Skip to main content

Market Tournament Framework: A Multi-Perspective Approach to Finding Sweet Spots

· 8 min read
Max Kaido
Max Kaido
Architect

In the ever-evolving landscape of algorithmic trading, finding optimal market conditions—what we call "sweet spots"—remains a constant challenge. Today, we're excited to introduce our innovative Market Tournament Framework, which approaches this challenge through pure indicator-based analysis, removing subjective pattern interpretation.

Core Principles​

  1. Pure Indicator Focus: Eliminate subjective candlestick patterns in favor of quantifiable indicators
  2. Multi-Factor Confirmation: Each setup requires multiple indicators to align
  3. Timeframe Optimization: Each tournament type is optimized for specific timeframes
  4. Risk Management Integration: Built-in position sizing and stop-loss guidelines
  5. Clear Trigger Sequences: Well-defined action signals based on market conditions

Tournament Types (Sweet Spots)​

1. Market Condition Based Tournaments​

These tournaments focus on observable market states, using quantitative metrics to identify opportunities.

STRONG_TREND (Momentum Sweet Spot)​

  • Purpose: Identify strong directional movements with volume confirmation
  • Optimal Timeframe: 4H/1D
  • Key Indicators:
    • RSI (Relative Strength Index) for trend strength
    • MACD (Moving Average Convergence/Divergence) for momentum confirmation
    • Volume Analysis for participation
  • Sweet Spot Characteristics:
    • Strong directional RSI (>70 or <30)
    • MACD histogram expanding
    • Volume increasing with price movement
  • Risk Management:
    • Stop: Beyond the last significant swing
    • Position Size: 1.5x standard size due to trend confirmation

VOLATILITY_BREAKOUT (Compression Sweet Spot)​

  • Purpose: Detect high-probability breakout setups
  • Optimal Timeframe: 1H/4H
  • Key Indicators:
    • Bollinger Bands for volatility contraction
    • Keltner Channels for normal volatility range
    • ATR (Average True Range) for volatility confirmation
  • Sweet Spot Characteristics:
    • Bollinger Bands inside Keltner Channels
    • ATR decreasing to historical lows
    • Price coiling near significant levels
  • Risk Management:
    • Stop: Beyond the compression formation
    • Position Size: Start small, add on breakout confirmation

PRICE_EXTENSION (Overextension Sweet Spot)​

  • Purpose: Identify price exhaustion points
  • Optimal Timeframe: 4H
  • Key Indicators:
    • Distance from moving averages
    • Rate of price change
    • Volume profile analysis
  • Sweet Spot Characteristics:
    • Price significantly deviated from moving averages
    • Rapid price movement with declining volume
    • Extreme readings on momentum indicators
  • Risk Management:
    • Tight stops beyond recent swing
    • Reduced position size due to counter-trend nature

2. Pattern Based Tournaments​

These tournaments identify specific technical formations and indicator relationships.

INDICATOR_DIVERGENCE (Hidden Opportunity Sweet Spot)​

  • Purpose: Spot divergences between price and indicators
  • Optimal Timeframe: 4H/1D
  • Key Indicators:
    • RSI Divergence patterns
    • Price Action structure
    • OBV (On Balance Volume) trends
  • Sweet Spot Characteristics:
    • Price making new lows/highs
    • Indicators showing opposite direction
    • Volume confirming the divergence
  • Risk Management:
    • Stop loss beyond the divergence structure
    • Scale in approach recommended
  • Implementation:
function calculateDivergenceScore(market: Market): number {
  const priceSlope = calculatePriceSlope();
  const rsiSlope = calculateRSISlope();
  const obvSlope = calculateOBVSlope();

  return evaluateDivergenceSignal(priceSlope, rsiSlope, obvSlope);
}

TECHNICAL_CONFLUENCE (Multi-Factor Sweet Spot)​

  • Purpose: Find areas where multiple factors align
  • Optimal Timeframe: 4H
  • Key Indicators:
    • Multiple moving averages
    • Support/Resistance levels
    • Fibonacci retracements
  • Sweet Spot Characteristics:
    • Multiple technical levels converging
    • Volume showing interest at level
    • Clean price action structure
  • Risk Management:
    • Stop below the confluence zone
    • Standard position size with clear invalidation

