trading_bot_v4/scripts/trade_analysis.py

#!/usr/bin/env python3
"""
Trade-Level Analysis - Find Patterns in Losing Trades
Purpose: Identify what conditions lead to losses
"""
from __future__ import annotations

import argparse
import sys
from pathlib import Path
from typing import List, Dict

PROJECT_ROOT = Path(__file__).resolve().parents[1]
if str(PROJECT_ROOT) not in sys.path:
    sys.path.append(str(PROJECT_ROOT))

import pandas as pd
from pathlib import Path
from tqdm import tqdm
from backtester.data_loader import load_csv
from backtester.indicators.money_line import MoneyLineInputs, money_line_signals
from backtester.simulator import TradeConfig, simulate_money_line, SimulatedTrade


def analyze_trades(trades: List[SimulatedTrade]) -> Dict:
    """Deep analysis of trade performance."""
    if not trades:
        return {"error": "No trades to analyze"}
    
    winners = [t for t in trades if t.pnl > 0]
    losers = [t for t in trades if t.pnl < 0]
    
    print("\n" + "="*80)
    print("TRADE PERFORMANCE BREAKDOWN")
    print("="*80)
    print(f"Total Trades: {len(trades)}")
    print(f"Winners: {len(winners)} ({len(winners)/len(trades)*100:.1f}%)")
    print(f"Losers: {len(losers)} ({len(losers)/len(trades)*100:.1f}%)")
    
    if winners:
        print(f"\nWinner Stats:")
        print(f"  Avg PnL: ${sum(t.pnl for t in winners)/len(winners):.2f}")
        print(f"  Max Win: ${max(t.pnl for t in winners):.2f}")
        print(f"  Avg MFE: {sum(t.mfe for t in winners)/len(winners):.2f}%")
        print(f"  Avg Bars: {sum(t.bars_held for t in winners)/len(winners):.1f}")
    
    if losers:
        print(f"\nLoser Stats:")
        print(f"  Avg PnL: ${sum(t.pnl for t in losers)/len(losers):.2f}")
        print(f"  Max Loss: ${min(t.pnl for t in losers):.2f}")
        print(f"  Avg MAE: {sum(t.mae for t in losers)/len(losers):.2f}%")
        print(f"  Avg Bars: {sum(t.bars_held for t in losers)/len(losers):.1f}")
    
    # Total P&L stats
    total_pnl = sum(t.pnl for t in trades)
    print(f"\nTotal P&L: ${total_pnl:.2f}")
    print(f"Avg Trade: ${total_pnl/len(trades):.2f}")
    
    # MAE/MFE Analysis
    print("\n" + "="*80)
    print("MAE/MFE ANALYSIS (Max Adverse/Favorable Excursion)")
    print("="*80)
    
    # Winners that gave back profit
    if winners:
        large_mfe_winners = [t for t in winners if t.mfe > 2.0]
        if large_mfe_winners:
            avg_giveback = sum(t.mfe - (t.pnl / 8100.0 * 100) for t in large_mfe_winners) / len(large_mfe_winners)
            print(f"Winners with >2% MFE: {len(large_mfe_winners)}")
            print(f"  Average profit given back: {avg_giveback:.2f}%")
            print(f"  💡 Consider: Wider trailing stop or earlier TP2 trigger")
    
    # Losers that could have been winners
    if losers:
        positive_mfe_losers = [t for t in losers if t.mfe > 0.5]
        if positive_mfe_losers:
            avg_peak = sum(t.mfe for t in positive_mfe_losers) / len(positive_mfe_losers)
            print(f"\nLosers that reached >0.5% profit: {len(positive_mfe_losers)}")
            print(f"  Average peak profit: {avg_peak:.2f}%")
            print(f"  💡 Consider: Tighter TP1 or better stop management")
    
    # Direction analysis
    print("\n" + "="*80)
    print("DIRECTION ANALYSIS")
    print("="*80)
    
    longs = [t for t in trades if t.direction == 'long']
    shorts = [t for t in trades if t.direction == 'short']
    
    if longs:
        long_wr = sum(1 for t in longs if t.pnl > 0) / len(longs) * 100
        long_pnl = sum(t.pnl for t in longs)
        print(f"LONGS: {len(longs)} trades, {long_wr:.1f}% WR, ${long_pnl:.2f} PnL")
    
    if shorts:
        short_wr = sum(1 for t in shorts if t.pnl > 0) / len(shorts) * 100
        short_pnl = sum(t.pnl for t in shorts)
        print(f"SHORTS: {len(shorts)} trades, {short_wr:.1f}% WR, ${short_pnl:.2f} PnL")
    
    if longs and shorts:
        if long_wr > short_wr + 10:
            print(f"  💡 LONGs outperform: Consider quality threshold adjustment")
        elif short_wr > long_wr + 10:
            print(f"  💡 SHORTs outperform: Consider quality threshold adjustment")
    
    return {
        "total": len(trades),
        "winners": len(winners),
        "losers": len(losers),
        "win_rate": len(winners) / len(trades) * 100,
        "total_pnl": total_pnl
    }


def find_exit_opportunities(trades: List[SimulatedTrade]):
    """Analyze if different exit strategy could improve results."""
    print("\n" + "="*80)
    print("EXIT STRATEGY ANALYSIS")
    print("="*80)
    
