Bullish direction has foundation | Ensure correct rebound direction |
Trading Logic Analysis:
This is a typical counter-trend bottom-catching signal, but not blind bottom-catching. Through the combination of four conditions, the strategy achieves the design concept of "finding oversold opportunities within a trend."
- RSI < 35 and fastd < 35: Two independent indicators simultaneously confirm oversold status, greatly reducing false signal probability
- ADX > 30: This condition is crucial — it ensures bottom-catching occurs in an environment with trend support, not catching a falling knife in a downtrend
- plus_di > 0.5: Ensures bullish momentum exists, avoiding premature entry in a pure bearish market
Applicable Scenarios:
- Pullback lows in an uptrend
- Lower boundary areas in ranging markets
- Not suitable for so-called "low points" in a downtrend
3.3 Entry Signal 2: Strong Trend Momentum Chase Mode
(
(dataframe['adx'] > 65) &
(dataframe['plus_di'] > 0.5)
)
Condition Breakdown:
| Condition | Meaning | Design Intent |
|---|---|---|
| ADX > 65 | Extremely strong trend | Capture major moves |
| plus_di > 0.5 | Bullish direction confirmed | Follow the trend |
Trading Logic Analysis:
This is a pure trend-following signal. When the market exhibits extremely strong momentum, the strategy enters directly without waiting for pullbacks. The logic behind this design:
- ADX > 65 indicates trend strength has reached extreme levels; at this point, market momentum is strong
- plus_di > 0.5 confirms the trend direction is bullish, avoiding counter-trend trading
- No price conditions set: Acknowledging "the trend is your friend," accepting current-price entry
Applicable Scenarios:
- Surging markets triggered by sudden positive news
- Acceleration phase of a long-term uptrend
- Trending moves in high-volatility markets
3.4 Synergy Between the Two Entry Signals
These two entry signals form a complementary relationship:
| Comparison Dimension | Oversold Rebound Signal | Strong Trend Signal |
|---|---|---|
| Entry position | Relatively low | Any position |
| Market state | Oversold reversal | Strong trend continuation |
| Risk characteristics | Moderate bottom-catching risk | Higher chase risk |
| Profit potential | Rebound gains | Trend continuation gains |
IV. Exit Conditions Details
4.1 Exit Signal 1: Overbought Reversal Mode
(
(
(qtpylib.crossed_above(dataframe['rsi'], 70)) |
(qtpylib.crossed_above(dataframe['fastd'], 70))
) &
(dataframe['adx'] > 10) &
(dataframe['minus_di'] > 0)
)
Condition Breakdown:
| Condition | Meaning | Design Intent |
|---|---|---|
| RSI crosses above 70 | Entering overbought zone | Price may be too high |
| fastd crosses above 70 | Stochastic overbought confirmation | Double confirmation (OR relationship) |
| ADX > 10 | Basic trend exists | Avoid noise in trendless markets |
| minus_di > 0 | Bearish pressure appears | Trend may reverse |
Trading Logic Analysis:
This is a profit-taking signal with carefully considered logic:
- RSI or fastd crosses above 70: Uses an "OR" relationship — either indicator triggering is sufficient, improving signal sensitivity
- ADX > 10: Loose trend confirmation, avoiding frequent trading in extreme volatility
- minus_di > 0: Key condition, ensuring bearish momentum exists, increasing reversal credibility
Signal Characteristics:
- Uses crossover signals rather than static thresholds, capturing reversal points more timely
- Multiple condition combinations reduce false signals
- minus_di condition adds trend reversal confirmation
4.2 Exit Signal 2: Extreme Trend Reversal Mode
(
(dataframe['adx'] > 70) &
(dataframe['minus_di'] > 0.5)
)
Condition Breakdown:
| Condition | Meaning | Design Intent |
|---|---|---|
| ADX > 70 | Extremely strong trend | Trend may be overheated |
| minus_di > 0.5 | Strong bearish pressure | Reversal signal appears |
Trading Logic Analysis:
This is an extreme market exit signal, based on the market law that "extremes must reverse":
- ADX > 70 indicates trend strength has reached extreme levels; historical experience shows extreme strength is often followed by adjustments
- minus_di > 0.5 confirms bearish momentum has accumulated, increasing probability of trend reversal
- Timely profit-taking: Exit at the peak of trend madness, avoiding greed
Applicable Scenarios:
- Profit-taking after consecutive sharp rallies
- Risk avoidance at top areas
- Forms a corresponding relationship with Entry Signal 2
4.3 Logical Symmetry Between Exit and Entry Signals
The strategy design demonstrates excellent logical symmetry:
| Entry Signal | Corresponding Exit Signal | Design Concept |
|---|---|---|
| Oversold rebound (ADX>30, plus_di>0.5) | Overbought reversal (ADX>10, minus_di>0) | Buy oversold, sell overbought |
| Strong trend chase (ADX>65, plus_di>0.5) | Extreme reversal (ADX>70, minus_di>0.5) | Enter on trend, exit at extreme |
V. Risk Management
5.1 Built-in Risk Control Mechanisms
The strategy controls risk through multi-layered mechanisms:
1. Indicator Combination Filtering
- Single indicator signals are insufficient to trigger trades
- Multi-indicator cross-confirmation reduces false signals
- ADX filtering ensures trading in trending environments
2. Phased Take-Profit