PATTERN_COMPLETION (Formation Target Sweet Spot)​

  • Purpose: Identify completion of technical formations
  • Optimal Timeframe: 1H/4H
  • Key Indicators:
    • Technical formation completion
    • Volume pattern analysis
    • Time-based factors
  • Sweet Spot Characteristics:
    • Complete formation structure
    • Volume confirming completion
    • Price reaching measured targets
  • Risk Management:
    • Trailing stops based on ATR
    • Reduce exposure near target levels

3. Market Psychology Based Tournaments​

These tournaments analyze supply/demand dynamics and trader behavior patterns.

VOLUME_EXHAUSTION (Capitulation Sweet Spot)​

  • Purpose: Identify potential reversal points through volume analysis
  • Optimal Timeframe: 1H/4H
  • Key Indicators:
    • Volume Profile distribution
    • Price range expansion
    • Momentum oscillators
  • Sweet Spot Characteristics:
    • Volume climax
    • Price exhaustion
    • Extreme sentiment readings
  • Risk Management:
    • Tight stops due to reversal nature
    • Start with minimal exposure, increase on confirmation
  • Implementation:
function calculateExhaustionScore(market: Market): number {
  const volumeProfile = analyzeVolumeProfile();
  const priceRange = analyzePriceRange();
  const momentum = analyzeMomentumOscillators();

  return evaluateExhaustionSignal(volumeProfile, priceRange, momentum);
}

PRICE_EQUILIBRIUM (Balance Point Sweet Spot)​

  • Purpose: Find stable price zones after volatility
  • Optimal Timeframe: 4H
  • Key Indicators:
    • Volume-Weighted Average Price (VWAP)
    • Time-weighted price levels
    • Order flow analysis
  • Sweet Spot Characteristics:
    • Price finding stability
    • Volume profile showing accumulation
    • Reduced volatility after expansion
  • Risk Management:
    • Stop below the equilibrium zone
    • Standard position size with clear levels

SMART_MONEY_ACTIVITY (Distribution Sweet Spot)​

  • Purpose: Detect institutional distribution patterns
  • Optimal Timeframe: 4H/1D
  • Key Indicators:
    • Volume distribution analysis
    • Price action patterns
    • Time in price zones
  • Sweet Spot Characteristics:
    • Increasing volume on small moves
    • Price unable to make new highs
    • Time spent at price levels
  • Risk Management:
    • Reduce exposure on distribution confirmation
    • Adjust protection levels to protect gains

Complementary Tournament Pairs​

One of the framework's key innovations is the concept of complementary tournament pairs. Instead of strict entry/exit pairs, we combine tournaments that look for different but related market conditions.

High Confidence Pairs​

  1. Volatility Cycle Pair

    • Tournaments: [VOLATILITY_BREAKOUT, STRONG_TREND]
    • Market Philosophy: Markets move from low volatility to high volatility
    • Optimal Timeframe: 4H
    • Trigger Sequence:
      1. VOLATILITY_BREAKOUT identifies compression phase
      2. STRONG_TREND confirms directional strength
    • Use Case: Identifying early trend development
    • Risk Management:
      • Tighter stops during compression
      • Wider stops during momentum phase

  2. Technical Confirmation Pair

    • Tournaments: [INDICATOR_DIVERGENCE, TECHNICAL_CONFLUENCE]
    • Market Philosophy: Technical factors align with underlying strength
    • Optimal Timeframe: 4H/1D
    • Trigger Sequence:
      1. INDICATOR_DIVERGENCE spots hidden strength/weakness
      2. TECHNICAL_CONFLUENCE confirms with multiple factors
    • Use Case: High-probability reversal conditions
    • Risk Management:
      • Build position during divergence
      • Increase exposure on confirmation

