Maximum drawdown control system development

Development of Maximum Drawdown Control System

Maximum Drawdown (MDD) is the largest decline from peak to trough of a portfolio over a period. This is a key risk metric: a trader can survive many bad periods, but catastrophic drawdown is psychologically and financially devastating.

Drawdown Calculation

import numpy as np

def calculate_max_drawdown(equity_curve):
    equity = np.array(equity_curve)
    peak = np.maximum.accumulate(equity)
    drawdown = (equity - peak) / peak
    max_drawdown = drawdown.min()
    
    # Find period of maximum drawdown
    end_idx = drawdown.argmin()
    start_idx = equity[:end_idx].argmax()
    
    return {
        'max_drawdown': max_drawdown,
        'start_date_idx': start_idx,
        'end_date_idx': end_idx,
        'drawdown_duration': end_idx - start_idx
    }

def calculate_current_drawdown(equity_history):
    peak = max(equity_history)
    current = equity_history[-1]
    return (current - peak) / peak

System Response Levels

System works on traffic light principle:

Green (drawdown < 50% of limit): normal trading.

Yellow (50–75% of limit): warning. Reduce position size by 25%.

Orange (75–90% of limit): serious warning. Reduce size by 50%. Only high-quality signals.

Red (> 90% of limit): almost reached limit. Close half of open positions.

Stop (= 100% of limit): automatic closure of all positions. Complete trading halt until manual confirmation.

class DrawdownController:
    LEVELS = [
        (0.50, 'yellow', 0.75),   # at 50% of limit: trade at 75% size
        (0.75, 'orange', 0.50),
        (0.90, 'red', 0.25),
        (1.00, 'halt', 0.00)
    ]
    
    def __init__(self, max_drawdown_limit=0.15):
        self.limit = max_drawdown_limit
        self.peak_equity = None
    
    def update_and_check(self, current_equity):
        if self.peak_equity is None or current_equity > self.peak_equity:
            self.peak_equity = current_equity
        
        current_dd = (self.peak_equity - current_equity) / self.peak_equity
        dd_ratio = current_dd / self.limit  # as fraction of limit
        
        for threshold, level, size_mult in reversed(self.LEVELS):
            if dd_ratio >= threshold:
                return {
                    'level': level,
                    'current_dd': current_dd,
                    'dd_ratio': dd_ratio,
                    'size_multiplier': size_mult,
                    'halt': level == 'halt'
                }
        
        return {'level': 'green', 'size_multiplier': 1.0, 'halt': False}

Drawdown Recovery Tracking

Time to recovery: how long it took to recover from previous drawdowns. Helps assess strategy quality.

Recovery factor: total profit / maximum drawdown. Shows how much system "pays for" the risk.

Underwater chart: graph of time portfolio spent in drawdown. Important tool for trader psychological assessment.

Limit Calibration

Typical max drawdown limits by strategy type:

We develop drawdown control system with multi-level response, automatic position reduction, circuit breaker halt and recovery time reporting.