- Decremental take-profit design controls holding time
- 40-minute mandatory close avoids overnight risk
- High profit targets trigger rapid profit-taking
3. Directional Confirmation
- plus_di/minus_di ensure trend direction
- Avoid counter-trend trading
- Reduce trend judgment errors
5.2 Potential Risk Points
1. Chase Risk
- Entry Signal 2 enters at any price, possibly buying at highs
- Recommendation: Add price position judgment, such as distance from moving average
2. Wide Stop-Loss
- 10% stop-loss is relatively wide on a 5-minute timeframe
- Recommendation: Combine with position management or reduce stop-loss magnitude
3. Oversold Bottom-Catching Risk
- Oversold conditions may persist longer
- Recommendation: Add trend confirmation or use time-based stop-loss
5.3 Improvement Suggestions
Technical Level:
- Add volume confirmation indicators
- Utilize Bollinger Bands for price channel reference
- Use MACD to add trend confirmation dimension
- Add EMA/SMA golden cross/death cross signals
Parameter Level:
- Optimize indicator thresholds through backtesting
- Adjust stop-loss/take-profit ratios
- Adjust timeframe based on market characteristics
Execution Level:
- Add market state filtering (bull/bear market identification)
- Consider trading time period filtering
- Add position management logic
VI. Strategy Pros & Cons
6.1 Strategy Advantages
- Multi-Dimensional Confirmation: Comprehensively employs various technical indicators, reducing false signals
- Dual-Track Signals: Balances oversold rebound and trend chase, adapting to different markets
- Trend Filtering: ADX-centric design avoids trading in trendless environments
- Controlled Risk: Phased take-profit + fixed stop-loss provides bounded risk
- Logical Symmetry: Buy and sell signal designs are symmetric, easy to understand and optimize
6.2 Strategy Limitations
- Hard-Coded Parameters: Lacks flexibility, inconvenient for optimization
- Wide Stop-Loss: 10% stop-loss on 5-minute timeframe is relatively wide
- Indicator Redundancy: Some calculated indicators are unused
- Chase Risk: Entry Signal 2 may enter at highs
- No Position Management: Lacks dynamic position adjustment
VII. Applicable Scenarios
7.1 Market Environment Suitability
| Market State | Suitability | Reason |
|---|---|---|
| Clear uptrend | ★★★★★ | Entry Signal 2 captures perfectly |
| Ranging upward | ★★★★☆ | Oversold rebound signals effective |
| Sideways range | ★★★☆☆ | ADX filtering may reduce signals |
| Ranging downward | ★★☆☆☆ | Bottom-catching signals higher risk |
| Clear downtrend | ★☆☆☆☆ | Not recommended |
7.2 Trading Instrument Suitability
Suitable Instruments:
- Mainstream cryptocurrencies (BTC, ETH)
- Trading pairs with good liquidity
- Instruments with moderate volatility
Unsuitable Instruments:
- Extremely low-liquidity instruments
- Stablecoin trading pairs
- Highly manipulated small-cap coins
VIII. Parameter Optimization
8.1 Backtesting Configuration Suggestions
Data Requirements:
- Time span: At least 6 months or more
- Should include bull and bear cycles
- Multi-coin combination testing
Parameter Optimization:
- Use Freqtrade Hyperopt functionality
- Focus on RSI, ADX threshold combinations
- Test different stop-loss/take-profit ratios
Evaluation Metrics:
- Annualized return
- Maximum drawdown
- Sharpe ratio
- Profit/loss ratio
- Win rate
8.2 Live Trading Suggestions
Phased Deployment:
- Paper trading test for at least 2 weeks
- Small-capital live trading test
- Gradually increase capital scale
Monitoring Metrics:
- Deviation between actual execution price and signal price
- Signal trigger frequency
- Holding time distribution
- Profit/loss trade analysis
IX. Live Trading Notes
9.1 For Beginners
- Use this strategy as a template for learning strategy development
- Understand the design philosophy of multi-indicator combinations
- Focus on ADX application in trend judgment
9.2 For Advanced Users
- Conduct sufficient backtesting validation
- Adjust parameters based on market characteristics
- Add risk control and position management logic
- Optimize using unused indicators
9.3 For Professional Traders
- Use this strategy as part of a combined strategy
- Coordinate or hedge with other strategies
- Dynamically switch parameters based on market cycles
- Combine with fundamental analysis for trading decisions
X. Summary
EXPERIMENTAL_STRATEGY is an experimental quantitative trading strategy that demonstrates considerable professionalism in its design despite the "experimental" label. Its multi-indicator combination, dual-track signal design, ADX-centric trend filtering, and multi-layered risk control mechanisms make it a strategy worth studying and extending.
However, users should note that this strategy carries inherent risks, particularly parameter sensitivity, wide stop-loss settings, and trend-chasing risks. Before live deployment, thorough backtesting and simulated trading validation are essential.
Disclaimer: This strategy is experimental in nature and is provided for learning and research purposes only. Live trading involves risk. Please make decisions cautiously based on your own risk tolerance. Past performance does not guarantee future returns.
Document Version: v1.0 Last Updated: 2024