  3. Accumulation-Distribution Pair

    • Tournaments: [PRICE_EQUILIBRIUM, SMART_MONEY_ACTIVITY]
    • Market Philosophy: Smart money accumulates quietly and distributes loudly
    • Optimal Timeframe: 4H/1D
    • Trigger Sequence:
      1. PRICE_EQUILIBRIUM finds quiet accumulation
      2. SMART_MONEY_ACTIVITY spots institutional selling
    • Use Case: Trend strength to distribution transition
    • Risk Management:
      • Larger position during accumulation
      • Scale out during distribution

Medium Confidence Pairs​

  1. Psychology Shift Pair

    • Tournaments: [VOLUME_EXHAUSTION, PRICE_EQUILIBRIUM]
    • Market Philosophy: Markets move from extreme to balance
    • Optimal Timeframe: 1H/4H
    • Trigger Sequence:
      1. VOLUME_EXHAUSTION identifies extremes
      2. PRICE_EQUILIBRIUM confirms new balance
    • Use Case: Counter-trend reversal trades
    • Risk Management:
      • Minimal exposure at extreme points
      • Increase exposure during stabilization

  2. Pattern Evolution Pair

    • Tournaments: [VOLATILITY_BREAKOUT, PATTERN_COMPLETION]
    • Market Philosophy: Patterns develop from consolidation
    • Optimal Timeframe: 4H
    • Trigger Sequence:
      1. VOLATILITY_BREAKOUT identifies consolidation
      2. PATTERN_COMPLETION targets pattern targets
    • Use Case: Pattern-based trading strategies
    • Risk Management:
      • Increase exposure when pattern forms
      • Reduce exposure at target levels

Experimental Pairs​

  1. Volume Profile Pair

    • Tournaments: [STRONG_TREND, SMART_MONEY_ACTIVITY]
    • Market Philosophy: Strong trends end with distribution
    • Optimal Timeframe: 4H/1D
    • Trigger Sequence:
      1. STRONG_TREND identifies strong moves
      2. SMART_MONEY_ACTIVITY spots distribution phase
    • Use Case: Trend strength to distribution transition
    • Risk Management:
      • Full position during momentum phase
      • Reduce exposure on distribution

  2. Reversal Anticipation Pair

    • Tournaments: [VOLUME_EXHAUSTION, PRICE_EXTENSION]
    • Market Philosophy: Extreme moves lead to reversals
    • Optimal Timeframe: 1H/4H
    • Trigger Sequence:
      1. PRICE_EXTENSION identifies extreme moves
      2. VOLUME_EXHAUSTION confirms reversal potential
    • Use Case: Counter-trend trading
    • Risk Management:
      • Small positions at extremes
      • Add on confirmation

Configuration Interface​

interface TournamentConfig {
  // Core identification
  version: string;
  name: TournamentType;
  description: string;
  timeframes: {
    primary: string; // Main timeframe for analysis
    secondary: string[]; // Supporting timeframes
    correlation: {
      type: 'CONFIRMATION' | 'DIVERGENCE';
      rules: string[]; // How timeframes interact
    };
  };

  // Technical setup
  indicators: {
    primary: string[]; // Indicators on primary timeframe
    secondary: string[]; // Indicators on secondary timeframes
  };
  conditions: {
    primary: string[]; // Must-have conditions
    filters: string[]; // Additional filters
    timeframeRules: string[]; // Multi-timeframe conditions
  };

  // Comparison logic
  comparisonPrompt: {
    systemPrompt: string;
    temperature: number;
    maxTokens: number;
  };

  // Risk management
  riskManagement: {
    prompt: string;
    stopLossType: string;
    defaultSize: number;
    timeframeAdjustment: {
      sizeMultipliers: Record<string, number>; // Size adjustments per timeframe
      stopMultipliers: Record<string, number>; // Stop adjustments per timeframe
    };
  };

  // Scoring & optimization
  scoring: {
    weights: Record<string, number>;
    minimumScore: number;
    targetScore: number;
    timeframeWeights: Record<string, number>; // Weight for each timeframe's contribution
  };

