Overview

This strategy combines the Voss Predictive Filter and Ehlers Instantaneous Trendline indicator to identify cyclical turning points in the market for quantitative trading. The Voss filter provides early buy/sell signals, while the trendline indicator determines the overall trend direction to avoid misleads from the Voss filter in trending markets. This strategy works well on instruments like Bitcoin that exhibit cyclical patterns, as evidenced by good backtest results.

Strategy Logic

Voss Predictive Filter

The Voss filter comes from John F. Ehlers' article "A Peek Into The Future". Its calculation is as follows:

            pine
            
            
_filt = 0.5 * _s3 * _x1 + _f1 * _s2 * _filt[1] - _s1 * _filt[2] 
_voss = _x2 * _filt - _sumC

Where _x1 is the 1st order price difference; _x2 is a smoothing factor; _s1, _s2, _s3 are filter parameters; _f1 is the cycle parameter; _filt is the filter output; _voss is the final output.

The filter can be seen as a smoothed filter that emphasizes current and past cycle information to generate early signals. Due to inherent group delays, it "looks into the future" to produce predictive signals ahead of other indicators.

Instantaneous Trendline Indicator

The trendline indicator is calculated as:

            pine
            
            
_it = (_a-((_a*_a)/4.0))*_src+0.5*_a*_a*_src[1]-(_a-0.75*_a*_a)*_src[2]+2*(1-_a)*nz(_it[1])+-(1-_a)*(1-_a)*nz(_it[2]) 

It plots a trendline in real-time that closely fits the price action, accurately determining trend direction and strength.

Strategy Logic

A buy signal is generated when the Voss crosses up through the filter result.

A sell signal is generated when the Voss crosses down through the filter result.

Trading signals are only accepted if confirmed by the trendline indicator. This avoids wrong signals from the Voss filter in trending markets.

Advantages

• Voss filter provides early predictive signals to catch cyclical turns
• Trendline indicator accurately determines trend direction, avoiding wrong signals
• Parameters can be optimized for different cycles and market environments
• Stops can be added to control risk

Risks and Mitigation

• Strategy relies on early signals, potentially missing some trends
• Counter-trend signals may occur in strong trends, causing losses

Risks can be reduced by:

• Optimizing period parameter for instrument cycles
• Reducing filter bandwidth to decrease false signals
• Adding trend filters to avoid signals against strong trends
• Using stops to control loss per trade

Enhancement Opportunities

The strategy can be improved by:

• Trying different price sources like close, moving averages etc.
• Adjusting filter period for specific instrument cycles
• Optimizing trend indicator parameters for better trend detection
• Trying different trend indicators for better combinations
• Adding stops, trailing stops for better risk control
• Parameter optimization for best parameter sets

Conclusion

This strategy combines the Voss filter and trend indicator to effectively identify cyclical turns in the market. With optimized parameters and risk controls, it can produce a robust quantitative trading system. It is widely applicable to instruments exhibiting cyclical patterns, as evidenced by good backtest results. Overall, the strategy has unique predictive capabilities, and broad potential for enhancement through multi-dimensional optimization.