    # Current strategy stats
    tp1_hits = sum(1 for t in trades if t.exit_type == 'tp1')
    tp2_hits = sum(1 for t in trades if t.exit_type == 'tp2')
    sl_hits = sum(1 for t in trades if t.exit_type == 'sl')
    max_bars = sum(1 for t in trades if t.exit_type == 'max_bars')
    
    print(f"Current Exit Distribution:")
    print(f"  TP1: {tp1_hits} ({tp1_hits/len(trades)*100:.1f}%)")
    print(f"  TP2: {tp2_hits} ({tp2_hits/len(trades)*100:.1f}%)")
    print(f"  SL: {sl_hits} ({sl_hits/len(trades)*100:.1f}%)")
    print(f"  Max Bars: {max_bars} ({max_bars/len(trades)*100:.1f}%)")
    
    # Simulate tighter TP1
    could_take_tp1 = [t for t in trades if t.mfe >= 0.5 and t.pnl < 0]
    if could_take_tp1:
        saved_pnl = len(could_take_tp1) * 0.005 * 8100  # 0.5% profit instead of loss
        print(f"\n💡 Tighter TP1 (0.5%): Would save {len(could_take_tp1)} losing trades")
        print(f"   Estimated improvement: +${saved_pnl:.2f}")
    
    # Check if runners are worth it
    tp2_trades = [t for t in trades if t.exit_type == 'tp2']
    if tp2_trades:
        tp2_avg = sum(t.pnl for t in tp2_trades) / len(tp2_trades)
        print(f"\nTP2/Runner Performance:")
        print(f"  {len(tp2_trades)} trades reached TP2")
        print(f"  Average TP2 profit: ${tp2_avg:.2f}")
        if tp2_avg < 100:
            print(f"  💡 Runners not adding much value - consider 100% TP1 close")


def compare_to_baseline(csv_path: Path, symbol: str, best_inputs: MoneyLineInputs):
    """Compare best config to baseline."""
    print("\n" + "="*80)
    print("BASELINE COMPARISON")
    print("="*80)
    
    print("\nLoading data and running simulations...")
    data_slice = load_csv(Path(csv_path), symbol, "5m")
    df = data_slice.data
    config = TradeConfig(position_size=8100.0, max_bars_per_trade=288)
    
    baseline = MoneyLineInputs()
    print("\nRunning baseline config...")
    baseline_result = simulate_money_line(df, symbol, baseline, config)
    print(f"  Trades: {len(baseline_result.trades)}")
    print(f"  PnL: ${baseline_result.total_pnl:.2f}")
    print(f"  WR: {baseline_result.win_rate:.1f}%")
    
    print("\nRunning best sweep config...")
    best_result = simulate_money_line(df, symbol, best_inputs, config)
    print(f"  Trades: {len(best_result.trades)}")
    print(f"  PnL: ${best_result.total_pnl:.2f}")
    print(f"  WR: {best_result.win_rate:.1f}%")
    
    improvement = best_result.total_pnl - baseline_result.total_pnl
    print(f"\nImprovement: ${improvement:.2f} ({improvement/baseline_result.total_pnl*100:.1f}%)")
    
    return best_result.trades


def main():
    parser = argparse.ArgumentParser(description="Trade-level analysis")
    parser.add_argument("--csv", type=Path, required=True, help="Path to OHLCV CSV")
    parser.add_argument("--symbol", default="SOL-PERP", help="Symbol name")
    args = parser.parse_args()
    
    print("="*80)
    print("V9 TRADE ANALYSIS")
    print("="*80)
    
    # Use "best" config from sweep
    best_inputs = MoneyLineInputs(
        flip_threshold_percent=0.6,
        ma_gap_threshold=0.35,
        momentum_min_adx=23.0,
        momentum_long_max_pos=70.0,
        momentum_short_min_pos=25.0,
        cooldown_bars=2,
        momentum_spacing=3,
        momentum_cooldown=2
    )
    
    trades = compare_to_baseline(args.csv, args.symbol, best_inputs)
    analyze_trades(trades)
    find_exit_opportunities(trades)
    
    print("\n" + "="*80)
    print("ACTIONABLE RECOMMENDATIONS")
    print("="*80)
    print("Based on this analysis, the next optimization steps should focus on:")
    print("1. Exit strategy improvements (TP1/TP2 levels)")
    print("2. Direction-specific quality thresholds")
    print("3. Stop loss positioning")
    print("4. Entry timing refinement")


if __name__ == "__main__":
    main()