  // Version tracking
  metadata: {
    created: string;
    author: string;
    performance?: {
      accuracy: number;
      sampleSize: number;
      period: string;
    };
    replacedBy?: string;
    replaces?: string;
    reasoning: string;
  };
}

Example Configuration (v2.ts)​

import { TournamentConfig, TournamentType } from '../types';

export const strongTrendV2: TournamentConfig = {
  version: 'v2',
  name: TournamentType.STRONG_TREND,
  description: 'Strong trend with multi-timeframe confirmation',
  timeframes: {
    primary: '4h',
    secondary: ['1d', '1h'],
    correlation: {
      type: 'CONFIRMATION',
      rules: ['1D trend direction matches 4H', '1H shows no reversal signals'],
    },
  },

  indicators: {
    primary: ['rsi', 'macd', 'volume', 'adx'],
    secondary: ['supertrend', 'ema200'],
  },
  conditions: {
    primary: [
      'RSI > 50 && rising',
      'MACD histogram > 0 && expanding',
      'Volume > 2 * avgVolume',
    ],
    filters: ['ADX > 25', 'No major resistance within 5%'],
    timeframeRules: [
      '1D Supertrend aligned with 4H direction',
      '1H above 200 EMA',
    ],
  },

  riskManagement: {
    prompt: `Determine stop loss and take profit based on multi-timeframe analysis...`,
    stopLossType: 'SWING_LOW',
    defaultSize: 1.5,
    timeframeAdjustment: {
      sizeMultipliers: {
        '4h': 1.0,
        '1d': 1.5,
        '1h': 0.75,
      },
      stopMultipliers: {
        '4h': 1.0,
        '1d': 1.5,
        '1h': 0.75,
      },
    },
  },

  scoring: {
    weights: {
      rsi: 0.3,
      macd: 0.3,
      volume: 0.2,
      adx: 0.2,
    },
    minimumScore: 0.65,
    targetScore: 0.8,
    timeframeWeights: {
      '4h': 0.6,
      '1d': 0.3,
      '1h': 0.1,
    },
  },

  metadata: {
    created: '2025-03-15',
    author: 'maxkaido',
    performance: {
      accuracy: 0.72,
      sampleSize: 1000,
      period: '2025-01-01/2025-03-15',
    },
    replaces: 'v1',
    reasoning: 'Added multi-timeframe confirmation logic',
  },
};

Future Development​

We're currently working on:

  1. Machine learning models to optimize tournament pair selection
  2. Dynamic timeframe adaptation based on market conditions
  3. Integration with our shadow portfolio system for automated validation

Conclusion​

The Market Tournament Framework represents a significant evolution in our approach to market analysis. By combining multiple perspectives and leveraging complementary tournament pairs, we've created a robust system for identifying trading opportunities. The framework's strength lies in its pure indicator-based approach, clear timeframe optimization, and integrated risk management guidelines.

Stay tuned for future updates as we continue to refine and expand this framework.

Other articles:

Implementation Details​

The framework is implemented in TypeScript and integrated into our DIKE module. Here's a high-level overview of the architecture:

export enum TournamentType {
  // Market Condition Based
  STRONG_TREND = 'STRONG_TREND',
  VOLATILITY_BREAKOUT = 'VOLATILITY_BREAKOUT',
  PRICE_EXTENSION = 'PRICE_EXTENSION',

  // Pattern Based
  INDICATOR_DIVERGENCE = 'INDICATOR_DIVERGENCE',
  TECHNICAL_CONFLUENCE = 'TECHNICAL_CONFLUENCE',
  PATTERN_COMPLETION = 'PATTERN_COMPLETION',

  // Psychology Based
  VOLUME_EXHAUSTION = 'VOLUME_EXHAUSTION',
  PRICE_EQUILIBRIUM = 'PRICE_EQUILIBRIUM',
  SMART_MONEY_ACTIVITY = 'SMART_MONEY_ACTIVITY',
}

Configuration Directory Structure

Overview​

The tournament configurations are organized in the DIKE module following a clear, versioned structure that maintains separation of concerns while keeping related configurations together.

