Understanding Market Regimes indicators in StrategyQuant Coding Base

Over the past few months, we have added several technical indicators to our coding base. When selecting these indicators, we focused on ensuring that they help users better identify market regimes.

In the first part of this article, we will briefly explain what market regimes are. In the second part, we will discuss how to configure each of these indicators and how to effectively use them during strategy development.

What Are Market Regimes?

Market regimes are distinct periods where markets exhibit specific statistical properties. Beyond simple “bull” and “bear” classifications, markets operate in multiple states:

• Trending markets with persistent directional movement

• Range-bound markets where prices oscillate within boundaries

• High-volatility periods featuring large price swings

• Low-volatility periods with minimal price movement

• Correlation shifts between related instruments

• Liquidity variations affecting execution quality

A critical insight for StrategyQuant users: strategies optimized for one regime typically underperform when conditions change. This explains why a profitable strategy can suddenly start losing without any code changes.

Why Your Strategies Suddenly Stop Working

Strategy failure usually isn’t due to flawed logic but to undetected market regime changes. Consider these common scenarios:

• Your trend-following system that performed beautifully for months suddenly accumulates losses when markets become choppy

• Your breakout strategy that captured major moves now faces repeated false signals

• Your mean-reversion setup that thrived in range-bound conditions gets steamrolled when strong trends emerge

Standard indicators like moving averages, RSI, or Bollinger Bands typically don’t detect these fundamental shifts until significant losses occur. They track symptoms rather than identifying underlying statistical changes in market behavior.

​Enhancing StrategyQuant with Advanced Statistical Regime Detection

StrategyQuant platform users gain access to four advanced statistical indicators specifically designed for market regime detection:

• Kolmogorov-Smirnov Test – Detects statistically significant changes in price distribution

• Wasserstein Distance – Measures the magnitude of distribution shifts

• CUSUM (Cumulative Sum) – Identifies subtle, persistent changes in market behavior

• DTW (Dynamic Time Warping) – Recognizes similar pattern formations regardless of timeframe stretching

These tools evaluate fundamental aspects of market behavior that conventional indicators miss:

• Changes in return distributions

• Shifts in temporal structure

• Pattern similarities with historical periods

• Persistence of deviations from established norms

Practical Applications in StrategyQuant

For StrategyQuant users, these indicators can help you in theese situations:

• Strategy Switching – Create custom blocks in your strategies that activate different trading logic based on detected regimes

• Parameter Adaptation – Automatically adjust stop-loss, take-profit, and position sizing based on current market conditions

• Portfolio Allocation – Shift capital between strategies optimized for different regimes

• Robust Backtesting – Test strategies across multiple detected regimes to ensure consistent performance

• Forward Testing Validation – Compare current market conditions with historical regimes to better validate forward testing results

The following sections will show you exactly how to configure each indicator within StrategyQuant to detect regime changes before they impact your trading performance.

Market Regime Indicators in StrategyQuant Coding Base

In the following section, we will look at how to properly set up and use each indicator.

Setting Up the Kolmogorov-Smirnov Test Indicator

The KS Test compares recent price action against historical distributions to detect statistically significant changes. Here’s how to set it up effectively:

Key Parameters

Period1 (Recent Sample Size): This defines how many recent bars will be analyzed.

• Start with 20-30 bars for daily charts
• Use 50-100 bars for intraday charts
• Larger values (50-100) provide more statistical confidence but respond more slowly to changes

Period2 (Historical Sample Size): This determines the historical reference period.

• Generally set equal to or larger than Period1
• A good starting point is setting Period2 = Period1
• For more sensitivity to recent changes, make Period2 larger (e.g., 2-3x Period1)

SignalThreshold: The statistical significance level.

• The default value of 0.05 (5%) is standard in statistics
• Lower values (0.01) reduce false signals but might miss some regime changes
• Higher values (0.10) increase sensitivity but generate more false signals

Practical Setup Examples

Conservative Setup:

• Period1 = 50
• Period2 = 100
• SignalThreshold = 0.01
• Best for: Weekly or daily charts, lower frequency trading

Balanced Setup:

• Period1 = 30
• Period2 = 50
• SignalThreshold = 0.05
• Best for: Daily charts, swing trading

Responsive Setup:

• Period1 = 20
• Period2 = 30
• SignalThreshold = 0.10
• Best for: Intraday charts, more frequent signals

Interpretation Guidelines

• When the indicator value is 1, it suggests a statistically significant regime change
• Look for persistent signals (multiple consecutive 1s) rather than isolated instances
• Combine with other indicators to confirm the change

​You can download the indicator and custom blocks qui.

Setting Up the Wasserstein Distance Indicator

Wasserstein Distance measures the dissimilarity between recent and historical price distributions on a continuous scale.

Key Parameters

Periodo: This defines the window size for comparison.

• Start with 20-50 bars for most timeframes
• Shorter periods (20-30) react more quickly to distribution changes
• Longer periods (50-100) provide more stable readings but lag in detection

Practical Setup Examples

Fast-Response Setup:

• Period = 20
• Best for: Catching quick regime transitions, intraday trading

Standard Setup:

• Period = 50
• Best for: Daily charts, balanced between responsiveness and stability

Trend-Focused Setup:

• Period = 100
• Best for: Identifying major regime shifts while filtering noise, weekly charts

Interpretation Guidelines

• The indicator provides values on a 0-100 scale
• Typically, values below 20 indicate similar distributions (same regime)
• Values of 40+ suggest significant distribution differences
• Watch for sustained increases in the indicator value rather than brief spikes
• Establish baseline readings during known market regimes for your specific instrument

​You can download the indicator and custom blocks qui.