Directory Layout​

dike/
└── tournaments/
    ├── types/
    │   ├── tournament.types.ts        # Core type definitions
    │   ├── scoring.types.ts           # Scoring related types
    │   └── risk-management.types.ts   # Risk management types
    │
    ├── configs/
    │   ├── market-condition/          # Market condition based tournaments
    │   │   ├── strong-trend/
    │   │   │   ├── v1.ts
    │   │   │   ├── v2.ts             # Current version
    │   │   │   └── metadata.json     # Version tracking & performance
    │   │   ├── volatility-breakout/
    │   │   └── price-extension/
    │   │
    │   ├── pattern-based/            # Pattern based tournaments
    │   │   ├── indicator-divergence/
    │   │   ├── technical-confluence/
    │   │   └── pattern-completion/
    │   │
    │   └── psychology-based/         # Psychology based tournaments
    │       ├── volume-exhaustion/
    │       ├── price-equilibrium/
    │       └── smart-money-activity/
    │
    ├── pairs/                        # Tournament pair configurations
    │   ├── high-confidence/
    │   │   ├── volatility-cycle.ts
    │   │   ├── technical-confirmation.ts
    │   │   └── accumulation-distribution.ts
    │   ├── medium-confidence/
    │   └── experimental/
    │
    ├── utils/
    │   ├── scoring.ts                # Scoring utilities
    │   ├── timeframe.ts             # Timeframe management
    │   └── validation.ts            # Configuration validation
    │
    └── index.ts                      # Public API exports

File Structure Examples​

Tournament Configuration (v2.ts)​

import { TournamentConfig } from '../../types/tournament.types';
import { baseConfig } from '../base-config';

export const strongTrendV2: TournamentConfig = {
  ...baseConfig,
  version: 'v2',
  name: TournamentType.STRONG_TREND,
  // ... configuration as shown in previous example
};

Pair Configuration (volatility-cycle.ts)​

import { TournamentPairConfig } from '../../types/tournament.types';
import { VOLATILITY_BREAKOUT, STRONG_TREND } from '../configs';

export const volatilityCyclePair: TournamentPairConfig = {
  name: 'VOLATILITY_CYCLE',
  confidence: 'HIGH',
  tournaments: [VOLATILITY_BREAKOUT, STRONG_TREND],
  timeframes: {
    primary: '4h',
    secondary: ['1d', '1h'],
  },
  correlation: {
    type: 'CONFIRMATION',
    rules: ['1D trend direction matches 4H', '1H shows no reversal signals'],
  },
  scoring: {
    timeframeWeights: {
      '4h': 0.6,
      '1d': 0.3,
      '1h': 0.1,
    },
  },
};

Base Configuration (base-config.ts)​

import { BaseConfig } from '../types/tournament.types';

export const baseConfig: BaseConfig = {
  comparisonPrompt: {
    temperature: 0.3,
    maxTokens: 500,
  },
  scoring: {
    minimumScore: 0.65,
    targetScore: 0.8,
  },
  // ... other common configuration
};

Benefits​

  1. Clear Organization:

    • Tournaments grouped by type (market condition, pattern, psychology)
    • Each tournament type has its own directory
    • Clear separation between types, pairs, and utilities

  2. Version Control:

    • Each version is a separate file
    • Metadata tracks performance and changes
    • Easy to roll back or compare versions

  3. Reusability:

    • Base configurations can be shared
    • Common utilities centralized
    • Types ensure consistency

  4. Maintainability:

    • Clear import paths
    • Separated concerns
    • Easy to find and modify configurations

  5. Scalability:

    • Easy to add new tournament types
    • Simple to create new pairs
    • Flexible for future extensions

  6. Documentation:

    • Structure mirrors the conceptual framework
    • Self-documenting organization
    • Easy to navigate and understand

This structure follows the Mercury Pantheon conventions while providing a clear, maintainable organization for our tournament configurations.