Setting Up the CUSUM Indicator

CUSUM detects subtle, persistent changes in price behavior by accumulating deviations.

Key Parameters

Periodo: The lookback period for calculating mean and standard deviation.

• Start with 20-50 bars
• Shorter periods are more responsive to recent changes
• Longer periods create more stable reference statistics

Soglia: Controls sensitivity to deviations.

• Default value of 2 corresponds to 2 standard deviations
• Lower values (1-1.5) increase sensitivity and generate more signals
• Higher values (2.5-3) reduce false alarms but might delay detection

Drift: Accounts for expected natural drift in the process.

• Default value of 0.5 is suitable for most markets
• Increase drift (0.7-1.0) in more volatile markets
• Decrease drift (0.2-0.3) in less volatile, range-bound markets

Practical Setup Examples

Early Warning Setup:

• Period = 20
• Threshold = 1.5
• Drift = 0.3
• Best for: Getting ahead of regime changes, accepting some false positives

Balanced Setup:

• Period = 50
• Threshold = 2
• Drift = 0.5
• Best for: Most market conditions, daily timeframes

Confirmation Setup:

• Period = 100
• Threshold = 3
• Drift = 0.5
• Best for: Confirming established regime changes, reducing false signals

Interpretation Guidelines

• Monitor both Positive and Negative CUSUM lines
• When either line rises significantly above zero, it indicates a potential regime change
• Positive CUSUM rising suggests upward pressure/regime shift
• Negative CUSUM rising suggests downward pressure/regime shift
• The longer a CUSUM line stays elevated, the stronger the signal
• Reset your expectations when both lines return to near zero

​You can download the indicator and custom blocks qui.

Setting Up the Dynamic Time Warping Indicator

DTW identifies similar patterns in price action regardless of timeframe stretching or compression.

Key Parameters

WindowSize: The lookback period to search for patterns.

• Start with 20-50 bars
• Larger windows (50+) allow for finding more distant patterns
• Smaller windows focus on recent market behavior

PatternSize: The length of the pattern to match.

• Start with 5-10 bars
• Shorter patterns (3-5) identify micro-structures
• Longer patterns (10-20) identify larger market formations

DistanceType: The method used to calculate differences.

• Absolute (0): More robust to outliers, recommended for most cases
• Squared (1): More sensitive to large deviations, useful for detecting volatility regime changes

Practical Setup Examples

Micro-Pattern Setup:

• WindowSize = 20
• PatternSize = 5
• DistanceType = 0 (Absolute)
• Best for: Short-term trading, identifying quick repeating patterns

Standard Setup:

• WindowSize = 30
• PatternSize = 10
• DistanceType = 0 (Absolute)
• Best for: Most trading timeframes and instruments

Macro-Pattern Setup:

• WindowSize = 50
• PatternSize = 15
• DistanceType = 1 (Squared)
• Best for: Longer-term analysis, identifying major market structures

Interpretation Guidelines

• Lower DTW values indicate similarity to historical patterns (regime continuity)
• Sudden increases suggest unfamiliar price action (potential regime change)
• Look for unusually low values to identify strongly repeating patterns
• Establish baseline readings specific to your market and timeframe

​You can download the indicator and custom blocks qui.

Combining Indicators for a Complete Regime Detection System

For the most robust regime detection, consider these multi-indicator approaches:

Early Warning System

• Start with CUSUM (sensitive setup) for first alert
• When CUSUM signals, check Wasserstein Distance for confirmation
• Use KS Test as final statistical validation

Strength-of-Change Measurement

• Use KS Test to determine if a statistically significant change has occurred
• Measure Wasserstein Distance to quantify how different the new regime is
• Monitor DTW to identify if the new pattern resembles any historical regimes

Timeframe Integration

• Apply indicators across multiple timeframes
• Look for confluence of signals (e.g., KS Test signaling on both daily and weekly)
• Short-term indicators can provide early warnings for longer-term regime shifts

Practical Considerations for All Indicators

Market-Specific Adjustments

Azioni: Generally less volatile; use more sensitive settings

• Reduce KS Test threshold to 0.03-0.04
• Reduce CUSUM threshold to 1.5-2.0

Forex: Regime changes can be subtle; focus on early detection

• Use shorter periods across all indicators
• Monitor CUSUM closely for early warning

Commodities: Often exhibit sharp regime transitions; need robust confirmation

• Use KS Test with stricter threshold (0.01-0.03)
• Increase Wasserstein Distance period for stability

Cryptocurrencies: Highly volatile; need filtering of false signals

• Use longer periods across all indicators
• Increase CUSUM threshold to 2.5-3.0

Regular Recalibration

For optimal performance, recalibrate your indicators:

• Every 3-6 months for most markets
• After significant market events (crashes, major news)
• When you notice deterioration in signal quality

Final Thoughts on Implementation

Remember that market regime indicators are most powerful when:

• Used in combination rather than isolation
• Calibrated specifically to your trading instruments and timeframes
• Integrated into a complete trading approach rather than used as standalone signals
• Regularly reviewed and adjusted as market conditions evolve

In the coming months, we will continue adding similar indicators and aim to deliver tools that help better manage strategies in the current volatile period. We firmly believe that we will also succeed in bringing even more advanced algorithms to help traders become more profitable